Projects
![]()
No projects found
Investors
New
DAO Treasuries
DATs
Lend / Borrow APRs
Chains
Categories
Spaces
#FinanceDaily: NVDA surges | Elon visits Beijing | AI hype or reality?
Recorded: May 31, 2023 Duration: 2:00:12Space Recording
Short Summary
The discussion highlights the launch of a new crypto-focused show, Crypto Town Hall, and strategic partnerships by IBC to support AI and Web3 startups. Trends in the crypto space include a shift towards citizen journalism, innovation during bear markets, and the emergence of new narratives. The absence of hype suggests a focus on substantive developments, with innovations like Bitcoin's use as a software platform gaining attention.
Full Transcription
Hey everybody. Just waiting a few minutes for everybody show up, but yeah, today should be interesting. It was like, should we talk about AI specifically, or should we talk about AI and how China is getting completely annihilated right now? I was trying to figure out which one to focus on, but...
their economy is in just complete disarray.
And maybe that's another topic for another day.
But kind of the intersection of Elon going to Beijing was interesting.
All right.
I'm not fucking thinking Kucci King.
Hell no.
Yeah.
Am I going to have to keep referring to Koochie King?
Can I just call you CK?
Is that okay?
Because I literally can't imagine doing that.
Yeah, let's go with that.
Let's go with that.
Yeah.
My wife in the back is going to be like, I'm sorry, what's going on?
A finance talk do you say?
Hey Jeff, I'm going to be relying quite heavily on your Tesla analysis today because I think a lot of people are asking the question, why is Elon in Beijing?
Why is it making the stock rise so much?
I think I know why.
I know there are rumors why, but I want you to, like, we want to be evidence-based here quite important for us to be.
So, you know, we're going to have to be very clear on what we actually know because Tesla stock moves quite aggressively.
So, you know, just want to make sure.
Yeah, we're just going to wait maybe another minute and then get started.
We're still waiting for folks to, we're starting to get there.
Yeah, let's give it another minute.
I wonder if I do the sponsor talk right now, whether I won't have to do it during.
Yeah.
I literally got a message in the back end from Romi, who's like one of the producers.
And she's literally just like, nope, that won't work.
God damn it.
I just want to be able to do that.
I hate sponsorship talks.
But I do enjoy trying to find a moment when no one's expecting it to throw it in and try to act like it all kind of flows together.
Oh, Mish is here.
Great.
And then Trevor's here.
Awesome.
We've got a full list here.
All right.
But I will say that for folks that are listening,
just so that everybody is on board on how we run our spaces,
if you're listening and you find something that you disagree with
or even agree with but have additional insight,
please go to the bottom right and go into the comments.
You can see the little comment bubble.
It looks like messages.
And go in the bottom right, put in a comment,
We often put the comments up in the nest, and we make sure to actually highlight the comments that were insightful or interesting, and then often we'll bring up the person that put in that comment.
It's a little bit different, right?
We think of ourselves, this is the beauty of spaces, is that you can do something like that.
And, you know, there are often speakers that have been able to.
kind of earn their way up.
You know, often our speakers are actually like experts,
but sometimes if you're insightful,
we learn that you are actually an expert
because you've been watching it.
And that's the beauty of citizen journalism.
We've seen this over and over again in our spaces
and many of the people that are up
have actually kind of quote unquote earned their way up.
And that's kind of how we run our spaces anyway.
So please go on the bottom right, put in comments, and please reach out.
Also, if you like what you're hearing, follow some of the people.
You know, it really does help them build their reach.
All right, we're getting started.
So, you know, for the past few weeks, the entire stock market has been going sideways.
I think across the board, we have seen a slight rise.
when you look at the breadth of the market,
it's been going sideways.
But over the course of the last few weeks,
we've also seen the rise of one big heavyweight.
And that's NVIDIA.
We keep seeing NVIDIA really kind of making a move
that is unprecedented.
Whenever I see something like that,
my spidey senses go off,
and I start thinking to myself,
huh, this is probably hype.
So what I want to understand...
And today's goal is going to be, can we understand what has led NVIDIA to this?
Why is AI suddenly, you know, AI has been around forever.
Why is it suddenly becoming center stage not only in terms of what people can do,
but why it's a good financial decision to invest in AI?
And I want to hear the bear case about why AI is just like every other hype cycle
like, you know, Metaverse, Web3, and others,
and whether, and I want to hear the barricades on why NVIDIA is,
is overpriced right now.
Now, you know, I do want to hear both sides of it.
If I don't hear both sides of it, I will try to bring people up that do believe, because I think it's easy to be the bull case, in my opinion.
AI is everywhere.
Everyone's talking about it.
And I will start with, I'll go to Jeff in a second to kind of give us a rundown in NVIDIA specifically.
But, you know, I think the...
there's clearly been a move towards AI.
You know, I remember tweeting, AI is going to eat software.
And it's because something has changed in AI in the last couple of years.
Number one thing that I believe has changed in AI is that we've gone from just AI as an analytics tool to a generative tool.
And I think that has had significant impact.
But why NVIDIA specifically, and how did this company, that by the way, most people were hating on last year, how did it suddenly become a trillion dollar company?
So, Jeff, I want you to lay it out for us a little bit on your, I know that you're probably an NVIDIA bowl, or an AI bowl at least.
I don't know if you're going to speak to the pricing of it, but more just what is going on that has brought NVIDIA to this point?
Well, Nvidia has had a history of being many different things, right?
Started off as kind of a gaming play.
Then it went to a crypto play.
Then it went to a data center play.
And now it's gone to this AI play.
And, you know, and through its life, you know, it's up.
I think over the last decade, it's done over 100x.
in the last decade.
It's one of the most successful stocks, number one, actually, over the last 10 years.
I think AMD, actually, is number two, and Tesla was three or four.
So those are the transitions it's made.
It's like in 2018, it cut in half last year.
I think in the 52-week range.
It hasn't it gone down to like 108, so it's basically quadrupled in the last...
year on the stock.
And so what happened recently is basically there's been,
obviously this NASDAQ meltup,
where everybody was basically caught off guard.
You know, every talking hat on TV,
you know, coming into this year was not saying go in
and buy, you know, large cap, you know,
or by tech or by NASDAQ.
Everybody was staying away from it.
And sure enough, you know, the NASDAQ surprised,
Obviously, there is the chat GPT kind of moment in November, December of last year,
and everybody's like, okay, what's powering that?
What's the hardware behind that?
And obviously, that's Nvidia.
And then if you look at it in terms of picks and shovels, however you want to look at it,
there really isn't a competitor that's anywhere near them in terms of supplying,
this hardware and silicon.
And I think, you know, and again, they've, you know, this silicon they've had for a bit of time.
They've been at this for a while, but they've been really basically building up their supply of it and kind of, you know, shaking off data center issues from last year.
So, yeah.
They melted up with the NASDAQ.
I would say Jensen, obviously, is great product person, great product visionary, great technologists.
Also, he's been a great steward of his managing a stock price.
I mean, they've basically melted up with the NASDAQ, obviously with the chat GPT moment.
But this whole move in the last two weeks with earnings, I've talked about this.
By the way, again, I've been ahead of supply chain for three different Fortune 50 companies.
And I bought Silicon from Nvidia before.
And I know, I mean, if I just look at this current quarter, right, there's seven weeks into the current quarter and they had their earnings and they raised their guidance, right?
They raised it from $7 billion to $11 billion.
And I know when those orders had to go in, those orders had to go in January probably.
And, you know, there's probably some that dropped in in February, but they had a pretty good idea that,
their Q2, or their, they should call it their April to June end period, was going to be a blowout months ago.
But I think they timed it properly.
They timed the release of information properly.
And it's really just...
help cat up they've been great stewards of managing their stock yeah it's interesting because
the the bulk case in my opinion on invidia is that no matter which a i company wins a i uh
vizia wins but that's not completely true because google is not using nvedia for their chips correct
Yeah, Google has their own proprietary.
But the underlying base hypothesis here is that Google is not going to be the AI winner.
And I don't know, man.
I don't like betting against Google.
But I do think that that is the underlying assumption here.
Why not AMD?
Like, why aren't they getting any attention?
Because AMD is not good for AI.
At least it's not good.
Oh, shit.
We're about to get into it.
All right.
No, that's good.
I want to know that.
Why is, why is, I don't have an opinion on that at all.
I just want to know why AMD is not good for AI.
Yeah, so a lot of...
Eugene, go ahead.
Yeah, sure.
So a lot of the...
Well, first, by the way, I think the near-term bear case, I think overall for AI, I am bearish in the short term, but overall in the long term, meaning I think it's being undervalued in the long run, like five to ten years.
But in the short run, I think there's quite a, you know, I think there's quite a bubble.
And in the Gartner hype cycle, we're about to, you know, we're getting to the peak of inflated expectations and we'll soon enter perhaps the drop of disillusionment.
But to answer the question, yeah, I mean, Justin, you asked why not AMD.
You know, AMD can be used.
And by the way, isn't AMD the company so glad it bought ATI, which was the precursor to the current graphics chip set in the early 2000s, right?
Can we imagine if it was a one horse race right now with just Nvidia?
So at least there is a two horse race.
So any gamers in the audience will know that, you know, AMD and Nvidia are kind of the two, you know, the two rivals, right?
And a lot of gamers will say, hey, you know, should I get an AMD or an Nvidia?
And from a consumer side, I'd actually advise probably, you know, if you can get your hands on a good AMD, you know,
a graphics processor for a gamer, probably do that because the value for the money is probably better for AMD this generation versus the Nvidia 40 series.
I think...
the speaking of the nvdia 40 series the bear case is actually in
NVIDIA is sitting on quite a lot of 40 series graphics cards.
It has actually quite an oversupply of them in certain respects,
even though it's really hard to get a 40-90,
which is the top end of the 40 series.
So, you know, a lot of the NVIDIA stuff right now
is better run on NVIDIA graphics cards
because of the way that the graphics cards are architected,
because of the way the programs are,
and because they're just better, right?
So for the last several years,
Nvidia Tesla, the chip, not the car,
Nvidia Tesla GPs like the V100 were the ones that were really good for AI applications, right?
Because for AI, you need a lot of V RAM.
So it's only actually in the more recent times, actually 30 series, like the 3080 TI,
and now the 49D, which is the flagship product for AMD that, sorry for Nvidia,
that in fact we're getting a lot of,
it has 24 gigabytes of VRAM basically.
So it makes it great for things like stable diffusion,
which is what's behind mid-Journey,
makes great for,
you know,
GPT4,
including things like chat GPT.
So basically it's just a great,
you know,
it's just a really excellent,
even consumer card.
We're actually seeing the best AI card
and also the best consumer card for gaming
and it's actually all coming together.
so yeah,
so that's why Nvidia's great,
but also they have too many of those,
a card sitting around that I think in the near term they might have some pricing pressure on the way down.
Can AMD just adapt? I mean, maybe that sounds like an idiot question, but you know, you should
treat me like an idiot because I don't know how chips are filmed. Yeah, I mean, sorry, who said that? Someone said
something. But I was going to say, you know, can they adapt? Caleb, go ahead. Were you,
were you weighing in? Yeah, I mean, any of these semiconductor companies can adapt. I think
You know, I don't have any teen insight on AMD specifically versus
Invidia.
I've done deep dives on both, but I'm not an expert in the semiconductor space by any means.
Let's go back to the initial question here, which is Bear versus Bull on
Nvidia.
I can present both cases because I think it's kind of a strange position to be in.
This is a company I was pitching in the fourth quarter of last year.
has an amazing opportunity to be buying this stock, right?
And so obviously I de-risk a significant portion of my position in Q1 of this year.
So, you know, I also shot myself in the foot by exiting Nvidia a little bit too soon.
But nonetheless, I'd say in the short term, we're in a bubble.
That does not mean that the bubble implodes tomorrow.
So I expect pretty significant price appreciation for both AMD and Nvidia in the days, weeks, and potentially even months ahead.
But here's a thing.
The company right now is trading at a price to sales ratio of 38%.
That is obscene.
Also, their price to earnings ratio is almost at 230.
