Projects
![]()
No projects found
Investors
New
DAO Treasuries
DATs
Lend / Borrow APRs
Chains
Categories
Spaces
D A R K pt2
Recorded: May 6, 2025 Duration: 0:55:10Space Recording
Short Summary
Dark is gearing up for significant growth with new product launches and strategic partnerships, focusing on enhancing its brand and community engagement in the AI tooling space.
Full Transcription
Music All right, we are going to try this again.
I noticed people kept getting kicked out.
I have no idea why.
Hilariously, it's probably an infrastructure problem.
So if you want to know why infra is valuable, stuff like that.
When technology works smoothly, it's because some infra nerd is deep in San Francisco, keeping your shit alive 24-7, living on some
10-minute SLI on-call, and trying to explain to his wife why he gets woken up in the middle
of the night by pager-duty sounds that are scientifically engineered to keep him awake.
Let's see how we're doing.
We're going to hang out for a little bit.
We're going to try to get to 20, 25 people
before we kick off.
All right, so drop some comments in the chat.
Like, what did you guys hear?
Did people generally hear, like,
were people getting kicked out so frequently
that they heard effectively nothing?
Where are we at?
Should I just live stream on pump fun again?
That's what I was thinking of doing.
The pump guys keep pinging me to remind me that I can live stream on pump and I was going to actually multicast today.
But,
I don't know if that's the case
okay
it feels stable
is it feeling stable to you guys
alright
nobody get upset
if you get kicked out it's's not me. I promise. I want you here,
even the losers and the haters.
Although everyone in the chat looks nice so far. So, all right, cool. So let's do the following.
Let's do a concise review of dark product, of what we're thinking about, right?
Let's do a concise review of the landscape, and then let's go into product, and then let's open up for questions. If you have questions or
you want to chat about anything, please feel free to request being a speaker anytime and I'll keep
you in the queue while I'm talking. All right. So very quick review, right? What is dark building hosting infrastructure for AI tooling?
Why?
The number one way that AI is going to get smarter from where it is today is going to
be through access to more capabilities.
That's going to be tooling. It's going to be better evaluation, tooling evals, testing, verification.
There's a whole host of things that ultimately fall under tools that LLMs will need to access.
There is a huge hosting problem when it comes to all of these tools. Specifically, there are auth and monetization problems that I think have made it difficult to host these types of tools remotely. We're basically, we're building a new internet where agents are like a new class of internet users, right?
It's not just people anymore.
It's now also technology that's browsing the internet
and they need new ways of managing
interactions on the internet securely.
So we're building for that.
What have we been doing the last few weeks?
Where has my head been at?
I've been spending time across internet communities, both in Web 2 and Web3, to understand where AI tooling is going and what the right go-to-market is for Dark, right? Everything about being early in a startup and building
an early stage venture, right, at the beginning is about go to market. You can have attention,
you can have, you know, engineering resources, it's understanding how to take those things,
combine them and execute into a product that drives usage. And ultimately, like a sustainable user base that is important, right?
This AI tooling kind of landscape
has been in front of us for a couple months now.
If you've been following like model context protocol
servers and their wave,
there's a number of players across the space
that are pushing on building infrastructure for them, Cloudflare and Stripe,
Coinbase, right? Some smaller startups in the Solana ecosystem, for example, like Send, right?
A lot of people are thinking about this stuff. My view kind of going into working on Dark was that hosting infrastructure was the main kind of problem space.
That view I continue to hold.
My question around what to do about it has evolved quite a bit in the past month.
bit in the past month. Initially, if you guys remember, I was talking about how it continued
to be difficult to access remote servers. I can see we're definitely stable now because
our space is growing. So my view initially was that, okay, everyone's putting up GitHub repos of new tools for agents.
That helps agents get smarter, but it's really hard to use.
Maybe an MCP directory is valuable to ship.
Initially, I looked at this problem and I thought,
a directory on its own is super boring.
There are a ton of MCP directories online already.
Web2 has a lot more people working on it than Web3.
We have Web3 distribution because Dark is
a project aligned with than Web3. We have Web3 distribution, right? Because Dark is like, you know,
a project aligned with the Web3 community.
