Projects
![]()
No projects found
Investors
New
DAO Treasuries
DATs
Lend / Borrow APRs
Chains
Categories
Spaces
AI Hardware Frenzy, OpenAI’s $100B Bet, And Apple’s Chip Power Play🎙️
Recorded: Sept. 23, 2025 Duration: 1:08:19Space Recording
Short Summary
In a dynamic discussion, industry leaders explore the rapid advancements in AI technology, highlighting significant project launches, strategic partnerships, and emerging trends that are reshaping the tech landscape, particularly in the context of global power dynamics and consumer-driven innovations.
Full Transcription
Thank you. Thank you. Music Thank you. Thank you. Music Thank you. Music Thank you. I'm going to go to the next video. Thank you. Thank you. Music Thank you. Thank you. All right. Welcome, everybody.
Hey, Noah. How are you?
Good morning, Lewis. I'm good. I'm good. I'm just waking up like an hour ago on the West Coast.
It is bright and early here in sunny California.
Excellent. There's a lot happening. It's an hour ahead in sunny Utah. So welcome to everyone to
this week's AI newsletter and news coverage. It's been a really powerful week. In the last seven days,
you know, AI proved it's not, you know, so much about the algorithms. It's who owns the power
plants, you know, and the chips and rules for sovereignty. We're seeing lots of different
countries starting to get in on that. I mean, OpenAI wants $100 billion worth of nuclear
scale servers. but while I can
launch an app in minutes building power infrastructure really takes a lot lot
longer certainly the blockchain people learned that lesson and Apple just
turned your phone in into a sovereign state we'll talk about that and
meanwhile the global South is coding a future that no no longer needs Silicon
Valley's permission and so it's's really interesting as we watch the broader geopolitical and economic stuff
happening with all the infrastructure rolling across.
It's been a powerful week for that, that's for sure.
I mean, Apple unveiled that they're using their A19 Prochips explicitly optimized for
on-device AI.
As we saw, NVIDIA pledged $100 billion into open AI data centers while investing $5 billion into Intel, which is a really interesting play.
So lots of stuff happening on that.
And then what else have we got going on?
Altman teased an expensive new chat GPT features coming out, while regulators
are kind of getting their knickers in a knot about monopolistic control. And first time
AI scales, economics is directly clashing with public utilities and national grids as
they really start to suck a lot of power out of the environment.
The other part of the story was the Apple launching of the chips,
and they're folding MCP support into the Mac and the iPhone and the iPad. So unlike cloud-first AI,
this makes AI a baseline feature of personal devices. So that's a big shift as well.
Noah, did you see anything pop up that caught your interest in the last seven days?
interest in the last seven days.
I think the Apple chips are exciting to me because we all use iPhones, and I'm curious
to know what the implications of that are.
These AI-optimized chips, do you think Apple's going to be able to close the gap?
I mean, if they could, that'd be incredible.
With OpenAI, if everyone on Apple, if everyone on their Macs and their iPhones
had LLM access across the board,
I think that would be a game changer.
Absolutely.
I mean, the thing I've noticed is
Apple always seems to be late to the game.
I mean, you know, you look at a lot of their stuff.
It's true, they've done some, you know, leading edge stuff,
but for the most part,
they seem to want to watch the marketplace. I mean, we, they've done some, you know, leading edge stuff. But for the most part, they seem to want to watch the marketplace.
I mean, we've seen Meta release, you know, those, the VR, AI glasses, right?
And yet we've heard rumors and, you know, seen design stuff around Apple Glass
and that it's coming, it's coming in 2026.
So I think it's interesting.
I think they really want to watch the market, see what's happening, and then they just come in and kind of crush it.
So I wouldn't be surprised that what we're seeing, especially with the recent OS26 with the glasses view.
I mean, for me, the way they're treating the icons and the way they're treating the UI is definitely a precursor to some sort of VR, AR play.
I mean, you just look at that and you go, OK, yeah, eventually I'm going to be able to have glasses that look like that.
But again, I think they're just playing it safe.
You know, this idea of folding MCP support into the Mac, into the iPhone, into the iPad, into the chip first so that
it's on-device AI, right? Because it's clear that cloud AI is really expensive. And so
trying to push this down further into the hardware, it seems to be a big focus. And I think
until then, they don't, my guess is they don't believe they can really deliver a top quality
product or a service.
I mean, that's my thought.
What do you think?
You know, the best artists steal.
And I think Apple is just watching its competitors innovate, make mistakes, innovate, make mistakes.
They're taking notes.
And then they're going to come out of the trenches running because that's what they've done historically.
And they seem to be able to close the gap,
if not just completely overtake their competitors,
even though they had a late head start.
So I think, yeah, I don't bet against Apple.
There's few people I don't bet against
and companies I don't bet against.. There's few people I don't bet against and companies I don't bet against.
Apple's one of them.
I think the AI hardware movement that you highlighted where consumer tech, software compute, and energy converge is also really interesting to me.
You know, the sub-flagship hardware that can rival GPUs at fractions of the cost, the democratization of that kind of compute is exciting to me.