I thought it was pretty obscene at 150.
This is getting crazy.
And what I said when the stock was trading at a PE of 150 is that this can go way higher.
And at the end of the day, I think investors should be thankful that this company even has earnings, right?
Because sometimes you can get a price to earnings ratio of basically infinity because a company has a really high price, but they have no earnings.
So think of Tesla six years ago.
Right. Just because a company has a high PE or a high price to sales or an NA price of sales,
right, not applicable does not mean that the stock is bad or that the stock can't keep going
higher. And it certainly doesn't mean that the stock is going to fall tomorrow, right? So I think
that's super important to understand. So I'd say right now I'm pretty short term bullish on the
stock. I'd say over the next maybe...
Do you mean bearish?
I just want to make sure.
No, no, I'm bullish on the stock in the short term.
I think over the next four to eight months, though, if the market starts to show some signs of capitulation, stocks like NVIDIA are going to give up the most, right?
Because it's way harder to justify a price to earnings ratio of 230 or a price to sales of 38.
Right.
Right. So this is this is where it gets very tricky.
So I would generally classify myself as being short-term bullish.
We're in a bubble. So ride the wave, right? It's here.
So the music's playing. You have to dance.
However, at some point, I don't know when that's going to be.
At some point, investors are going to start to question the valuation here and start to question that valuation
relative to the broader market and other opportunities that are still available in the semiconductor industry.
Right.
So in the medium term, I'd say there's justification to be bearish on the stock.
However, over the long run, this is going to be the clear winner because semiconductors as a whole, whether it's AMD,
and Vindia, you can even look at companies like Lamb Research, Qualcomm, Broadcom.
All of these companies are massive beneficiaries of a 5, 10, 15 year AI wave, right?
If you think this is a bubble now, wait until you see what it looks like 10 years from now.
Right? Seriously.
But we go ahead.
Get your financial advice thing in there, Donish.
I will in a second.
Sorry, stock.
Go ahead.
So on a rebound, and especially during these mini bubbles that we're seeing with the precipice of the AI bubble is about to take over and consume about 30 to 40% of all jobs in corporate America with robotic process automation and just workflow automation on the back end.
You have to understand, too, we should be used in forward PE for Navidia, which is right around about 69 is right now, as they guided up $4 billion.
So that's about a 46% increase.
And so you have to understand that they're about to take over a majority of the jobs out there.
And then you have to also just add in the fact that Intel's roadmap with the HPE servers, Dell servers, in the past has eroded completely.
Up until about 2017, 2018, they had a great relationship.
But then you saw Spectrum Meltdown, which is where it's completely displaced these Intel servers, the old X-Aad6 servers.
And so now you see AMD and the video, which...
are completely taken over that market share.
And so that's why we've seen Intel lag,
and we've seen AMD in the video,
just sore 2,3X the last year and a half, you know?
So you just have to understand that there's a bigger picture too with this.
And then should just give you an idea too.
So, Navidia, they're technically like the graphic processing unit.
It's very expensive chip,
but it's the standard in the data center
because you're able to actually use it as like the brain.
So up until early 2020,
Tesla actually use Navidia for their full self-drive, but when you consider that it's about $17,000, $18,000 for a chip, how can you justify a $35,000 vehicle when that's the main component of your vehicle?
You know, and so you have to understand that they're trying to figure out how are we going to create the brain, because that's really what the chip is of AI.
And so that's why Navidia is the leader.
And then AMD is more so...
They've been able to take over in the data center side of the house.
So for the servers, they've created a great roadmap.
Intel was completely displaced.
A&B has been able to supplement them.
So that's why you've seen a rotation from Intel,
which has been sold off, barely even been moving,
into this huge disruption of AMD and the video.
So I just wanted to stop there,
because I think Ford P is something big that we're
I was looking at P.E. during the rebound in October, January time frame, but now we need to use forward P.E.
Because what we're seeing is like, so they already devalued, de-risk all of these stocks.
Now, what's the forward P.E. of these going for? Because look at CRM, it has like a thousand P.E.
And they're going to report today. So I'm very curious to see how they do, because they're going to talk about their Einstein, their AI component.
So it's interesting to see how the market's going to reprice them going forward.
Speaking of Intel, yeah, I think Intel is like far in the third horse in the race by far when it comes to AI for their, you know, their version of their GPUs.
You know, they've been struggling a lot since Apple moved away from them with the, you know, the M1, Apple Silicon chips and now the M2.
You know, AMD, actually, you know, going back to that, it's not, you know, the market's not sleeping on them.
I mean, they are up something like 17% or so the last five days.
So, you know, they're riding some of that AI and video wave.
But to kind of put a fine point on things.
InVita about seven, eight years ago, they did a really smart thing.
They actually brought out, you know, in deep learning, machine learning and sort of these researchers around the world who didn't have lots of money, right?
They were mostly in research universities.
Now they're mostly in AI companies.
A lot of them moved over.
But these PhDs and these universities got these Nvidia at GPUs.
to use for things like DLSS and stuff like that.
And so what NVIDA uses is KudaCorps,
AMD has their own version of that.
But basically an entire library of stuff
has been developed based on the NVIDIA architecture,
the Kudakores.
And that happened because of 70 years ago,
all these people got free NVIDIA GPUs
to develop their AI machine learning stuff.
So basically that's kind of the reason
why NVIDA has a head start.
I'd say Eli 5, like the simple version is
kind of like, Nvidia kind of created like a closed,
Sort of closed architecture, it called semi-closed, and there's an open version of the architecture that AMD can utilize.
But because of that closed architecture, Nvidia, GPs can take advantage.
And even though AMD on a performance per watt is better, right?
The 4090 by Nvidia is a power-hungry chip.
Not great from an energy perspective, but it does run AI better, right?
So if AMD had those same libraries or could access those,
you know, it would be, you know, you could use those for, you know, pretty efficiently as an AI deep learning developer.
But as of now, Nvidia is your best bet from a cost to performance perspective.
I will say one thing before Scott goes on to respond, AMD actually,
has done something that Jeff was saying earlier, which is stock management, earnings management, having that conversation.
AMD has actually done a really poor job of that.
You know, missing, they closed more than 9% lower on May 3rd because they were disappointed.
You know, the market was disappointed with the results.
And so, you know, they missed their earnings.
And what's really interesting is NVIDIA during the last year or so was incredibly bearish on their outlook.
And then suddenly came out like hitting hard like Jeff was saying.
And I think it speaks to leadership and how to play the market correctly.
It was very interesting because AMD, if you look at it on a full trajectory,
is actually not doing that poorly.
And the AI hype, which I will keep saying the word hype,
and Scott will explain why.
There is crazy hype right now.
We're seeing it both in the public and the private markets.
Go ahead, Scott.
Yeah, I would say that I agree, I think it was Caleb speaking before, that there's cause to not try to time the top of a bubble, obviously.
You ride the wave until, you know, the trend is your friend until they bend in the end, as they famously say.
But I would like to point out, and I just pinned it up in the nest.
I tweeted this just for this purpose.
Yeah.
People seem to forget now that we're riding this AI hype wave over the past few weeks with Navidia,
but they seem to forget that the chart is literally exactly the same as Bitcoin because the last wave up and down in the video
was simply because of Bitcoin mining hash rate and the excitement around Bitcoin.
So listen, I'm a huge bitcoiner, but you can't look at these charts and say, hey,
Navidia hasn't capitalized massively from two of the biggest hype cycles.
in history, right? I mean, people love to point to the crypto and Bitcoin bubble,
while Navidia stocked within a month of Bitcoin price topping at 69,000 and went nothing
but down along with Bitcoin, as Bitcoin bottomed around, you know, 15,000, so if you look at those
two charts and you think that this hype cycle or bubble might eventually pop, yes, you ride
the wave, but you accept it as this AI bubble
eventually, you know, calms down or cools off,
you're probably going to see Navidia also doing the same.
I mean, it's very, very hard not to look at this
and see the similarity.
I mean, just look at the price.
I mean, that's Navidia in the blue and gray candles
and Bitcoin in the orange candles.
All the way up to the all-time high
and then all the way down together to the bottom
within literally a month on both.
Hey Scott if you want to see something really cool instead of looking at in video look at the long and that's a great relationship
I'm not trying to debunk it or anything like that but just to add on to your point that's a very short-term correlation you could
From the studies that I've done the best long-term correlation I found for Bitcoin on a long-term basis is
Micron Technologies MU right so look at this chart that I just pinned up there
That goes back to 2016
right so this is a very long-standing relationship you can certainly have short-term deviations
but again kind of plays on to the same thing same theme that we're talking about right now
where semiconductors are
are really enabling kind of the edge of innovation and technology as a whole, right?
That's no secret, whether it's cloud, whether it's data centers, whether it's AI, whether
it's crypto, whatever the case is, it's a puzzle piece that you could slide into anything
and it fits, right?
And so it's pretty interesting to see that, especially with a correlation, such a long-term
correlation with something like Bitcoin.
So just to add on to the point that you just brought up there because it's fantastic,
everyone should know about the correlation with Nvidia, but
Everyone should 100% know about this correlation with Micron as well.
Yeah, I agree with that.
That's a really, really compelling chart as I'm looking with it.
And then clearly the difference is that Micron remains effectively, you know,
correlated very closely with Bitcoin right now.
And you can see Navidio diverging massively, really over the past couple of weeks,
Bitcoin started to sort of top in the 30s.
And then we saw the AI hype and it jumped.
So maybe the Micron one is a better...
correlation, but the point being that we've seen the Vidia jump massively on the height.
So maybe it's leaving the Bitcoin chart and now it starts to correlate with whatever's
happening in AI, which inevitably will cool off.
Perhaps NVIDIA is just like all the VCs that pivoted.
That's exactly right.
I mean, that's exactly the point is that the same people who are passionate,
listen, how many people who are passionate about Bitcoin and crypto now do you see only talking
about AI?
Right. I mean, there's obviously people who, for better or for worse, chase these hype cycles
and the video's chart is probably the best visual reflection you're going to get of that.
So, Scott, Scott, which is the logical question to ask next is with this recent divergence,
who converges on the other one, right? Does Bitcoin catch up to Nvidia and the rest of
semiconductors or do semiconductors revert back lower towards where Bitcoin is right now? What's your thoughts on that?
If I had to give a guess, I would say they meet somewhere in the middle, which would be the most logical expectation.
I know it's a laughable answer, but I think that that's the truth.
Because, I mean, if we're talking about relative percentages, I mean, Bitcoin would have to effectively be testing all-time highs.
to make a similar move as NVIDIA's made up in the last few weeks, right?
We're talking about doubling, tripling, and price off the bottoms.
I don't see that happening or any reason to see that happening.
And I think, Caleb, you really made the strongest point, which is a lot of this hype is going to be largely determined if it continues on what happens with the market as a whole, because that mean reversion trade is a bitch.
every single time.
Whatever pumps the hardest always falls, right?
I mean, it's not really, it's not rocket science to expect that the easiest short is going to,
if the market goes down, is going to be the thing that rose to the clouds.
Yeah.
Yeah, exactly, because it's been interesting to see something like equal weight S&P 500 just generally kind of consolidating sideways.
And in that environment, Nvidia and the rest of the semiconductor universe is just, you know, skyrocketing higher, right?
So the rest of the market hasn't even necessarily been that strong.
It certainly hasn't been weak, but trending sideways is exactly that.
It's just trending sideways.
Well, here, I'll give you guys kind of my insight.
So I'll give you a bare case.
probably a year, year and a half.
Hey, your audio is awful.
Sorry, your audio is really bad.
Can you drop down, come back up, my friend?
Was this me or his audio was pretty bad?
Yeah, it wasn't.
What tested it.
So it was like throughout it, I believe that
Nvidia right now has...
Oh, he fell down.
Come on back up, my friend.
Sorry.
Also, I won't, as I was saying earlier,
I will not, I refuse to call...
call people
Kucci King
so I'm going to have to
switch it to
see King
I think you just did
not going to do it
not going to do it
he just got you
Yeah, you just did it.
That was a gotcha.
I just did it.
That's true.
Can I add one thing to what's got?
I really enjoyed what Scott said.
So, and I love that quote.
The trend is your friend until the bend at the end.
So I got to pick that one.
I can't take credit for that, by the way.
That's a famous trading quote.
Nice.
Yeah, yeah, yeah.
I forget it from somewhere, but I do intend to borrow slash steal at a great, great quotation.
So the, the, uh,
You know, I'll give Nvidia this.
So since we're all kind of like, I don't call it the bull case for Nvidia, but the one good thing is I remember, like six, seven years ago, right?
So there are several companies over the last, several public companies that had sort of like venture like returns over the last call a decade.
Right.
So in there like a decade ago was like Tesla was a way to have public market investors to sort of have venture like upside.
And, you know, in that case, it's a good.
was proven to be true.
And Nvidia was the one talked about a lot, like, five, six years ago.
There was, like, a mini AI hype cycle, you know, in like, you know, the 2016-17s of the world,
where AI also went through, like, a peak of inflated expectations and then went down to a drop of disillusionment.
So what I think Nvidia did is what a lot of people in finance always tell you to do.
This is not financial advice, but it's just the general principle of, like, diversify, right?
So they did lay the foundations for Metaverse, VR, and...
AI and crypto like for years, right?
So they did, they did do that.
It's not like, you know, I do agree that Donish, it's really funny that VC suddenly
become banking experts to suddenly becoming, you know,
AI experts to suddenly becoming.
I think there's even memes in the VC community specifically about that.
Which, by the way, is like a defense tech.
But they're like all over all these different spaces, which is so fascinating.
So hype after height.
I believe so.
Yeah, well, I just specifically on Viti, I agree with that.
And I think there's a meme to be had there somewhere about VCs and probably about
Nvidia too.
But to Nvidia's credit, and I do have some questions about where it's going, but to
video's credit for years, like, you know, five to ten years, they've been laying the
foundations for these three specific industries.
In addition to gaming, which is sort of the core.
So this isn't like a yesterday kind of thing.
It's kind of like a, this has been a decade in development.
And.
one of the parts of their portfolio has gone up a lot.
And Metaverse, by the way, is still really levered to Invidia, right?
I mean, we'll see what Apple does next week with potentially its XR headset,
which will probably rely on the M1M2 chip based on rumors.
But, you know, I mean, you know, Metaverse might be back in the hype cycle.
So, Nvidia did a good job with diversification, and it's enjoying the reaping the benefits of that as well.
I have one or two bear cases.
I've been longer than I think I pulled out a little bit early.
I think Caleb did as well.
It's from what it sounded like.
I think obviously there's there's one good thing here.
Like if you have what you feel like is an overinflated stock,
you can always rotate some returns and buy puts, you know,
far out of the money and always have some hedge for the downside of,
if stocks are being a bit parabolic.
I think there are two things that I didn't even recognize up until what I threw into the nest here.
But, you know, it's...
The diversification or the lack thereof of operating revenue for NVIDIA is a bit concerning.
And then secondarily, I think the only other bear case is acting as though
Nvidia will always be the leader.
I think we talked about this a couple days ago.
But I think the software element and architecture that NVIDIA has is not overly impressive.
I've used it for commercial solutions before and it's,
It's workable, but I think one of the things that ChadGBT did that we haven't necessarily seen Nvidia do yet,
and perhaps somebody else comes to market and does it better or a competitor can stand up something.
But the tools that are basically worked into Nvidia's platform are subpar at best.
And I think one of the improvements that if you have $13 billion in cash,
is beefing up R&D to figure out how to commercialize and normalize the use of AI at a simplistic level.
I think this is kind of something that whoever leads that spearhead and it may or may not be NVIDIA could possibly really, really change things up quickly.
So those would be my only two bear cases on the notion.
So interestingly, so I brought up Shakar.
Many of you might know Shakar.
He's actually solutions that copy AI, one of the fast-growing AI companies.
And I wanted to get your thoughts, Shakar.
one broadly how much hype is AI you know we know it there is some hype obviously but you know are we
it's hard to time the top of the market now I will say this is not financial advice but but you
know Shakar wanted to get your thoughts on the hype around AI and then specifically around
NVIDIA and how you all think about NVIDIA being inside the being inside instead of outside the business
Yeah, that's a great question.
Frankly, you know, I'll address the hype on AI comment first.
I, um,
I don't think it's hype, to say the least.
I think there are a lot of projections that are speaking about potential improvements.
Obviously, I've shared in the past that I'm pretty bullish on AI.
I think in the long run, it'll lead to a lot of positive externalities.
That said, I think this idea that AI is not going to take your job
or at least not fundamentally change the nature of work in the short run is pretty silly.
We're already seeing that in the tech sector.
Now, what does that actually mean for an infrastructure company like NVIDIA?
I think it kind of depends on how a few things play out.
First, I don't think this is the top of the market.
They've got a 10-year advantage.
They've done an incredible job doing a lot of the fundamental advances that are required
in order to be able to do compute at the scale that is required to support these massive,
massive clusters.
I mean, forget the chip name.
They added a few more letters to it as usual, but if you look at the metrics on this
thing, it's unreal.
And what we know about infrastructure advances in technology is that they typically create fundamentally just completely new opportunities for companies.
Like, I mean, take cloud for an example, right?
For the longest time, it wasn't possible for most companies to be able to be competitive doing, you know, incredibly advanced things, mostly because...
There's only a handful of ways that you can build a data center.
But cloud compute fundamentally changed that, right?
By giving us, you know, software as a service,
but more importantly, platform as a service.
So everyone could spin up, you know, a giant cluster to, you know,
for example, host a massive distributed application.
I think Nvidia is a similar unlock at the infrastructure layer
for, you know, a lot of the leading industries.
between, you know, obviously AI, even gaming, right?
I mean, the idea that we could inject AI into games
so that you could actually have a real conversation
with an NPC.
That was nuts.
That was nuts.
When I saw that, for people that are not aware
of what Shakar is talking about,
it's as part of their demo,
uh nvita when uh when they were walking through it they gave an example of a video game
where all the npc's all the you know non-main characters uh you could walk up to any one of them
sit down at a bar have a conversation with them and a i was generating the whole conversation
they had a backstory but that backstory was being uh uh
you know, used to generate the entire conversation.
It was unreal, Shakar, and very natural.
Yeah, and I mean, that's unreal, right?
And that's what's possible at the demo announcement of this chip.
Now think about what that means for other industries that aren't just gaming, right?
It's cool to be able to chat with someone in Grand Theft Auto, but even more interesting would be able to create a completely VR world that allows people to have real or at least seemingly human connections.
There's obviously a lot of business impact to this for being able to train people to do some complex things or deal with real world scenarios.
All of these are net new opportunities that are going to create massive changes in the way that we live, changes in the way that we work, and frankly, lots of opportunity for people to consume more of Nvidia's core tech.
Now, that's the both side for me.
Can I play devil's advocate for a second, my friend?
So, you know, if you talk to Web 3 folks like a year ago, you sound a little bit like them.
I'm just saying, not trying to hate.
I'm not trying to be difficult.
But you know, you know, I respect you.
That's why you come up on these stages.
And so, you know, I do want to say you can't.
It does just like a little bit of the, hey, this has been going on forever.
Y'all just didn't know about it.
this is real, don't worry.
But if you have to ask the question, is it real?
Don't you think that automatically means that there is a little bit of truth
about whether there's a little bit of hype?
And again, let's talk very short term.
Cody, what are your thoughts about this?
I know you're a little bit of a bear on this.
Oh, definitely.
I mean, I don't even think we have clear regulatory guidance
on how this is going to play out.
I think...
You know, it's easy to have perceived, you know, creation and perceived job destruction.
But, you know, in my mind, this is all perceived scarcity.
It's all piled into the short term.
And, you know, I mean, this stock could fall 50%.
That wouldn't be surprised.
So I'd love to love to jump in on that.
Oh, yeah, man. I was expecting you to respond. Go ahead.
Hey, I play both sides. I play both sides. I'm pretty excited about and optimistic about what this could mean in the long run.
In the short run, as it relates to Nvidia stock, you know, one thing that's obviously very difficult to ignore is the advancement of a lot of these open source models on much, much, much smaller data sets and requiring much less compute.
Now, today that still probably is using an NVIDIA GPU somewhere.
But we know Microsoft has also put its hat in the ring and being able to produce GPUs.
I think at some point we could see an eclipse where maybe you just don't need as many of the insane, newly graphic cards that NVIDIA is producing.
And, you know, we could see, you know, kind of what happened in cloud where suddenly cheap bare metal and new architectures make it a lot easier for people to be able to do things without necessarily, you know, reinventing what Nvidia did.
But again, you know, when we think about, go ahead, Cody.
I'll say, I mean, Tesla is building their own chips. Apple's building their own chips. You know, AMD is obviously a competitor. I mean, like, I.
How is it that we believe that
Nvidia is just going to corner this entire market
and no competition is going to come into play?
I mean, that doesn't seem like a positive case at all.
There's already competition though.
Right.
So like for example, on the data center side of things, I think Nvidia had about 4.3 billion in revenue, whereas AMD had 1.3, right?
So AMD is competing in the market.
They could certainly compete more.
It's just, you know, leaders lead, right?
And it's pretty clear right now who the leader is.
Well,
Well, that was actually the point I was going to make because it seems like
Nvidia has kind of just walked into this position where they became CNBC and Wall
Street's darling for an AI company.
And I'm actually extremely bullish on Nvidia in the future.
It's one of my largest positions.
It's grown into that.
I haven't been buying it at these levels.
Right.
But in my opinion, Nvidia is going to do great in capitalizing NAI.
It's just that the stock is really pricing in everything that's going to come to it over the next couple of months and years.
I really do think the best place to really be positioning now, obviously no financial advice, is the companies that everyone's kind of sleeping on that are also taking advantage of AI.
Someone like Tesla, who's really been leading the way in one of the biggest and most powerful...
powerful artificial intelligence platforms out there and maybe even a company like
Palantir who's also doing a lot with artificial intelligence.
I just think the market has really said, okay, there's only one artificial intelligence company
out there.
That's Nvidia and that could not be farther from the truth.
I just think it's one of the easiest for Wall Street to invest in and see.
And two years ago, we were having that exact same conversation, but not on CNBC within the crypto community about how Navidia was the singular company that was going to capitalize from this massive Bitcoin trend when Bitcoin was at $69,000.
And once again, guys, I am a bitcoider, but this is the exact same conversation that has now
reached the mainstream media.
That's usually a very easy counter trade.
Not saying, like I said, doesn't mean you short it today.
I'm just saying for all the reasons you guys just listed, the video will revert to the
mean as the others rise to meet it, even if this AI trend is real because of that competition.
And this is just, I mean, it's literally like deja vu for anyone who is in Bitcoin.
It was buying Navidia because of.
of mining and because Bitcoin was going to $250,000 or a million dollars.
I don't believe that AI itself is hype,
but that doesn't mean that the trading around it is not a hype wave,
just like I don't believe that Bitcoin is hype,
but I believe at that moment it was a wave of hype for investors.
Talk about Bitcoin.
Hold on, Justin.
Hold on. This is my transition. God damn it, Justin. Just kidding.
You know, this is my transition. Talking about Bitcoin,
wanted to share that we are starting a new show. Scott is doing a new show with Mario and Ran.
It's super exciting. It is a pretty big deal. KBFD.
Really excited about it.
Scott,
tell us a little bit about your show.
It's starting today, isn't it?
Isn't today the first day?
Finally starting today after months of planning,
effectively and coming together,
obviously with our three teams
to make this happen.
Crypto Town Hall,
it's going to be at 10.15 a.m. Eastern Standard Time.
Every weekday,
we're just trying to be like you, Donis.
We want to just show up and be there every day
and be the source for citizen journalism on crypto.
But we think that there's a...
woefully lacking citizen journalism in the crypto space right now because it has received so much sort of negative mainstream media or even at the time like I said positive mainstream media hype.