I wonder how far out we are from that, though.
Well, I think what we're seeing are two different drives.
I think the drive on the cloud infrastructure with the, you know, buying nuclear power plants and sucking up all the electricity.
I think that's driven by corporations wanting to get as fast as they can to AGI, to artificial general intelligence.
Meanwhile, Apple is looking at it from a consumer base, which is saying, okay, the consumers
want AI.
How can we deliver them a quality AI service that's personalized and that's efficient and
that flows really well?
And that has to be device, to some extent, that has to be on device, right?
With some cloud probably, but a lot of the stuff has to happen on device.
So I think the drive from Apple is really to focus on the consumer.
I mean, that's their power.
The Apple ecosystem is just such a powerful thing and that's their sort of crown jewels,
if you will.
And so I think given that crown jewel, that's in some ways why they're really okay to sit back and watch the market.
And then once, like you said, the competition works the bugs out, they crash and burn in some areas.
Then Apple watches that.
They tweak, they tweak, they tweak in-house.
And then finally, they just go wham.
And you get an amazing product that's easy to use, simple.
You don't need to be a PhD to use it.
And it just works.
That's that.
And that that that progression to get to that point of that level of technology just takes time.
And it takes a lot of things to break.
And they know that their ecosystem expects their anything that Apple delivers to not break.
So I think we're just seeing two different massive trends occurring.
A lot of the cloud stuff also is just many countries, you know, like I mentioned the global south and people like, well, what's the global south?
Right. So the Global South essentially is regions that may be historically a disadvantaged in global trade and political influence like Africa, Latin
America, some parts of Asia, Middle East, Oceania, excluding Asia, Australia and
New Zealand, probably. But, you know, how are they setting themselves up? And what
we're seeing is massive investment in cloud and setting up massive data centers.
I mean, we're seeing that obviously across India at the moment, a bunch of announcements
on that, but all of those things require power and open AI and Nvidia's hundred billion AI
plan is apparently going to require 10 nuclear reactors of power.
And the trouble is the power infrastructure, it's hard to build fast.
I mean, it really, really is.
I saw it when the blockchain and the Bitcoin stuff were happening and they were trying to set up crypto mining.
It takes a while to set up the infrastructures to support all the electricity that's needed for all those devices.
And the same thing's true with GPUs and cloud data centers.
So that's my thought.
What do you think? Yeah. Also, where are they building this infrastructure? I was reading that
it's the first time. Well, we're seeing data centers. There's been announcements of new
data centers, I think in Chennai in India. And we saw, you know, with Trump and NVIDIA and
Sam Altman being in the UK, I think it was last week, announcements
of building the UK part of it.
I mean, that's the whole, I remember the whole Stargate announcement, essentially building
a global network of AI centers across the world, right?
So obviously, we saw the announcements in the US.
We now saw the announcements in the UK.
We're seeing announcements in India.
Where's it going to happen next?
It's really interesting.
Someone's got their hand up.
Dap it.
Go ahead.
Oh, God, I hit create clip by mistake there.
Almost did something weird.
How are you guys?
Good to hear you.
We're good.
How are you?
Good, man.
Thanks for having me.
I wanted to comment on the data situation there.
So I live in Prince William County, two counties away from DC.
And property value here is it's not low enough to not have infrastructure, but it's right
in that sweet spot.
I have seen in my area just recently recently probably eight or nine data centers open up
but the issue that we're running into in as a county is we have enough power infrastructure
to handle it but not enough to do it well so ironically all of us are having to pay more
to support these things and I just saw today um Meta started a super PAC to fight state regulations because this is happening all across America.
And like, obviously, people don't want to pay more on their electric bills for data centers. Right.
So states are starting to create barriers to these data centers. And it's pissing off the big guys.
Meta is going to be fighting that big. big they just dropped a few how many millions of
dollars to to start lobbying to make sure that state lawmakers don't listen to us and listen to
that it's messy generation go ahead i'm sorry it's it's messy for americans right um we need the ai
but we haven't been investing in our infrastructure the same way a country like China has over the past, you know, 10, 15 years.
And now we're paying for it.
Yeah, there's been a couple of stories even across the last two or three weeks about electricity prices going up as what people believe to be a direct result of, you know, new new data centers getting pushed out there and starting to suck a lot of power.
I mean, we saw that in Texas when a whole number of Bitcoin miners went down there and started setting up Bit9 mining.
The difference is, though, that they did a deal with the state
where if there was an electrical emergency because of the miners,
you know, it's all software, and they could turn around and basically turn it off and allow the grid
to kind of recalibrate itself so that they weren't blowing the grid up. I'm not
sure if AI cloud, AI data centers are doing the same thing. I suspect not, but
it's definitely a problem. And so I think that's why maybe, you know, my reference to the Global South with, you know, Africa and other countries like that, where electricity is a lot, lot cheaper.
And maybe they don't have the infrastructure that's really kind of, you know, using a lot of electricity, maybe better positioned to put those data setters in, you know, putting it in a first world country like
the US or any of the others where we essentially using like maybe 90, 95% of the grid anyway,
all the time. And now we're adding, you know, AI data centers in, it's only going to crack it up.