So the goal is really to get down to the core of what's happening in that space on a day-to-day basis, break down the news with
experts. I mean, the guests we have lined up are going to over the coming weeks and months are going to absolutely blow your minds.
But really, we just think that it's very important that people are speaking accurately in real time about the crypto space.
So we're committed to showing up every single day now because I need like another two or three hours of talking about crypto in my life like a hole in my head.
But here we are.
But it's interesting, you know, one thing that's really interesting, and this is the thing that makes citizen journalism so different, is that when everybody is zigging, you've got a zag.
And ultimately, there's something really exciting happening, in my opinion, in the crypto space right now, which is the hype is gone.
The hype is gone.
Of course, there's bullshit like Pepe and crap like that.
Meme coins and stuff like that's still around.
Sorry, I had to take my dig.
But what we're seeing right now, even in some of the spaces where I disagree,
utilization of Bitcoin as an actual software platform,
it is interesting that those BRC 20 tokens, all that stuff.
Crazy stuff is happening right now and no one's paying attention.
That actually is a good sign.
that i agree and i agree and i think that it's very clear if you look at the these once again
hype cycles in crypto and what's happening
The best products, platforms, protocols of the future are always built in the bare market
when teams can bear down and not be concerned with price and hype and can actually focus
on what they're building.
So, you know, inevitably in the next bull market, whatever narratives we're talking about
today, the things that are being built, those are going to be the winners.
It's always laughable.
People are always like, which coin do you think is going to benefit the most from the next
bull market?
My answer is usually something you haven't even heard of yet.
Right?
Exactly.
The coins from each cycle are dead by the next one.
You have so many people holding bags or looking to sell on any move up, and it will be whatever this next narrative develops.
And so we really want on a daily basis to be digging in to the positive side.
What's being built?
We're all tired.
We're going to talk about it because we have to.
But we're all tired about talking about United States crushing it and regulation and legislation.
None of that is focused on what...
these absolutely brilliant entrepreneurs
are building in this space right now.
And if you are an entrepreneur building in this space...
Oh, good segue.
Yeah.
IBC incubates and accelerates AI and Web 3 companies.
We've been partnering with the VCs we were just hating on a few seconds ago and funds directly.
So if you are a VC and you do like the hype cycle, maybe it's time for you to work with actual people.
Sorry, I have to do it.
And Mario is going to kill me for saying this.
And we will work with your portfolio companies to actually help them build
a community. They're also doing some Shark Tank style pitches. And I know Scott's helping
on the crypto space. And so if you're interested, please reach out to Mario and his team to learn more
about how IBC helps.
All right. How was that?
Donis, you are a season.
Epic professional.
Smooth.
There is nobody.
Nobody does it better.
Hopefully one day.
You can get through the transition
without everyone openly mocking him
while he's doing it.
Thank you so much.
Justin, I appreciate you standing up for me there.
People really shouldn't mock my ability
to do these sponsorships.
And I feel like I'm potentially hurting
our ability to get more sponsorships.
And so please...
People, please tell some VCs.
So it seems like we're actually effective at this.
All right.
We're going to move on.
Can I briefly touch on the bull and bear for, I think everyone.
Yeah, because I want to put an exclamation point on what Scott said, which was, I think you can have a bull and bear case in AI at the same time.
I think the bull case, I think Sotia Malik, the CEO of OpenCV.
He told me AI is electricity.
He's like, it's going to transform literally every industry, every job.
But I think the bare case is, you know, obviously invidia soaring forever.
That's a bear case.
I think the bear case is also disrupting all jobs and taking everyone's jobs.
I think that's a lot of hype too.
Yeah, on the hype, I want to say one thing that I think has become pretty apparent, which is the one thing that everybody's like, well, how is this different?
I think previously, obviously, the notion of people, everyday people being able to use AI like chat GPT, etc.
is one thing.
But I think one of the interesting notions about, you know, the recent, you know, development of startups and implementation like, you know, copy AI and others is,
It's no longer just marketing hype.
I feel like four to five years ago,
plus everybody...
had an AI company or a, you know, crypto company or whatever.
And there was a lot of ambiguity around what that actually meant.
And I'm excited about what I've seen from recent startups who are actually deploying
AI in a meaningful way instead of just using like Microsoft BI tools to help with analytics.
I mean, I can't tell you how many, you know, decks I've been using Microsoft open AI tools now, Trevor.
Exactly.
Same exact shit.
Like,
they're just building wrappers around it.
Sorry,
I have to push back a little,
which is,
you know,
just because it looks cooler,
because technology has developed further,
doesn't mean that the actions have changed.
Sorry, Trevor,
keep going.
I was just going to say,
I've seen more actionable tools
that have been developed as of late
within the last,
you know,
six to 12 months than I've seen previously
when I've seen,
you know,
a large amount of hysteria and hype
over things that,
It's all branding and marketing.
And I think those who sit in a lot of these startups look to the left and right of them
and say, oh my goodness, like there's a lot of hype around me.
And it's actually creating a lot more difficult scenarios for startups who are actually
implementing unique novel approaches.
And I think now what we're seeing is the tool sets and unique capabilities that
startups can use AI at a level that's no longer like, man, I've got to start from
scratch on every single model that I develop.
and I can actually implement and deploy something that I can commercialize.
And I think this historically has been a real big problem that now startups are actually having
functional tools.
I mean, you could go and look at a lot of pitch decks over the last five to 10 years and
everybody's LTV and customer retention is absurd, right?
We're going to have this customer for 10 years.
Realistically, no, you're not.
because you're going to have a competitor that's going to run into the space and compete with you and basically commoditize, you know, a comparable product that you have or something near that, or they have better marketing and branding.
And now all of a sudden, everybody's, you know, always looking at the grass is greener on the other side.
So I think just in terms of what the momentum that I've seen and the excitement for not necessarily –
you know, the bull case for NVIDIA, but I think in general AI is the adoption of tools that are no longer just like,
hey, this is a $9.99 subscription that I use for four days, can't figure out how to use it or it's worthless.
versus, hey, I actually can pay a $20 subscription, and I'm using it on a regular basis more
than I would have four or five years ago with a quote-unquote data analytics tool.
So I think those are just some observations that get me excited.
I think some of those individuals who are in the startup space in AI specifically would agree
that we're seeing a lot more, I think, realistic application that is
is kind of pushing the excitement level
that historically, I think a lot of people
in the boom and bus cycles of AI hype
have been limited around.
Look, I think we get this.
Go ahead.
I'd love to jump in on that.
I 100% agree,
and this is exactly what I mean
about infrastructure raising the water level.
Frankly, I don't think it matters
if they're still just using some Azure API, right?
What this generation of AI has done is...
in fact, create this concept of wide AI that actually is incredibly useful.
Now, there's degrees to which, you know, people or companies are actually, you know, effectively
I'm pushing the same platform.
You're telling me that a startup that just builds a wrapper on freaking GPT4 is automatically
a company that has a moat.
Come on, Shakar.
We know that is this.
No, fuck no.
I think, uh,
Yeah, exactly. Come on, man.
Like, there's a bunch of people doing this right now and getting billion dollar
valuation.
This is insane right now.
I think that's what we're saying is the problem.
I think that's the issue.
That is a problem.
Go ahead, Shoucar, sorry.
No, it's a problem and it creates confusion for a typical consumer.
But it's hard to argue with the fact that that wrapper is still creating something that
consumers want and is delivering value to them.
So to Trevor's point, it's still useful, right?
I still feel like it's better use of my money if I spend $9.99 on a subscription for something
I use more than four days.
I think it obviously has downsides and, you know, the confusion aspect of it is, you know,
if I use a weak tool and, you know, find that it's not doing everything that I want, I could
write it off entirely.
And that's actually the frustrating part in working in this space and building, I think,
you know, something a little bit more real than the startup that was born yesterday.
Exactly.
I was going to say that what is doing is adding a lot of noise.
And ultimately, what's been fascinating for me is seeing that.
That what we saw, for example, in the early days of the internet, again, I'm a little bit older, but I was going to say that the early days of the internet, we saw a bunch of companies that built dot coms that each one of them could have just been in the Amazon store.
And they were essentially e-commerce companies that didn't have good platforms to sell.
And what we're seeing right now is a bunch of VCs investing in startups that really should just be one of many.
And they're good lifestyle businesses, but not venture scale.
And we'll never get the growth that they need to become what actually needs to happen.
What's happening instead is that people are investing in the wrong business.
companies.
It happened in Web 3.
It just keeps happening over and over again.
And I don't want to keep shitting on VCs because they're doing their jobs, right?
At some level.
But they're,
but they,
it's really hard to tell signal from noise.
You know,
You know, I know that a bunch of people
unmuted. Scott, go ahead.
And then, I think Cody then mesh.
Go ahead.
Yeah, I was just going to say really quickly,
I mean, to Trevor's point,
I think that they're correct.
And it's funny, we talk about bubbles
as if it's a negative thing,
but each of these bubbles,
once again, advances the ball massively,
and then you get the next wave
of the biggest companies.
People always compare...
crypto to obviously the dot-com bubble as if it's a bad thing because yes,
Pets.com and net taxis failed, but you also got Google and Facebook and Amazon as the
biggest companies in the world.
I think that that's largely what you're looking here to once again remind you that
the bubble can be in the actual stock prices and not in the technology underlying.
I mean, you guys remember in the Bitcoin hype cycle previous to this one,
there was a company that was an iced tea company in New York, Long Island,
and they literally named themselves like Long Island blockchain iced tea,
and their stock went up hundreds of percent.
You can look at what happens in the crypto market sort of as a pre-hype...
to what's going to happen in other markets.
Right now, you see AI coins or things that deem themselves AI coins going absolutely mad.
When Facebook rebranded to meta, you saw Metaverse coins that had literally nothing to do with it going absolutely mad.
And then you see it in the real market.
I'm not going to call the crypto market the real market, unfortunately.
Then you see it in the stock market.
And that's what I think we're going to see here.
A bunch of people rebranding, saying that they're AI companies that were never really AI companies.
he's capitalizing on the hype, then the bubble will pop and we'll see the big winners
come out from there.
Agreed.
Agreed.
You know, and I just, I want to reiterate this because I think it's really important for
everybody and especially the listeners is, you know, when everybody talks about AI taking
30 or 40% of the existing jobs, it's just actually not true.
And I mean, Chakar, you can probably attest to this.
I'm an angel investor and I venture invest in a lot of startups.
A handful of them are AI companies.
The ones that are successful, the ones that are doing a really good job are not taking jobs.
They are enhancing jobs.
And it's really important to reiterate this.
This process is not going to happen as rapidly as everybody thinks it will.
This is perceived scarcity at the moment.
And the successful companies in this AI space are going to 10x the workload of the existing employees.
This will be a productivity increase.
Yeah, I can attest to that.
Copy AI, we're 36 employees and, you know, have 8 million users.
And we use AI as literally every fabric of our DNA.
I'm torn here, right?
Because long run, I genuinely think that it's going to create net new jobs, net new opportunities.
Short run, it's just so attractive.
And it's honestly as a business owner hard to justify a lot of the roles that exist today.
So, you know, I think the metric to watch is how fast adoption is really going to drive that change
and how quickly people are going to uplevel their skill sets.
But, I mean, take like a data entry job or, you know, even like a basic marketing job or even a team that just tends to, you know, have a lot of resources required like, you know, even in enterprise tech, like a software development representative or sorry, a sales representative development representative.
A lot of those roles will, you know, really just have to evolve very, very quickly.
And the people that have those skills will have to change very, very quickly in order to not be immediately impacted.
I don't know.
Chikar, how much bigger would your team have to be
to produce the same output if you didn't use AI in your work?
I mean, put simply,
the last company I worked out
was the head of solutions architecture at MoveWorks.
When I joined, we were 120 people
and scaled to 700,
and I still think that we're, you know,
likely significantly more productive at copy AI.
So, I mean...
Kind of nuts.
Yeah.
Agreed.
Cody, did you have a comment before we moved to mention?
I was just going to say, I mean, you know, if you haven't built a company before, you know,
obviously a lot of people in this panel have done those things, but it takes time.
It takes a lot of time.