I mean, Noah, what's what's what do you think.S. going down a pathway where we're going to blow up
our electrical grid? Yeah, I was wondering if we're going to blow up our electrical grid or
just blow up other electrical grids. You mentioned that. That's why I was asking you where these data centers are going to be built.
And you said India.
I am curious to know if we have space in the U.S. to build them.
I'm just curious.
I saw someone post about how this is going to affect, quote, marginalized communities more so than anyone else.
And I can see there being some truth to that.
I just wonder if our infrastructure can handle the kind of compute power
these guys are trying to push out or if we're going to have to outsource space.
Like you said, countries like India.
Yeah.
Yeah, I think India, like I said there's this data center is going into the
uk um there's probably others around i'm just going to do a quick search into uh
yeah we're seeing ai revolution second yeah in video ce CEO I mean he's flagging electricity constraints as a key
limitation you know and they're saying basically AI will drain more electricity that's for sure
and uh what else I'm seeing uh we're seeing Korea uh also getting involved in in it as well with the
data centers and yeah it's it's definitely metas 50 billion ai data
center triggers energy supply worries climate crisis 24 7 um lots of things open ai oracle plan
a 4.5 gigawatt start gate data center so yeah lots of stuff happening on hardware i think but again
happening on hardware i think but again i my view is is that this push by major by the major
enterprises open ai meta oracle and all these others is because they're focusing one on the
enterprise and two there's a huge push to try and go towards agi um and and so we're seeing a lot of
different trends kind of pushing this to a brink.
Whether it actually breaks stuff, not sure.
But I think, as Dapit mentioned, it's raising prices or it has the potential to raise utility prices for, you know, everyone across the U.S. where these data centers are going in. And, you know, it's hard enough now with, you know, food and everything else going up to, you know, go, oh, my God, my utility is going to go up as well.
It makes you think we're kind of stuck between a rock and a hard place because, as Dapit was saying, this is the new arms race.
This is the AI arms race. And those that fall behind are
going to feel the pain later on. On the flip side, you're talking about energy constraints,
and you're basically highlighting that this is going to not just increase costs, but
disrupt power grids for human beings just trying to live. And so what's the answer, right?
What do you, do you regulate?
Do you slow it down?
Do you risk falling behind?
Or do you just hit the gas like open AI is doing pedal to the metal?
Because if you're not first, you're last.
It's disappointing that the US administration appears to be moving away from renewables, you know,
solar and wind. We've seen a couple of projects where they've pulled the money out, you know,
and I think, I mean, that just puts China and the rest of the planet in a much better position. I
mean, China's obviously done huge investments in renewables and solar and wind and so on.
And so that's a great way to really offset and to allow the support for AI, which is
going to demand more electricity overall.
I mean, but then again, you can also apply AI to forecast demands.
You can apply it to the balancing of supply in real time. You know, it's no doubt it's super complex, you know, in these markets where it's either a regulated market or it's a deregulated market or it's a fragmented market like across the U.S. in some ways.
I mean, Texas is a completely separate grid.
grid. So yeah, it's hard to say what's going to happen. But I do believe as an overall trend that
moving away from renewables is probably not a good idea, given we're suddenly getting this
massive need for an increase in electricity by AI. So yeah, I mean, there's a lot of potential for
inequities occurring, you know, with consumers
getting pinged on their electricity bills and a whole bunch of other stuff and rates
going up. It's just nuts. So whether it gets offloaded into other countries or not, I mean,
it seems as if the major companies are making huge investments into the u.s for cloud
but we'll see yeah ryan go ahead
i'll click yeah it looks like he rugged really quick let me bring him back up
okay no worries yeah i mean you know
that was me last week go ahead oh yeah i i don't it's funny i don't think i've ever been on spaces
where there hasn't been technical difficulties so it's kind of bizarre um that's a rabbit hole
we don't want to go down they should drop their their code base into Grok. Maybe it'll fix things.
There you go.
Or maybe they vibe coded this entire thing.
And that's why.
There's been stories going out of people who've done vibe coding and it's created spaghetti mess.
And now they're begging professional voters to come and fix this stuff.
So maybe this is what we're seeing.
Go ahead.
Sorry.
That sounds about right. come and fix this stuff you know so maybe this is what we're saying go ahead sorry that sounds
about right um yeah so it's funny because you know we're all looking at all the same indicators
here and we're all looking at all the same puzzle pieces and whereas like six months ago uh just to
dovetail what you just said six months ago we all said oh you know vibe coding is going to be the
end of the engineer and you know ais are going to replace coders and now it's, oh, you know, vibe coding is going to be the end of the engineer. And, you know, AIs are going to replace coders.
And now it's like, oh, wait, you mean you still have to have someone that knows how
this stuff works?
So we're finding actually, interestingly enough, there's a boom in job security for
good engineers because everyone and their mom has started vibe coding.
So all those ideas for apps that they've been sitting on forever,
now they can go to Lovable and they can create a mock-up in five minutes,
but now they need actually someone that knows how to do something with it.
So it's actually created a lot more work for engineers.
Now, eventually, will AI take it over?
Yeah, but not in the foreseeable future, maybe five, six years down the road, but a lot needs to happen between now and then.
market for years. And I started saying this, I think, three years ago specifically,
was that Bitcoin miners have unknowingly accrued some of the most valuable resources on the planet
without even realizing it. They've been scouring the planet over the past 12 years looking for the cheapest power because
Bitcoin mining is a bounty on cheap energy. So anywhere there was a favorable government,
stable infrastructure, and cheap electricity, a Bitcoin miner would pop up. And we've had that
all over the globe. Now what happens is all these Bitcoin mines that have all this electricity, now everyone's eyeing them for data center compute.
The difference is a Bitcoin mining facility and a data center are two very, very different animals.
One requires clean, relatively professionally run systems and security, whereas Bitcoin mining is people shoving stuff on racks in a barn and putting some dust filters and swamp coolers up.
It's very, very different.
The other aspect, too, is the bandwidth is different.
The bandwidth for training data models is crazy.
So you have to have very, very high-end network switches.
You have to have very, very robust networks.
And Bitcoin mining is not like that.
You could literally mine off of a cell phone signal with Bitcoin mining.
You don't need very much bandwidth at all.
So there's a lot of different factors that go into AI to the point where someone was telling me,
I want to say it was my buddy Russell was telling me that a typical AI compute costs about three times as much to deploy as Bitcoin mining.
So if Bitcoin mining is about a million dollars per megawatt, then AI compute is about three million dollars per megawatt.
So it's very, very expensive to deploy, time consuming, and it just requires a whole different level of infrastructure. But the power contracts are there. So the smartest miners are leveraging their ownership of power contracts in order to gain an upper hand in the AI market.
not financial advice, by the way, but this is actually one of the reasons why I invested in
Riot was not because of their mining prowess, but because their ability to trade energy contracts.
And I saw that they were doing that several years ago. And I was like, well, that's a pretty good
angle to have, especially going into AI. Don't necessarily build out the AI infrastructure,
but trade the commodity that you've been able to corner in the market.
Yeah. Yeah. That's really interesting.
I want to follow up on the comment you made.
You know, like six months ago, we thought that, you know, vibe coding was going to take over and everything like that. And I think what we're actually seeing is the distinction a little bit between the front end and the back
end of coding, right? I mean, I know professional coders span both front end and back end,
but, you know, I think with Vibe coding, you can quickly create a front end piece. It looks great,
but then when it comes down to systems architecture, data modeling, API integration, security, it's like, oh, wait a minute. And so I think, you know, the
term vibe coding actually does it a disservice. It might be great for proof of concepts, but
for production level stuff, we're still just not there yet. and and i think anyone who thinks otherwise is kind of crazy
i mean what's your thought i mean this has been my my life for the last 20 something years uh
you know there's there's there's a lot of a lot of thought goes into how a system is architected and
how data is structured right yeah um a certain data structure will work if you're, you know, you just have a couple thousand
records or you're working with just a couple users, but the data structure quickly changes
if you're working with, you know, hundreds of thousands, millions of users or hundreds
of thousands, millions, billions, or trillions of records.
The way you architect a data structure or a data schema in this case, and how you index
the data is, you know, make it or break it at scale.
So system architects and, you know, big system or at scale data engineers are still very
much in high demand because it's kind of an art to know how to work with the stuff.
Now, it's not saying that AI can't replace that
and know how to do that,
but I'd say it's still very, very early in its ability
to come up with data structures that are going to scale
and deploy and monitor and then maintain.
Because at some point you have to have not just the general knowledge to stand it up,
you have to have the general knowledge to maintain it and keep it going, right? So it's a difference
between knowing how to design a car, knowing how to build the car, knowing how to drive the car,
and knowing how to keep the car running.
Yeah. Yeah. Yeah. It's like we've got a whole community of drivers who are now building cars.
Exactly. Exactly. You know, so and vice versa is just just because, you know, you design the car doesn't necessarily mean you know how to build it or drive it.
So people are taking for granted the complexities of some of these systems.
Yeah.
Yeah.
So, but, you know, it's, it's, as you kind of raised in the first part of it,
it's like, you know, yet, you know, soon at some point, you know, AI will get that smart where it can start to pick up some of that stuff.
And, you know, the Lego blocks will be more sophisticated and so on.
But we certainly aren't there yet.
And now jumping back to the stuff around the utility stuff, I just picked up some stuff because we've talked about power and electricity needed for data centers.
Another thing that's coming up is water.
AI crunching data centers, they're sprouting across the West, but there's a lot of
not only local grid strain, but water usage strain as well.