It takes a lot of energy.
It takes a lot of resources.
You know, AI is not going to be any different, right?
This process is going to happen slowly, much slower than most people understand.
The infrastructure has to build out, you know, a product, you know, a product,
has to build out, the discovery of what works and what doesn't has to build out.
You know, a lot of these AI platforms will become more vertical silos that I think people realize.
And, you know, this will happen slowly, but it will happen for sure.
But it will really increase efficiency of productivity in most enterprise solutions, like your side.
Yeah, it's interesting because Mish actually sent me something in the back channel about how AI could affect investing.
But Mish, did you want to talk about that or jump in on something else?
Well, yeah, I mean, I definitely have been listening to everything.
And I think that having a balanced...
mindset towards AI in that it's going to be continuing to advance,
and yet at the same time, there could be some dangers,
but not the end of the world.
You know, to me, seems sane.
I always try to take the road of sanity.
But in terms of what the article, yeah,
I was called yesterday by BNM Bloomberg,
which is Bloomberg in Canada,
to ask about, it was so timely for us,
to ask about how investors...
and portfolio managers are using AI.
And it turns out that we are using it in ways that have made our productivity so much greater and faster in that if you don't really know how to code,
you can essentially use AI now to take all these different trading variables that you might consider,
especially us who are very quant model oriented.
and input them into AI, and it will basically do the work for you.
So if you like to look at trend strength based on moving averages or volume or RSIs, et cetera, et cetera,
you put that in and bam, I mean, we're getting coding and getting systems either better or improving or even creating.
new systems that are being AI generated.
So it's faster, more efficient.
And I believe that it could really only now just beginning to change the investment space.
Most people are already using it that are in the investment space.
And that's just the very, very beginning.
So I still think that...
we have a big, big, big future coming here when it comes to AI.
As far as the trading aspect, because you know I'm also a trader, we've had, and I think I've mentioned this to you,
Danish in the past couple of weeks, we've been long technology, we've been long Envidio, we've been long meta,
we've been long, a lot of stuff within the semiconductor space, but we take profits along the way.
And that's always, so there's the future of the trend.
But then there's also the actual strategy for trading.
And I think it's really important for people to understand that
that you want to take profits into this.
You don't necessarily want to buy when you have these PE ratios exploding
or the relationship between spy and some of this tech at 2000, the year 2000,
that's where the ratios are right now, similar to 2000,
in where the spy is versus the tech.
When you get that kind of extension, you know, you definitely want to understand that
And I think I read that Kramer said this today.
Buying now might be a little bit more risky than waiting for what looks to me like some kind of correction coming, particularly if you look at the...
That usually means buy now.
But go ahead, Mish.
When Kramer says don't buy, it usually means buy.
He said bye.
I mean, I don't...
Oh, crap.
Oh, no.
Oh, no.
Envidia's going down.
It'll be worth a billion tomorrow.
Sorry, go ahead.
No, no, I mean, I think it's dangerous to take one person as the absolute opposite of what you should do, but it has been incredibly accurate for him.
No, I mean, just looking at the numbers today, NASDAQ is coming in lower than yesterday's low after making a new high.
That's sort of like a typical reversal pattern.
That could be good for...
a 5 to 10% correction. So if you use those numbers, if we go down about 5 to 10% here in some of these
big names like NVIDIA, that may be a better buy opportunity is all my point. And we can talk more
about the general market if you want, but I'm really interested right now in the metals again.
So whenever you want to talk about that, let me know.
But I did want to recommend for everybody that's listening.
I know we've been going back and forth on AI, and we will get to Elon and his visit to Beijing.
But, you know, I really did want to encourage people.
Remember, if you want your voice heard, there is a way.
This is not like the radio.
We actually are reading comments.
So go to the bottom right, put in comments.
Michael, who's actually up on stage right now,
is a good example of somebody that got the ability to come up here,
and now he's a regular.
So again, anything is possible.
So, you know, I wanted to kind of move towards charts for a second.
for a second. I know that Caleb initially was mentioning both the bull and bear case and Patrick, you don't have to raise your hand. I'm coming to you already, man. I'm just kidding. But I was going to say that, you know, ultimately what we do want to know is, and I know that that's why people join us, even though everybody knows that this is not financial advice, but the markets do speak to us. Patrick, tell us what the markets are telling us by Nvidia today.
All right. Well, I just put in a chart. Look, with charts, it's very easy because it creates a way to objectively see
what the actual formal is.
So people say, oh, Kramer's talking about it.
It's talking.
Well, hold on.
He's saying something on the daily timeframe,
but how is that really in the scope of how far formal can get?
You know how people like the froth can keep going higher
and more exuberant than what people think?
So on the chart, I put it in the Twitter space.
in the tweet, I put a chart of Nvidia and I put the distance from three-year moving average.
So essentially there's bull runs and they usually peter out at when it's stretched away from three-year-moving average.
Historically, you see the previous times a stock within its own cycles is able to stretch.
Invidia likes to go about 200% above its three-year moving average before starting to slow down on the log scale.
And I drew it. So right now, the three-year moving average is about 190.
It's about halfway, it's about 100% above it right now.
So if it keeps 110% above its three-year-moving average, so you can extrapolate that.
If it would rock it up today to be able to be above, 200% above its three-year-moving average,
I have a target around 570, 580, before...
When the Nvidia hits about 575 and north of that,
I think the FOMO is going to be laser
Nvidia chips eyes in Twitter accounts, right?
That's what we're going to start seeing.
That's when really people should start taking profits
off of this train right now.
It doesn't mean it can't go higher,
but that's probably going to be a great place
where you have some crazy candles, monthly candle
that's like just skyrocketing to the moon.
That's probably going to be the place.
People should start taking profits.
And because that stuff, that three-year-moving average has to catch up eventually.
And if we really look at the chart, look, the last greatest entry in Nvidia, it was back in 2014, guys.
That's what it was, maybe in 2016, after a 10-year or 8-year base.
Then you had that crazy run from a sub-10 all the way to today.
Look, it's kind of sad.
The fact we have a space today we're talking about on the video is because it went up a lot.
It doesn't mean it's the end, but usually when we had these type of talks is we're 20% away from the top, 50% away from the top.
But those 10-bag removes, guys, they're in the past.
So that's what the chart's telling me that.
Same thing with Bitcoin, something with any other type of instrument.
Because we're talking about it on this Twitter space, we're getting close to the top.
And I just put in Bitcoin versus Nvidia.
Yeah.
a racial chart of it.
So a lot of people are trying to compare Bitcoin and Vida and they're overlaim the charts,
one on top of the other.
But to extract better information from two charts that seemingly have the same chart pattern,
you have to divide one by the other.
And Bitcoin has carved up a massive top versus Nvidia since 2013,
all the way to 221, where it did a blow off top versus Nvidia.
A huge trendline break happened in November of 2022 in favor of NVIDIA,
and the waterfall is continuing downwards.
So the better performing asset is Nvidia since...
about six, seven, eight months ago, and that could actually accelerate in invidious directions until the individual comes out.
What I'm hearing is, sorry, Michael, I'm hearing that there should be an inverse Mario, like there is an inverse graemer.
I don't like this, but I appreciate that. Go ahead, Michael.
Yeah, I remember people were saying the same exact thing in 2018 when
Nvidia had its first crash after that big crypto run.
And that's really when I started buying my position.
And the interesting thing is, is that's up over more than 10x from that point.
So I think this idea that, oh, it's went up a lot can be a little deceiving when you zoom out
on such a large time frame.
I mean, I think it really just comes down to is how important is
If they really do continue to dominate, then I don't think there's any way to say,
okay, this is it this time.
Michael, but I have to say that you use the word zoom out, that should be a negative, you know, that's what all the crypto folks were saying and Web3 folks were saying last year.
So I feel like sometimes we shouldn't zoom out too much.
But go ahead, Patrick.
You won't do you have a response.
Yeah, because it's not, look, guys, it's not the narrative.
Like, InVideo could have, AI could have been happening in 2001 and the price of Invideo could have kept crashing.
It's the narrative comes because the price is going up.
People like attaching a feel good narrative
to why the price going up.
2018, if you look at my chart that's put in the Twitter space,
and Vita crashed to the three-year moving average
and when consolidating and went sideways for six, seven, eight, nine months,
and then it bounced back up.
That was a time to buy and to fear when nobody wanted to buy it.
The charts show you that.
So right now we're 100%, 110% above
the three are moving average.
So we're getting up there.
Momentum's going up,
but that's the danger of people buying on narratives
is they won't be able to sell.
Once we get at 575, 600,
we're stretched away from the moving average
and that thing goes sideways for six months,
eight months, 10 months, who knows?
People will not be able to let go
because, and they'll be trapped opportunity costs
because they have this narrative that AI is the future
and goes forever.
Capital flows carve out these price charts, not the narrative.
Yeah, but you could argue that investors got trapped in the fourth quarter of last year, right?
When the stock was down, I think, 67% from the November 2021 highs.
And so, you know, those investors today who are trapped are now at all-time huts.
Right.
So, again, I think Mickle's point is accurate.
If you're, if, and, you know, even the contradicted what Danis just said is like,
You know, even Bitcoin now has gotten back. It was briefly above 30,000. Right now we're right around 27, 28K. And so I think, you know, the thesis is still there. The fundamentals of where the company is going and what they're doing are still there. Right. And so I don't know. I really agreed with what.
Nicol was saying. I think traders and investors are looking at this very differently. As an investor,
you look at that big dip or you look at that trapped opportunity cost, as Patrick was calling it.
And I really do see it as a company who's continuing to increase fundamentals as just a great way to say,
okay, the market's not realizing what's happening here.
The stock's not moving.
This is the time to slowly average in.
I mean, times like this are obviously, I don't think anyone would look at the chart of
NVIDIA and say, oh, this is the time.
So I just think it's maybe a little bit of difference in how a trader might look at this
versus an investor who's been really following NVIDIA for the last five years.
Technicals versus fundamentals.
And we all know technicals the lead, right, guys?
You agree, right?
Am I the only technical guy on the call?
Look, you got to be careful, guys, because look, SPX, NASDAQ, they all broke down
versus the produce price index.
Last time that happened was in 2001.
The previous time was in 1970.
And U.S. equities nominally, while they might grind sideways and up, they're the
underperforming a whole bunch of other assets.
Look, in 2007, Invidio, went sideways from
from 2007 all the way to 2014.
I don't mind going long in video,
but at the low risk, high reward setup.
And that has just, we're about five months away from four months
into the current move for Nvidia.
So we're halfway there.
Not saying it's stopping now,
but I'm saying the lower risk entries
is when you're bouncing off, the three-year moving average,
and you're breaking out to the upside.
Because below that,
You don't know how long is going to go sideways, right?
Yeah, I think both of you guys are right.
You're just speaking the two different time frames, that's all.
And Patrick, you are the only technical person on this.
I knew.
I knew you were there.
I knew, thanks.
We try to keep it balanced and it's helpful.
Now, I wanted to commend Bradford, who's joined us up on stage.
Remember how I said you have to comment in the comments, and if it's good, you get up here.
Well, Bradford's up here.
Tell us a little bit about what you messaged us and, yeah, I would love to get your question and comment.
Yeah, so I'm more of an investor, but I think it makes sense to add and trim at opportunistic times.
And we've been long in Vida for a couple of years, but it seems to me like we're close to a buyer capitulation, a FOMO panic.
Um, C3.A.I had five times volume up 33% pre-earnings, no news.
Lemonade up, uh, 10% no news. Palantir 2, 3x volume, up 8% no news. Um, and Nvidia, you know, at one point was up
8% on a third day follow through. So I feel like,
This is the most it can sustain short term, but I'm curious what you all think.
Yeah, what do you people think?
Patrick, Caleb, Mish.
Silence.
Nobody wants to talk.
What's that about?
I'm surprised.
Go ahead.
Well, yeah, look, it's somebody quote tweeted my tweet on video.
It's, look, there's...
People think in binary, I have to be bullish or I have to be bearish.
And again, guys, it depends on your time frame, right?
100% agree.