And an interesting lawsuit popped up, I think it was in Wisconsin, where an environmental group
is suing the city of Racine, Wisconsin,
to release the projected water usage data for Microsoft's $3.3 billion data center in Mount
Pleasant, you know, because the city apparently has failed to give them any data. So, you know,
the impact is not only in terms of the electrical grid, it's also in terms of the water needed to cool all this these data centers as
well um so it's interesting that water usage is also popping up yeah go ahead ryan yeah it's it's
funny that you brought this up so my my dad was uh an engineer a nuclear power plant for like 35 40
years um he he lives out in arizona now where where one of the larger nuclear plants in the US is,
the Palos Verdes, or sorry, I always mix up the name because Palos Verdes, and
yeah, I think it's the Palos Verdes nuclear power plant. And they produce about six gigawatts, I believe,
between six and seven gigawatts of electricity.
And Microsoft came in,
I want to say it was like a year and a half,
maybe two years ago.
And they basically said,
yeah, we're going to build a data center out here.
And they're going to take up about 25% of the load,
25 to a third of the load.
So you're looking at, wait a minute,
you're going to be building out a 1.5 to 2 gigawatt data center
in the middle of the desert.
And just what you said, water is incredibly in high demand when you're out in the middle of the
Arizona desert, and then you bring in a massive data center that's going to take up, you know,
1.5 to 2 gigawatts of electricity. That became actually one of the biggest sticking points
with the data center out there. I believe it's called the Goodyear data center that they were going to be building.
But then when you actually start looking at the power generation in the U.S. and the
majority of where the miners are in the U.S. because of the power generation, it's Texas.
It's desert climates.
I mean, we do have some upstate New York stuff because of hydroelectric,
but, um, but most of it is, uh, is along the Southern side of the U S so it's, it's, it's
really interesting, uh, that water is actually going to become one of the bigger sticking points,
I think, than electricity in the near term, uh, because it, it does, uh, it's, it's, uh,
not something that you can just create by
bringing in fuel and building infrastructure you're actually pulling it from natural resources
which is very limiting right i mean in in there's a story in tucson where project blue was a proposed
large data center in pima county it faced. It faced a lot of opposition. And so
Tucson passed a new ordinance requiring that large water users, which is data centers,
had to supply detailed water use plans. And apparently the whole thing was put on hold.
And so that's super interesting. Meanwhile, in Texas, San Antonio, two data centers used 463 million gallons of water over 23 to 2024, while the
region faced drought and water conservation.
So it is interesting that while we're hearing stories about the electrical grid, there's
also lots of equal concern and issues popping up around water.
I mean, that's just nuts.
Yeah, it's fascinating. So I've been going through this book by Ray Dalio called Principles.
And early in the book, he talks about how every economic system on earth can be looked at as a machine.
And that's kind of how he got started with his investing career was he was in the agriculture industry and he was looking at basically a chicken is a machine that equals basically a chick
plus feed. Therefore, you know, you can,
you can drive a chicken futures as a commodity from feed futures and they're
in, they're interlinked. And, and then he was looking at cattle and, you know,
cattle is, you know, water plus feed. And he started building
out these models of how everything's connected to drive his investment decisions. And now we're
looking at data centers. And now we see that data centers require water. So now you have,
it's not just electricity for data centers, but it's water for electricity, or sorry, it's water for data centers.
And now we have, you know, it's a very complicated machine, but it just kind of tickled me because the more I look at this stuff from that investment point of view with people popping up data centers is they can be really deconstructed back down to multiple
commodities and resources. And we're going to start seeing a crunch in this country and in the
world in general, in these different areas. So I think a lot of people are building investment
portfolios and models now around these different areas where it's, you know, they're looking for
the equivalent of the grain to the farmer when it comes to data center planning.
Yeah, I mean, we're also seeing, I mentioned about the Global South, there was an article in
the last week where Africa unveiled its first AI data center in Uganda,
which is expected to cost $1.2 billion and run on 100 megawatts of renewable energy, apparently.
I mean, you know, with this sudden craze for AI data centers across the globe,
I mean, it's one thing for first world countries, which probably have some level of a mature grid and water system, which is going to get strained
anyway. But countries that, you know, the poorer countries that probably don't have
great infrastructure, either in terms of electrical grids and or water, suddenly throwing AI data
centers seems to be like throwing gasoline under a fire, but it's just, I don't know.
I think it's going to stress the planet.
What's your, Ryan, your hand's up.
Is it still up or is it, you want to comment?
It should be down.
Captain Levi has his hand up.
Oh, go Captain Levi.
Hey guys.
So a quick one just popped into my head based on the way Ryan was explaining the complex ecosystem of moving power to the desert.
And obviously, why I like this AI space is because it really spirals out of a lot of scopes that we couldn't even think of.
I really wish that Ryan's know around that was here to you
know on these spaces to tell us exactly how you know this will be solved and as well as um a
geologist so to speak you know to tell us um what we are not seeing but one thing that concerns me
is that um for every action they they should be an equal or positive reaction so um dragging water to cool down this
uh middle of the desert uh data center so to speak unless the water is going to be drilled
right there in the middle of the desert or going to be drawn from a water source somewhere
are there any implications we may not be seeing and um how are those um how are those going to
be mitigated so so to speak.
Sorry, can you repeat the question?
Okay, so the cooling system itself.