Most of the debates, when I realize I'm arguing with somebody,
it's because we're not synced on the same timeframe.
But let's on the monthly chart.
Right now, like, it's on a bull run to a higher web target.
I would definitely never short Nvidia here.
There's absolutely no reason to short Nvidia.
Maybe north of 560, 570, but even then I would probably wait for,
for a multi-month topping pattern, really momentum to slow down.
And then once it breaks down, then I have a higher probability of having a nice short.
But right now, it's on the people who bought lower, they're the winners here because they won't,
they'll feel that their targets are reached once they hit 600, they'll take profits.
But people that are buying too close to when you get close to the top, they are trapped
because psychologically, they want to make the same gains that they heard the guys before
they make the gains.
And as a pattern tops out, they'll refuse to accept that it's a topping pattern.
And once it starts breaking down, they'll start dollar cost averaging in it.
Then they're trapped.
Right.
So that's why people who buy in early and good, they're fine because they'll be able to take profits.
But the people who buy close to top, like all the people who bought crypto's or Bitcoin in 2021, above 50K, 30K, they are 100% trapped because they psychologically, they can't sell.
They can't tell their wife that they're down 34%. So they're going to hold on for a few years.
Yeah. I mean, Bradford, any feedback?
Yeah, I think, yeah, I'm just, I'm longer term, but I feel like it's a good time to trim.
I totally agree that, um,
a longer move to 600's possible.
I think Wall Street is still underestimating them on a fundamental level
and underestimating the amount of growth in both revenue and earnings.
I'm bullish, but we're looking to trim today.
Maybe it's slightly premature, but we certainly missed the top last morning.
Yeah, I mean, it's interesting.
I think what I keep hearing, and Michael, I'll jump, I know I cut you off earlier, you know, what I keep hearing is that,
There might be a little bit of a run-up short-term, but there will be a peak and it will probably come down.
But as it comes down, that does not mean that that is the bubble, quote-unquote, popping.
It literally means that it's just the beginning.
And we're probably going to see it come down a little bit and then go very bullish long-term.
Is there anybody here who's not bullish long-term on AI?
I feel like we're starting to coalesce towards a solution.
If there's anybody here who is not bullish,
who thinks that this is super duper hype,
this is the next metaverse,
this is the next web three,
is there anybody here who,
who is willing, please do not make me play that role.
There's got to be somebody here.
If there's somebody here, going to the comments on the bottom right,
I want you to tell us why.
And if I like what you're saying, I will bring you back up.
I'll bring you on up because I think there's got to be a bearish case.
I think it's easier to say certain applications might be overhyped for AI rather than the technology itself.
The technology is just so incredibly broad and seeping into so many areas.
I think it's just.
Honestly, it's blown my mind how broad it really is.
So I've seen things before with startups and small little companies doing these things that I look at.
And I'm like, ah, I can't really see that.
And I think that one specific use case is overhyped.
But in terms of the technology as a whole, I've been shocked with how many different places it's put its tentacles in.
And I will say that in terms of the narrow areas, oh, Jeff's about to kill me.
But I actually think that the physical aspect is,
of how AI applies to the physical world is way overhyped.
And what I mean by that is very specifically, as Jeff already knows where I'm going,
you know, I think that it's always going to take longer to get to full, full service driving, right?
We're going to see autonomous vehicles have already taken forever, right?
Obviously, when they get there, it'll be a big deal.
But mapping the physical world onto the digital world has been incredibly challenging.
incredibly challenging.
We have seen that over and over again.
Of course, the more challenging it is, the more of a moat there is.
I do not believe that robotics is going to take off.
As many people may know, that is actually my area of research.
So, you know, I know a ton about robotics.
And so, you know, I think robotics is going to take longer than people think.
I think the ability for AI to eat the digital world will be a lot faster.
Yeah.
than the AI's ability to eat the physical world.
We have seen this over and over and over again.
You know, we saw vertical SaaS grew way faster than tech-enabled services did.
And tech-enabled services still, which is, by the way, the technology component of the service world,
has been, you know, entering the physical world has been incredibly slow.
Jeff, why am I wrong?
Am I wrong?
I want to push back on that, but Jeff, go for it.
Yeah, I'll let Jeff go first and then you do.
I mean, just a couple of things.
One is like, you know, when you get in a Tesla with the full self-driving software, it's very easy and it's very obvious.
Yeah.
if a mistake happens, like, oh, I had to do an intervention.
I had to hit the break or I had to hit the accelerator or I had to hit the wheel.
With ChatGPT in the digital world, I mean, I hear your point.
And I think it makes sense.
But like, who's measuring the number of interventions is?
you know, like wrong data, like wrong responses.
Like there's a lot of like bad output that comes out of some of these things today.
So I think it's a, I agree with Michael too.
Like I think there's this underestimated like where could this go?
I think there's things we're just not even thinking about where AI can go.
At the same time, I think this whole thing of like it's taking everybody's job.
by the end of this year, that's not happening.
It's more of a timing thing, but I think the world,
there was a moment, right?
There's this moment with, you know, generational AI and this conference,
you know, this with chat GPT.
I think there was a moment.
There's a moment now happening with Bard.
You know, Bard is connected to the, you know,
connected to the internet.
You're getting real-time data and the responses are getting better every day.
Right.
There's a moment with all these plugins.
There's this thing happening now in the digital world that it's just,
it's hitting so many people right now that it's,
it's amazing.
That same thing is going to happen with full self-driving,
but it's going to be in a couple of different steps.
And this is what.
when you when you watch mainstream media they're like you know Tesla promise robo taxis you know
Elon prompt by you know that's the that could be an end state but between now and then like
you're going to be able to get in your car right right now there's people getting their Teslas
and driving for an hour without an intervention um so you're going to have this you're going
to have this intermediate state where you're
your Tesla is going to drive fat is going to have better crash prevention performance than a human driver.
Right now, it's already six times better.
So it's going to be this gradient.
It's not going to, it may be a series of moments that occur, but it's, I think it's a bit misunderstood.
And then taking that.
that aspect of computer vision and applying it to robotics.
I mean, I see it in manufacturing already.
It's having a massive
impact and I do think how would so just just I'm sorry we both are we're both we're both
engineers I love let me walk let's walk through this together okay how does driving data from a
freaking vehicle help a humanoid robot can we just be honest I mean I literally work in the space
there is no way you can convince me that data coming from a vehicle that is moving at completely
different speeds and is using completely different depths of view well in any way
help our ability to do actual task using a humanoid robot.
Well, I'm not advocating that.
I'm advocating more of the, yeah, I'm not saying the training data from the car
is going to train the robot.
It's more of the training infrastructure that Tesla's built.
And with their dojo computer and their own silicon that they've custom designed,
that infrastructure is going to be in place from full self-driving.
Yeah, yeah.
Okay, 100%.
I want to push back on the humanoid thing.
Well, related, but how we get there.
I'd love to paint that picture.
So love what Jeff said about Tesla.
I think it's, um, FSD is going to be really exciting and what it's going to do for,
for driving.
But, you know, going to meta, you know.
Meta, my old former employer, good old former employer, meta, I hate to say it, I think has a gigantic lead on computer vision that's going to enable digital to physical interfaces, right?
And that's because of the fact that it has the leading virtual reality platform today.
that is inside out tracking, right?
I mean, years ago when I was at Oculus
and then subsequently we called it Facebook at the time,
we acquired some of the best computer vision people in the world,
you know, acquired startups in Israel, et cetera,
and, you know, laid the foundations for what we have today.
And even though the MetaQuest 2 is still, you know,
nascent product, you know, few million units,
the kinds of data that is being generated there is absolutely incredible.
And even, you know, for anybody here who's either used a MetaQuest 2 or owns one,
you can see that thing, like, for example, the hand tracking data, right, the ability to track
individual digits so that you can type on a virtual keyboard.
I mean,
i mean that's that's a fairly difficult problem you know from computer vision side uh you know it's
it works pretty pretty darn well right if anyone's used the meta quest pro and able to use
virtual offices with the you know meta horizon workrooms um so i think this data the the digital
data is being captured uh for meta and hopefully we'll see with what apple does with xr next
week donnish to your point i think will lay the foundations and have the kind of data at mass scale
I don't know about humanoid robots directly,
but to be able to do lots of really interesting things
from the physical to the digital side.
Physical to digital is a lot easier than digital to physical.
That is very well established.
We saw a good example of this is how long has Boston Dynamics
been working on those freaky-looking robots?
Yet today, you see none of them actually being used by enterprise
or consumers to actually do any major tasks.
whereas
Can you buy them?
Can you even buy one?
Outside of the dog one.
I'm nother governments are actually.
They're very expensive.
Yeah, they're very expensive.
And they don't work that well.
You should watch their videos of Boston Dynamics robots
walking on,
on different terrain and it is hilarious to watch them.
There's like a...
The blooper reel?
Are you talking about the blooper reels they put up sometimes?
Well, it's even more than the blooper reels.
There's competitions every year where, you know, these robots are without pre-training.
That is the key here.
Without pre-training are given new environments and they continue to fail.
The humanoid robot that Tesla...
put out there is at least a couple of years behind everybody else.
And that's okay because Tesla does things differently.
They believe that they are the accelerant to the industry.
And I think that's what makes them great.
But at the same time, anybody that knows robotics knows that what Tesla put out there was not state of the art.
It was okay.
It was cool.
It looked like a government, like a, like a really high quality PhD project, which I think is great.
Again, you know, I think what do you, what are you referring to?
Like, what do you say what they put out there?
Well, when?
I mean, they've been, they've been showing it now for a year.
The first one that they unveiled was after six months, I think, of hardware development.
Yeah.
Exactly.
That's my point, which is, you know, again, the key with Tesla is never to look at them
at a snapshot in time.
The key to Tesla is to look at the speed and acceleration with which they build.
I think that was my main point, which is I'm excited about the work that's happening here.
But I am not bullish that this humanoid robot will add any value in terms of revenue or
or in terms of earnings anytime soon.
And anybody that's like, oh my God, the humanoid robot is going to change the world.
It's not.
We're like 10 years away, at least.
And maybe I'm underestimating.
And as far as Facebook and the work that Facebook is doing, I think that that will be
incredibly valuable in terms of the ability to map the physical world into a digital domain.
Putting the physical world and digital domain is a,
very big part of this. There's no doubt about that, Eugene, but it is the easy part. That is the easy
half. Going back from the digital domain back to physical infrastructure and then the feedback
between, sorry, I'm going too deep into this probably, but, and then I always ask you all not to.
But I think it's helpful, but I want to class a clarifying question actually. So agree, digital
to physical, much, much harder problem. So, so.
It's completely agree with you there.
But I guess trying to understand what the idea of what you're saying, the goal is, is the goal like literally like humanoid robots?
Or, because in my vision, it's going to start like you're going to have your virtual reality robot, you know, in a virtual space.
And that's going to give you some training data that could help you out.
And by the way.
humans moving around, right?
Like legs, arms, right?
The six degrees of freedom you have now with the hand controllers,
you're going to be able to produce stuff.
I mean, meta can probably already produce stuff
where you can have digital, you know,
basically digital robots like in a virtual world
that really emulate humans,
the way they move, the way their eyes,
they're face tracking.
I mean, that's going to be the start of it, right?
But, yeah, I guess clarification, like,
are you specifically saying, like, humanoid robots,
like ones that walk around?
I'm specifically saying humanoid robots.
And what I'm referring to is fully autonomous robots.
And so that, I believe, is what people think of what they think of.
And for people that don't know, if you actually...
Googled my company and then Googled X Prize,
you would see something very funny.
Because I'm talking about the bear case,
but if you actually look at some of the R&D that we're doing,
it's hilarious.
We're literally building robot avatars for doctors
in the background as part of our R&D.
And the only reason I'm able to say this is because...
It's public information anyway.
But, but, but,
but,
and that's like super duper in R&D.
But what I was going to say is that,
you know,
the problem is I know so much about this problem that I,
I realize that AI is a great tool today.
I think when it comes to its ability to disrupt the digital world,
I think there is actually a there there.