For instance, I think one of the experimental cooling systems that Microsoft did was, you
know, putting a data center in the middle of the sea and um following that thread in super so they
were you know engineers from all walks of life of all genres of all skill sets while some were for
it others were against it so um you know the problem of you know these innovative solutions
is that sometimes they being innovative they also know, bring about their own problems. So are there any problems we may not be seeing that may be affiliated with,
you know, these innovative solutions, so to speak?
Yeah, I mean, in terms of, you mentioned data centers in the ocean.
I mean, China's been putting data centers in the ocean to keep them cool since 2025.
And, you know, we've seen the project Project NATIC which was Microsoft's underwater data center
project as well and there's been some others as well. In the last week we also
saw researchers are exploring putting data centers into Earth orbit which is
interesting that it's finally cheap enough to do that because I would have
thought that lifting hardware would have thought that
lifting hardware into space like that you know might be cost prohibitive but um at least people
are starting to look at that but like you said you know um there are there seems to be some
solutions around this that maybe we haven't explored fully um but yeah go ahead captain levi
fully um but yeah go ahead captain levi oh no uh my hands no longer raised i dropped my hand because
i wasn't listening to what you're saying no worries i mean they're even looking at you know
throwing stuff on the moon uh like i mentioned putting it in orbit but i think the ocean stuff
seems to be you know uh cooling and you know it solves a bunch of problems and also there's the opportunity for
renewables to really have better access to them there as well so yeah interesting stuff happening
around water electricity and ai data centers and infrastructure as it as it pushes stuff through
um some other stuff that's popping up in the news as well while we're kind of going through stuff
Some other stuff that's popping up in the news as well while we're kind of going through stuff.
Robots.
A lot of things.
If you've seen the newsletter, there's been tons of stories rolling through on finally logistical robots, you know, already earning returns.
Obviously, Amazon's a big pusher of that.
There's been a huge hype cycle on robots and so on.
But I think it's masked the ROI.
We're now starting to see some transparent ROI benchmarks occurring where they're paying off.
And we're seeing a lot of development around that to actually say, hey, how can we actually make money on this?
actually make money on this? How can we improve stuff? Forklift, robots, capital loves the hype,
How can we improve stuff?
but industry needs the uptime, as they say. So lots of stuff around robots and getting more
into the commercial viability. As I mentioned last week, it's still my prediction that in the
next two to three years, we'll start to see it on the consumer space, but it's going to take
a while to get the AI up and running in that.
NVIDIA has been doing tons of stuff in terms of the robotics systems and setting up environments
in software environments for them to understand the physical environment.
So we'll see, I think there's lots going to be happening in that, and we'll start
to see more stuff coming out. At least the payback's starting to not be so murky, but we're
starting to see some real numbers popping out. Noah, what's your thoughts around humanoid robotics?
Do you need one in your home? Yeah, we've talked about this before. I'm looking forward to the humanoid robot that makes me breakfast every morning and does my laundry and cleans my floors.
I think it's cool. of robots and how they're going to basically displace a lot of
made type work for lack of a better term,
or even just do things that humans don't like to do on a daily basis so that
we have more time for creativity or exercise or self-expression or whatever
it is that you want to spend your
life doing.
So yeah, I think the robot stuff is always fun to talk about and fun to imagine.
I'm excited about it.
I just, you know, I prefer if it was like a, I don't know, a weak robot, though.
I don't want that thing all of a sudden going crazy and trying to kill me.
Yeah, a lot of the designs I've seen, they're only like a meter and a half.
They're short.
But there's been a bunch of stories around AI companionship slash robot stuff.
You know, it seems like mental health is starting to become an economic category in AI.
We're seeing stories on regulators cracking down on AI companions,
you know, teens, you know, doing discussions and then, you know,
self-harm a bunch of stuff.
I mean, if we start to get these robots into our homes, you know,
is AI companionship not that far around the corner?
companionship not that far around the corner I saw this video of it was a
humanoid or rather it was it was a robot but the face was a humanoid and
kisses another human yeah I wonder how long till we reach a point in history
where we look around and can't particularly tell the difference between a robot
and a human. And I wonder, you know, I can't speak on the mental health stuff, to be honest with you.
I didn't, I'm not too aware of whether AI has been a net positive or a net negative for mental health
on a broad scale. But I have a couple of friends talk to, and they use it for mental health stuff,
and I've heard mostly positive things.
But, yeah, just kind of pivoting back, I wonder when we get to a point,
we will eventually get to a point, I think,
where the AI robot is a better therapist than your psych PhD who went to school for over 10 years.
And then I do also wonder if, or wonder when, rather, AI companions also become AI lovers for people? Because, I mean, if you can just code the perfect partner into a robot
and have it be indistinguishable from a real human, you know, does that solve the loneliness
epidemic, if you will? Or is it going to increase it? I don't know. I mean, I just, it's, I go down these rabbit holes a lot.
Ryan, what do you, what do you think?
Yeah.
Oh man.
Well, one, it does seem like Elon is trying to push towards this very quickly.
I wondered this, like if it's just for personal reasons where he's just like tired of trying
to like have a relationship with a real human.