I am still not convinced that we're going to see AI completely disrupt the physical world.
And we are eons away from that, in my opinion.
But again, Jeff, tell me why I'm wrong.
Go ahead.
Well, yeah, briefly.
Oh, I didn't hear what you said.
Sorry, what you say?
No, I was saying I was going to go to Jeff to tell him, tell me why I'm wrong.
Yeah, I'll take, Doc.
I mean,
I'll take a coffee or beer bet on the 10 years.
I think within five years,
Tesla will have...
It's not next year. It's not the year after. But I would say within five years, they will have a humanoid robot working in their factories doing tasks, doing complex tasks, actually, doing repetitive tasks. I think they're going to start in the industrial environment. They're going to start in the labor environment, and they're going to get it fully refined in that environment. And then at some point, it will go...
commercial to consumers.
But I would say within five years, they'll have robots,
and they'll be useful.
And again, it's not going to be like,
their whole factory is going to be run by robots.
It'll be no humans.
First off, there are industrial robots in their factories today.
They're just not in humanoid form.
And they do have vision-driven robots as well
in terms of some of the tasks.
but humanoid form, I would say, within five years, doing tasks in the factory.
Yeah, absolutely.
Jeff, why do they have to be, why do they have to be humanoid, by the way?
And this obsession with humanoid robots is really interesting.
But yeah, what's the point?
I mean, there's a point if you have like some humanoid robot that's your friend in your house.
But is there, and I'm earnest asking the question, I'm sure there is.
But what's the usefulness in a factory?
Yeah, and in a factory...
There's many things that are automated via robotics.
There's conveyors, there's arms, there's even material robots that are pulling material out of stock rooms and are kind of running them on a train track.
That already exists in factories today.
But you're not going to be, there are assembly, like when you're, when it's basically not everything is Z-axis assembly.
When I say by Z-axis assembly, like it's in front of you and you're taking an object, you're lifting it up in one plane, you're moving it to this plane, you're moving it down.
That's what today's robotics can do.
But if you want to go in any kind of like more complex movements, and in the case of a car, if you need to go in and, you know, there's more like if you're putting a seat inside of a car, some of these things are difficult tasks for, you know, robotic arms to be able to do.
So...
Does it have to be?
There's a gradient you can walk up in a factory in terms of like how much you can automate it with today's automation.
And you can do a lot.
I mean, the factories, the two new factories that Tesla built have a significant amount of more automation than the first two factories they built.
But yeah, going humanoid will allow you to do.
Basically every task that you would put a physical human or laborer in, you'll be at some point you'll be able to replace that.
You know, I'm very much, so there's two schools of thought, all right, in the robotics world.
One is that we're going to see hyper specialization that I like to call the Wally model.
And the other side, which is we're going to see multimodality and
exploration of the humanoid robot that can do everything, which I like to call the I robot model.
And so I just wanted to put those in there because those are in fan fiction and people do understand
kind of what I'm trying to say. So Wally model being, you know, where you have a specific robot with a specific
task that can actually use intelligence within that task, right? Because it knows everything about
that specific task. And then the iRobot model being like one robot that is being trained on all
the information and that will allow that robot essentially to win over. I actually, you know, I very
much sit on the Wally side. There's no doubt in my mind that there's so much
And to kind of talk a little bit about what happens in factories today,
we've seen that there's actually very little intelligence,
very little intelligence that is involved in those tasks.
The intelligence is involved in the physicality of those tasks,
not in the higher order decision making.
That is what's coming next, obviously.
A good example of that is what we saw with Google X,
and they literally have an entire warehouse with robots that are picking out items from a,
a bucket and no matter what item it is,
they figure out the right angle to get in there
and grab it and bring it out.
It's very interesting and that is,
it just speaks to the incredible challenges.
We're trying to not recreate the brain.
We're trying to recreate the brain
plus the spine. Those are very complex tasks. And so, you know, for people that are like true
roboticists, you realize how complex this is. And so, you know, from the robotics perspective,
that is what's challenging. And so, you know, we've not even done a good job of getting the brain
built yet. And now we're talking about the spine. So some would say that it's easier to kind of get
some of those tasks done, but I actually am very much
against that if you change the angle of approach if you change the motion and the speed and the
acceleration of approach you know we don't realize how difficult that is and so you know i i do want
to kind of push back on the fact that the humanoid robots are going to be helpful in enterprise
a lot of people that are actual deep actually deep in this space believe strongly that we're
going to see hyper specialization before we see some large humanoid robot take over
uh this really illustrates
This really illustrates the difficulty of understanding how AI will disrupt industries, I think.
Because when you look at a physical body, right, in which case robotics is trying to mimic the nuance of a physical body, to someone like you, Dr. Donish, that sees this and you understand the breakdown and why it's so difficult, it's obvious to you.
Sorry about that.
But...
I think it's even more complex to replicate some emotional and intellectual tasks, right?
And I see this-
Hilariously, sorry, I want to push back on that just a little bit because the social and emotional side,
you don't actually have to replicate it.
You just have to represent it.
And that's a very, you have to emulate it, in my opinion.
But sorry, keep going, Justin.
What you're saying is, like, truly having an emotional social task.
Sure, I agree with you.
But can you be 80, 20 there so that people actually think?
So, like, for example, with the elderly, there's actually this really interesting example of a seal that emulates a pet.
for people with Alzheimer's and it's gone it's gone like crazy like everybody wants it
and so there are examples of where it's just enough that somebody can can get it sorry I know
Matt you've been having a lot of trouble getting up on stage so wanted to give you a chance to
jump in wanted to give you a chance to jump in on and I have a very specific question for you
so you know weigh in first and then I have a follow-up
Yeah, sure. I think what we always tried to do is we go to the 1% when it comes to like how the future is going to look.
And I think we always forget about the other 99% of like the real world problems that we're trying to solve.
And I think a perfect example of that is just the U.S. Postal Service.
The U.S. Postal Service today in all the digital world that we live in after coming through the pandemic,
they still deliver over 125 billion pieces of mail a day or annually.
So we still have a lot of problems to solve in the real physical world that AI and robots and things will not overcome in the next five to 10 years.
And I really do believe that because enterprises and governments, governments, the largest employer in the U.S., they take forever to fucking change.
And we try to think that these enterprises are so fast moving and forward thinking like Tesla.
They're really not.
And there's so many different political challenges that need to be overcome to get a lot of this technology into the actual workforce that I don't like when people extrapolate the 1% of how these things are being used today by leading companies like Tesla and Nvidia and Google to the rest of the economy.
And we're at early stage investment fund at Ripple.
We invest very early in companies that hopefully will eventually break down those,
you know, walled gardens from large financial institutions and health care organizations
to change and use cutting edge technology.
But just as long as it took...
technology companies like Uber to change the taxi industry or Airbnb to change the hotel industry.
It took them 10, 15 years to get to where they are today. And I think that when we extrapolate what
we see today happening in the markets with companies like Nvidia and Tesla, it still is a very
small percentage of the rest of the economy that's going to be able to adopt this technology as fast
as you want it to. And then it opens up a whole other worms.
a can of worms on universal basic income social disruption workflow and work job placements and things like
that that if it comes too fast you're actually going to bite off way more than you can chew in terms of
a social dynamic there so i just want people to understand that matt so you know we were earlier
talking about i don't think you were in the room because i don't know if you would come up on stage
there was a lot of talk around uh around
the venture capital industry and how it has been trying to find the next hype cycle.
And, you know, and well, that's their jobs, right?
And so, you know, but, but what's been interesting has been one,
the same VCs that were pushing Web 3 and the Metaverse
are now pushing AI.
That to me is a bearish signal on AI.
Number two, you know, Matt, I keep it real, man.
I keep it real, so I got, you're out here,
you're representing your industry.
No, I'm not, I'm not, and I'm not.
I know, I know.
I get it.
I get it.
Listen, the shilling on, you know, on Web 3 from the people that were all, you know,
against it and then are now, you know, cheering rah-rah for AI, I think it's very hard to
swallow.
As an individual who works in the industry, I mean, we're in the most boring, non-sexy type
of investment focus with enterprise SaaS.
Obviously, machine learning and AI incorporated into the feature set of what our thesis is, but I think
The people that are trying to say AI is the next hype cycle,
they always forget about the fundamentals of actually how businesses
create value and cash and deliver that back to shareholders.
And they always skip to the finish line.
It's like when you look at a company that goes public,
And all you think about is that moment the company went public, but you don't understand the 10 years of hard work that went into getting there.
And because people look at investing as literally like the turning point in a company, investing is the easiest fucking thing in the world.
It's returning capital to shareholders that is the hardest thing to do.
80% of people die on the way down from Everest.
Why?
Why?
because they exert all their energy
and use all their resources to get to the very top of the mountain.
And they don't fucking realize that in order to actually survive,
you need to come back down to base camp.
And what a lot of people in this current moment are saying is it's so easy to invest in AI.
Come on, join the train.
It's going to be awesome.
What they realize is these companies still have to deliver.
Invidia still has to deliver the growth rate that they just forecasted in the last quarter.
And then that tweet that came out, I forget who it was a couple days ago, 100% of that revenue, not earnings,
that revenue has to go back to shareholders for them to justify the multiple on earnings right now
and the whatever what is it trading at right now in an EBITDA multiple like 70 times
it's crazy it's crazy to get that money back into shareholders hands so all those things
have to align I think people conflate investing with actually portfolio management and delivering
returns back to shareholders Matt I take the other side of some of this
all right let's hear the other side jeff go ahead yeah just like like it was kind of a
i agree a lot of the points by way matt it just the point of like companies like tesla going after
narrow aspects i mean i think automated driving uh could you're thinking about truck drivers taxi
drivers uber lifts chauffeurs bus i mean you're talking about 10 million jobs yeah and the u.s
we're going after narrow i said the rest of the economy
is not ready to adopt all of this technology just because Tesla is leading the way.
Of course, we want them to, and I believe in the use cases and the production of it.
I'm just saying when you think about implementing what Tesla has done,
Tesla has only been able to do what they do because they have the greatest entrepreneur of our generation leading the company
and saying if you don't do this, you're gone.
Try doing that in government.
Try doing that at the financial organizations.
Try doing that at the U.S. Postal Office.
It's impossible.
And those still represent a huge sloth of the workforce in the U.S. right now.
And that's where I'm saying it's like,
it's easy to do what Elon's done in getting this stuff adopted
because it's Elon doing it.
And he's telling people they have to adopt it or they have to build this part in-house
or else they're fired.
You cannot go buy this part from Magna,
or Daimler, Chrysler, or any of the other OEM,
you have to build this in house.
And if we have to do it with a machine or a robot,
that's the way it's going to have to be done.
Just like the way he did when he decided not to buy Russian rocket ships.
He decided to build them in house because he can.
But you can't do that when you work at the government
or some other fucking Fortune 500 company.
Yeah, I mean, that's, I think, an area of leadership, like, who leads the budget or, you know, for the federal government.
I mean, if humanoid robots, first off, the other thing that, you know, Tesla is doing, they're not designing a $150,000 humanoid robot.
They have a price target.
The whole program, the whole schedule is built on, you know, can this thing be $25,000 or less?
And over time with volume, can they get it down to $10,000?
What is that, what does that laborer cost?
You know, as a human.
And, you know, I just, I think the business case is, is fairly infinite.
Well, here's the perfect example of how the business case was infinite and then how it backfired.
Two companies just dissolved or sold off their logistics and delivery businesses, Uber and Shopify.
Two great technology companies couldn't figure out how to fucking make a logistics business work.
work. Uber sold off theirs and Shopify just sold off theirs to flex port. So the physical world
of implementing these types of technologies are very hard and they take forever. And Elon Musk almost
went bankrupt many times trying to do it until he was saved by, you know, Daniel or by the U.S.
government or some other things like.
it's very hard to implement this.
And I think where AI is easily implemented and we're seeing it today is when you take the human brain and match it with the output of voice,
the services that come with those types of jobs easily replaced.
And that's perfect because there's no interaction with the body and with other far away labor.
And that's where the easiest thing to just say, does my job, which I mean, I'll make fun of myself right now,
Me is a venture-cadro.
Bro, VCs are the first to go, Matt.
Of course, of course.
I'm disrupting myself here because my job all day is to sell on a fucking phone.
My wife jokes and says, your job is to take...
people's, you know, skills and basically interact them with other people's skills as a networker.
I'm a phone operator. I plugged the line into the other person on the other end, and that's all I do.
I just jack in phone jacks all day because I don't actually do anything.
And I can be replaced by AI for that.
Well, to pushback, actually, this is an interesting point. So as a former VC myself, you know, I think
Matt, man, you're right.
What you said earlier, you said investing is super easy.
Like, I think you meant like deploying capital is super easy, right?
But of course, getting it right over 10 years and cetera is hard.
So yeah, I took it to mean that.
But I do think there's some significant disruptions afoot.
And it's entirely due to the fact that before you could leverage capital because, you know,
you needed 50 people or so to make an MVP, right, 50 engineers to make an MVP to go to have product market fit, go to market and then, you know, leverage.
Now you need like one to five people to maybe do the same thing.
And, you know, I'm seeing that real on the ground, you know, like chat GBT as Mish and others said.
You just, okay.
Eugene, you just touched on something huge and you just kind of glossed over it, but I think it was a bigger point.
You just mentioned that essentially the trick of any job is in the nuance and that nuance is what most people don't understand.
And it's the thing that we say, like, so what exactly do you do all day?
And then, like, we have someone explain their job.
And it sounds like, well, I could have done that in the 30 minutes a week.
And nuances, like the devil's in the details with AI.
Yeah, I can do the broad stuff.
Like, I hear people talk about how doctors are done.
That's like they just diagnose all day.
But, of course, Donish will tell you, like, that's freaking hilarious.
That's not...
Like we'd like to focus more on patient care, but that whole like diagnosed, you know, running machines and cat scans crap, we'd prefer to have that all automated so we can focus on patient care.
So like the devil's in the details with this AI stuff, we can all see other jobs and say, oh, I see the stuff I can see the stuff I can see happening on a daily basis.
from my limited understanding of that occupation,
oh, that's simple, that'll be disrupted.
But then we don't understand the nuance of the job.
And I think that's where AI struggles.
Sorry, Eugene.
I just wanted to, that was a really good point you hit on there.
I like the meta commentary, actually.
The nuance and nuances, there was like some meta commentary there,
which I think was quite clever and nice.
Yeah, so I mean, I think that the disruption of foot
is about the amount of capital.
I mean, you know, VCs raised a lot of capital,
just like startups during the, you know, pandemic times.
There's a lot of capital sloshing around in the world.
And, you know, now there's potentially a few opportunities if the capital set is going to be lower.
I mean, it depends on the series, right?
Precede versus C versus Series A.
But yeah, I do see potentially disruption afoot because of directly related to AI in the VC space.
But it will be nuanced.
I mean, how, are you okay, Danish? How are you doing?
Yeah, hey, Fulman. Go ahead.
Yeah, yeah, Jeff. I mean, you mentioned about humanoid robots.
And my question to you is, and there's a reason I'm asking this, so,
I'm asking this, there will be like a follow-up question.
And why is it relevant to have humanoid robots?
Like, what extra utility do is it provide?
In the manufacturing environment, there are...
There's a set of tasks that you can automate, you know, with robotic arms, with things on trolleys and so forth, conveyors.
There's all kinds of automation we don't think about that are in factories.
But there are things, you know, based on the product itself.
Some products are very simple.
Can be built with 100% automation that exist today.
They don't have to be in humanoid form.
But there's things that require the human form or human body to be able to traverse through a factory,
to be able to kind of bend and wind through various different aspects of a factory that you can't just do with, you know,
planar movements in X, Y, and Z.
And you wouldn't want to build, you would not want to actually...
purposefully design automation for one product line that can't be used to the next,
to the next,
you actually would want something that's more reprogrammable with computer vision
that isn't fit to that product,
that size,
and so forth.
So to be able to,
you know,
there's this gradient of tasks in a factory.
You can build,
like for example,
a smartphone,
you can do the whole smartphone assembly manually.
And that's what,
there's more of it done manually than you think.
in Asia today. But by task by task and person by person, you pull those 300 people off the
assembly line, there's a level of difficulty of like, in terms of like, can I automate this?
And can I get a five-year multi-year return on this automation based on the form factor and the
product? Or is this form factor and product going to change so much over time that I would need
something that's more...
malleable and more reprogrammable.
And that's where computer vision and humanoid robots can come in and replace more of the labor that people just thought could never be done before.
I'll just push back a little because, again, you know, I'm very much convinced that you kind of said partly, partially some things that I agreed with and some things that I very strongly disagree with.
A simple example, a humanoid robot has two arms.
Wouldn't it be nice if they had four?
Wouldn't it be nice if they had six?
Yeah, like literally, man, like we, there's so much information around humanoid.
The idea that a humanoid robot would be the right form factor for a factory,
I have to push back a little.
Now, in terms of traversing your home, the home, all of the regulations,
everything that's built in the home has been built around the underlying, you know,
point that pay, uh,
humans will be traversing the home, right?
And so...
Even then, down after...
Even when it comes to the home,
like, if you could traverse the home
in a much more better, efficient manner,
if you weren't exactly in a human form,
for example, you didn't have two eyes,
but maybe you had, like, four eyes.
If you didn't have, like, two arms, we had four arms.
So the reason I asked that question specifically was because essentially the real reason people
want a humanoid is because of like that human feeling and the hope that basically that
robot may even have some form of interaction or intelligence.
That's the real reason why you want a humanoid form, isn't it?
Yeah.
And actually to build on that, I thought just answers is very practical.
And so Radonish's, I think even if there were no practical answers, like if there's no
reason to have a human robot, like no one could even invent one, which I don't think is
necessarily true.
Yeah.
You know, I mean, even since why did, why do we have Greek choruses, right, from like, you know, 500 BC or whatever, right?
Like the first time man was able to, you know, chip away a stone.
I mean, they made, you know, cats and dogs and other things, but they made human looking things, right?
This was like thousands of years ago.
Right.
right i mean there's this will to anthropomorphism right i mean ovid had this idea of you know pygmalion the
sculpture who created galatea and then fell in love with this uh sculpture that he made right i think we just
we humans just have this i mean this philosophical point now have this desire to anthropomorphize everything
and even if there was zero utility which i don't think there is or a humanoid robot eventually when we
can i think human will humanity will i think it's just inevitable right yeah yeah i agree so
essentially but then when you at
As you know, and again,
not to bring philosophy into this,
but when you anthropomorphize anything,
you limit it, isn't it?
So, for example,
from a philosophical perspective,
even when one of anthropomorphizes God,
you essentially limit,
limit it. So limiting things isn't probably the best way to go about things when you want to have
optimum utility. And so in reality, from what I believe the reason that it's done is what we've
mentioned, that essentially is because of, again, feelings, being seen something what's more amiable,
and the hope and desire, which I don't believe will ever come into fruition, is intelligence.
Right.
Yeah, Jeff.
I was just going to, I was just talking, I mean, what I, you know, before my current role,
consulting with Mega Cap, I mean, I designed how products were my design, how the first Android
phones were made me in my teams and designed those factories and design what level of automation
goes into them.
And what I will tell you, there's a whole economics piece that we just don't, we won't have
the time to talk about on this space, but I can summarize in, you know, maybe 10 seconds, which is,
There are many things we choose to not automate because the economic return will never be there
and both the form factor sizes of the device change versus, you know, current conventional automation that exists today.
When you can go humanoid and you can go computer vision and you have the ability to take these tasks
that will change over time and over products and be able to basically adapt and reuse that same
piece of equipment. Now you're, you, you, you can put a 10 year life on something that could have only
been a one year. And the key thing is, can you get the dexterity and can you get the performance
and can you get the speed and rate out of the humanoid robot or not? I believe you will be able to
over time. I just want to tell you, like, there's a whole economics piece.
of like, why can't you just automate conventionally?
Why does it have to be humanoid?
And that whole economics piece, I think we're going to find to be very interesting.
Because that's why things are not 100% automated today because there's, you wouldn't have the return.
You wouldn't, your fixed cost structure would just cave in on top of your variable cost structure.
You'd never be able to ship a product successfully economically over time.
That's a very that to me that is the most salient point right there.
That was the best point because ultimately if you can build one thing over and over and over again,
it's obviously going to be more economically viable.
You could scale that.
I love that pushback.
Patrick, was there anything else before we close up?
Your hand has been up for a while, so wanting to make sure.
I know.
I had time to fall asleep, guys.
Come on.
Oh, Jesus, Patrick.
Just because we're not talking charts doesn't mean it's not interesting.
Go ahead.
Patrick, go ahead.
You might have called me to go on these calls to do some charts, but I listened to you guys.
And I love listening to all that because it's more pieces of the puzzle that I put in, right?
So like, let's say I put my hand up because I think the original point was we had to somebody bearish, right, fundamentals on AI, right?
And I'm telling you guys, take a notice of this.
When the price action is going down on the daily chart, on the weekly month, notice how that's when people start blaming stuff, right?
Let's say when Bitcoin top and it broke down or whatever.
Like, I don't want to pick on Bitcoin, but then, oh, well, it's not the future.
CDBCs.
They're going to ban.
You know, this news.
They've always been there, these news events, because I remember Bitcoin back in 2012 or 13, when I was living in France.
Oh, the French government, they're going to do something about cryptos.
And that was when Bitcoin was, you know, a whole way back, you know.
But that type of news comes out when the price goes down and the inverse.
When price goes up, people, they look forward.
They feel happy.
Oh, this price going up.
Why is it going up?
Why?
People look for answers.
Then we talk about all the positive stuff about something.
But.
This is a takeaway, guys.
The price led what we're talking about, the price.
Think about this.
It's very hard to be a true contrarian.
To have bullish news when the price goes down week after week, we hit these people with rocks.
Let's say somebody's bullish Bitcoin while the price going down.
Nobody's going to want to hear this guy, right?
So it's, that's all I had to say about that.
Remember, price often guides us more than guides the actual news that we hear on spaces or CNBC or whatnot.
Price is speculative.
Sorry, go ahead, Cody.
I apologize.
I was just saying price is speculative, you know.
Remember that these are businesses, these are real businesses interacting in the real world.
And there is a time cycle there that is probably longer than what most people anticipate.
Yeah, I mean, Patrick, to be fair, if you just listened to the price, you know, two years ago, you probably wouldn't have invested in NVIDIA, right?
The time to go to NVIDIA, like I said, was back in 2014 or 16.
Nobody was talking about AI back then.
But that was sub-10-20, and you would have been riding a 10-year bull run as US
equities were breaking out versus PPI, as U.S. equities were breaking out versus gold.
That was the time to go bonkers.
That's when Bitcoin was sub a dollar, $5 and went up to $60,000.
Game-changing type of wealth starts.
After huge, huge beat down, after four, five, six, eight-year base, that's when you want to come in.
But isn't this just like, sorry, I have to close on this, but isn't that just hindsight?
You're just looking backward and saying, oh, you should have done that?
And it's like, well, yeah, I should have also invested in Apple when it was a dollar.
Yeah.
I mean...
It's not about investing at Apple and a dollar.
I put a tweet out.
It's about investing in Apple
when SPX has put in an important base
versus its competitors,
gold, inflation, PPI,
and it's breaking out.
That's when the probabilities
of headwinds becoming tell wins.
They shift.
That's when the...
the fundamental, the institutional money starts buying the dips and then you have a meaningful
bull run. That's chart trading. We have 100 years charts. You can look at my charts, guys.
When these cycles shift and they turn, it's not 100%, but it's telling you, hold on,
the momentum in favor of one asset versus the other is changing. And if you don't look at that,
you're fighting an uphill battle.
And on that really helpful advice and note, we closed today.
Remember, we're here at 8 a.m. Eastern every morning.
Love this space.
Thank you, everybody, for joining us.
And thank you everybody who's commenting on the bottom right.
We'll see you tomorrow.
Thanks, everybody.