So he's just like, nah, I'll have kids with real humans and then i just need a ai girlfriend um that does seem to be the push um but interestingly
enough um and i've thought a lot about this just because you know i have three three kids and the
thought has always been like how do i make sure that the kids are turning out, you know, as good human beings, right?
Like, I want the kids to turn out to be adults that I want to be around and not like, you know, little a-holes that nobody likes.
And it's like this conundrum for parents.
It's like you want to turn out good kids.
So when it comes to relationships,
especially with like AI and that type of thing,
if you just have an agreeable AI
that doesn't ever challenge you
or doesn't ever like put you through the paces
of a relationship or drama
or emotional versus
logical thought processes and communication, then you kind of stagnate as a human being on knowing
how to communicate and to connect with people around you. And this is what I realized with my
kids is, you know, the greatness is bred from adversity. And if you just always give the kid
everything they want, and you always just
kind of coddle them and nurture them, and they never actually have to strive or grow or build,
you know, muscles of their own, then they atrophy both mentally, emotionally, and physically.
And if we strive for this type of AI companionship, you know know whether it be sexual emotional physical whatever um we're gonna see a
massive epidemic of people atrophying in their ability to connect on a human level um which
does not breed greatness it does not breed you know good leaders it does not breed you know sound judgment it often breeds the exact opposite do we need do we need greatness of
AI robots can be great for us I mean you know eventually someone eventually someone's gonna
unplug it you know so yeah yeah but then it's like WALL-E. Remember the movie WALL-E with all those people floating around on the couches?
I mean, are we going to end up like WALL-E?
Like, you know, I think there's some potential, you know,
and Noah, you mentioned, you know, robots starting to look like us.
I mean, if you want to see where it's going, go to Clone Robotics.
They've got
this thing called a protoclon which uses myofiber artificial measure tech muscle tech and has over
a thousand synthetic muscles joints ligaments and so on that truly mimic the human structure
so they're getting super close to it um engineers are building muscles um mit is doing work on it so yeah if you want i don't
think we're going to have to wait long to have and i saw some amazing demos of just the head
where it looked like a real human head and it looked like it was experiencing emotions so
i think we're really close to that sort of stuff um but i think ryan's point you know if it's i mean it's
this quandary you know if if you don't have challenge and some struggle and difficulties
in your life then you know how are you going to grow and then you know noah's saying well you know
if it's if the robot's going to do anything do we really need that i mean it's you know we started to get very philosophical here
go ahead well I mean for example yeah humans humans still have to create other
little humans to propagate right and artificial womb systems I mean in
hospital Philadelphias is developing a system for
extremely premature babies. That's the beginning of the artificial uterus slash womb. So,
we may not be too far away from that. Sorry.
Okay. Well, in the near term, in the near term, you know, there is an aspect of, you know,
attraction and, you know, interpersonal relationships to be had to create little humans.
But I want to propose something else and bring up another topic here,
because I don't think I've ever really heard anyone talk about it to this point,
to this point is the effect of science fiction and fantasy on the propagation of our technology race.
We've been in the last hundred years writing sci-fi that a lot of engineers grow up on,
and then they spend
the majority of their career trying to recreate, you know, if, you know, if WALL-E,
you know, even though it was a kid's cartoon was never created, we would never
point at WALL-E thinking, oh, that's an option for the future.
But now that we have that in our heads, you know, that there's going to be
engineers that want to go out and create that starship, you know, know the life of leisure because you know why not um so it's a
fascinating thing that a lot of this stuff does emanate from uh you know humans imagination to
write stories and to create uh you know whether it be i robot and the positronic brain, you know, whether it be iRobot and the positronic brain to, you know, WALL-E.
It's, we're kind of writing our own future here
and engineers are just kind of lagging by like 20 or 30 years.
Well, yeah, but also maybe it's a way of framing possible futures
so that we can start to think about potential ethical and, you know, other issues around.
I mean, Black Mirror is a great show for that.
And, you know, Gattaca for genetic engineering, you know,
Terminator is the classic for, you know,
why you don't really maybe want to put a robot in charge of, you know,
your weapon systems.
So I think it's an interesting robot in charge of uh you know your weapon systems so i think it's
an interesting thing you brought up you know there's there's you know fantasy and sci-fi as a
genre uh is is that a seed for future technologies i think absolutely it is um but i think it's also
a way for to introduce to non-technical people the potential for
technology to go wrong or go right you know um go ahead captain levi oh um i i think that
bit's hand is off uh and it's been up for a while so uh are you are you wanting to chat
Are you wanting to chat?
Thank you, Levi.
No, I didn't want to comment on a few things here.
Go ahead.
Ryan, I love your commentary, and I love your healthy fear of AI and your children.
I also am a dad.
I've had the same thoughts, and I think we can all talk about the potential negative side effects a lot.
There is some danger,
in particular, for the family. As Ryan was saying, if everyone's got these echo chambers of lovers
that are really just code trying to make you continue to interact with it. But all that stuff
aside, what about the positives? So I'm a dad, and I grew up with a lot of science fiction and, you know, just older stories. And one of the things that I always thought was so cool back in the day was if you were like a royal or you were raised in a wealthy family, the way you were educated was you were given a tutor.
taught you at your pace at the correct level and taught as much as you possibly could on a one-on-one
basis. We can do that with a humanoid robot for all of our children. Soon, right? AI is going to
be capable of doing things that in the past were only limited to tiny segments of humanity. I think
that's really exciting. If we push this towards like a positive direction, this can be, I mean,
just such a powerful thing.
Like, can you imagine having a humanoid that can walk around and teach your kid how to, you know, build things as it sees that it's interested in it?
Like, oh, you like those logs?
You know, it's cool.
We could build a fire.
Let me show you how to do that.
And on a more bespoke level than we've ever seen before.
So I think there's a lot of positivity to this as well, guys.
Yeah. Yeah. I mean, I agree.
You know, it's been said that science fiction and fantasy
really act as a collective dream engine
for the rest of us, you know?
I mean, they create these radical futures, you know,
in a psychological, cultural,
it looks like it's economically viable.
I mean, look at Star Trek um you know i i think without
the this this genre technology the race for technology would be a lot slower and maybe a
lot less ambitious um in some ways the the dream engines of science fiction and fantasy is really
it's made it you know legitimized and and it's been inspiring so yeah i mean i'd love to
have i would have loved to have had you know a little part of friendly robot this you know where
i could ask a question like how do i do this and what why is that because i you know as parents
we're used to you know going oh my god a thousand questions from my kids i won't exactly like AI is going to be able to. Yeah. I mean, you know, and you see people get critical. It's like, oh, you know,
you're giving babies iPads and, and, you know, it's like, I still think there's a strong need
for parents to be parents with kids. And, you know, maybe it becomes a partnership with the
technology. I mean, that's the interesting thing.
We're starting to partner even more and more and more
with technology in raising children.
And, you know, there's tons of issues, I think, around that
that we're seeing both positive and negative.
So, Daphne, I totally agree, but I also think there's a need for
some consciousness around it go ahead as you just said it right there right is I
mean it's a this is like a tabula rasa is what it can be for a choice of
blank slate if we use it for positive things I mean think of how much you and
I seem to be about the same age right like how much time was spent playing
video games or watching TV shows. That could be used for positive, engaging education,
but we have to push the technology in that direction.
We have to not give the two-year-old an iPad
to watch YouTube because we're exhausted.
We have to actually do our best
to make sure these emerging technologies are used
to educate and grow as a society
and not just end up as like virtual
girlfriends where none of us leave our houses kind of thing yeah yeah I mean we we keep we've
mentioned this you know a couple of times week to week that the progress is really accelerating and
I think science fiction and fantasy it's it's a double-edged sword where it's
accelerating progress and it kind of amplifies the risks but at least it makes them obvious
at least it's bringing us to our attention um versus you know driving us to a cliff that we
can't see it we're going to run off um Captain Levi go ahead so um sorry to take us back to movies.
I think some movies that Ryan may have missed.
By the way, I really enjoyed Wally.
At least I know I'm that old.
Some of the movies that Ryan may have missed are, you know, things like Person of Interest.
And then, you know, talking about, you know, growing up with, you know, a personal assistant.
I think one of the spaces last month, we did talk about this.
I guess if it continues as it goes,
I guess we are going to be entering an era of geniuses
because people who know everything about a lot of things.
But then another movie that actually comes to mind in that regard is Megan.
We can see how things spiral really quickly out of control because even if it's just a learning assistant,
it was also programmed and could be programmed to do other things.
So that risk still remains there.
But I am heavily optimistic because I'm actually pro-technology.
So I'm heavily optimistic about the future
and where the wind is actually taking us. Yeah, yeah. I tend to be positive myself.
You know, we've talked about risk with, you know, robots and humanoid robots and
in the way technology is moving so fast that it's co-parenting, causing us to co-parent.
A bunch of stories popped up recently where scientists, I don't think it was the US,
but they used AI to create DNA animals.
And those animals lived and were existing.
And so we're seeing also AI, you know, doing break breakthroughs
in protein folding, drug discovery, you know, but it's starting to collide with a lot of biosafety
fears. And, you know, researchers are warning that AI could accelerate not only the cures,
which is great, but the classes of pathogens, which is not so great, know and i i my concern is is that bad actors are going to
get involved at this level and you know we're seeing just ai compressing evolutionary timelines
effectively and while it's allowing rapid discovery in medicine it's also just really
super enhancing the biosafety fears you know with ai enabled pathogens Ryan your hands up you want to comment
yeah it's um I think it's it's so hard to know the the fullness of where
everything's going just because there's so many unknowns right there's there's
so many unknowns on how different industries are going to change
and new technologies are going to be discovered. And the, the current things that we, you know,
deal with, whether it be, you know, not having enough water or electricity, or, you know, we sit
in traffic every day, right? Like if we, we, if we rewound the clock 500 years, it could have been, you know,
oh, my cart is stuck and I need to, you know, get this barrel patched.
And, you know, I, I need to find a, you know, a book for my kid