So from the perspective of like,
who knows our brand,
who our partners are going to be, right?
It's probably going to start on Web3.
And Web2 is like much bigger.
So, you know, if I'm just building like an MCP directory,
it might not be as,
we might not be as competitive on that front.
I've actually like kind of come full circle on this,
and I think that an MCP directory
for reliably hosted remote servers
is more interesting now
because all of these projects
that have been working on
these directories that host remote servers
have stagnated, have continued to be unreliable.
I think a lot of people are coming to terms
with the fact that hosting infrastructure is difficult,
difficult to ship in a stable and performing way,
which is pretty fascinating.
So I think it's one of the kind of go-to-markets that I'm weighing is like
a very, very simple MCP directory that just hosts, even if it's like literally a copycat
of some of the stuff that exists on the internet, you know, something easy that hosts popular MCP servers
like fire crawls, like sequential thinking, right?
Even stuff like Superbase's MCP server.
Those types of servers could do a really good job
of driving views to Dark's website,
give us a product foundation,
allow people to look at the website and be like,
okay, this is what Dark is actually doing, right? This is what I can go to the website for.
Another kind of interesting angle that I've been thinking about has been MCP bundles. So if you've
played with MCP servers at all or AI agents at all, you learn pretty quickly that agents have a pretty hard time picking the right MCP server, the right tool, right, to use for a given problem set once they have a lot of tools to pick from, right?
This is like very akin to like, one of my favorite things about ai is that um ai's
behavior is uh very human-like right we've we've definitely built machines like in humanity's image
uh and so if you imagine you know you sat down at a table somebody gave you a problem and you know
gave you i don't know 50 tools or 100 tools and said, okay, solve the problem, pick the right tool
for the job, you'd probably be pretty overwhelmed picking the right tool out of this mess of tools
that you have in front of you, right? Agents have exhibited the same exact performance.
One thing that's worked is routing agent decisions through different levels of
hierarchy, right? So allowing an agent to say, okay, I have a given problem. What type of problem
is it? Is it a coding problem? Is it a content problem? Is it a creative problem? Is it something
else? If it's a coding problem, okay, here are my coding tools, right? Now that I know that it's a coding problem, now that I've made that decision, I can kind of de-scope to just picking across a few tools that I'm interested in, right?
And determining which tool out of that small set is the right tool for the problem.
These, I think, you know, the term that I'm working on is MCP bundles.
These, I think, you know, the term that I'm working on is MCP bundles.
And I think MCP bundles or like MCP aggregators are going to get really popular.
I think that's like potentially a really interesting way to like capture attention on the internet for developers that are building with MCPs. So those are kind of the two like product directions that I'm thinking about
right now in parallel, right, to product, which is important, but not the most important thing.
Really what I've been focusing on over the last few weeks is building Dark's brand, right?
And, you know, it's funny because if, you know, if you've been on the internet, if you come from
the trenches, right, you might, you know internet, if you come from the trenches, you
might have seen people saying, oh, Edgar's a little quieter on the internet.
What's happening?
Some just incredibly worried individuals have been like, oh, is that you're still working
on Dark?
It's been a whole two weeks.
Why isn't the whole company built?
And what that really highlights is information
silos, which I think is really fascinating. There are like three core communities that I'm working
with and really, you know, one set of individual archetype that I'm working with across those
communities to build Dark's brand. The three communities, right, I would say are like, you know, call it like crypto trenches, right? Then crypto
like builder ecosystem, and then web to builder ecosystem, right? Oh, I just got a really exciting
message I can't talk about on Telegram. Fuck. All right. That'll be a couple of weeks, but stay tuned. If that partnership
works out, it'll be good. Okay. So what I think is really fascinating about AI tooling, right,
is that it's relevant for non-crypto companies as much as it's relevant for crypto companies,
right? In order to build legitimacy,
right, with companies in general, crypto or non-crypto, right, the method of communication
and, you know, negotiation is pretty different than the method of building brand publicly.
And this, I think, is really fascinating, right? So, like, publicly, you know, if you want to grow in reputation with the trenches, you tweet a lot, right?
You jump on PumpFind, you do live streams, you do spaces, and all those things are relevant, right?
If you want corporate partnerships, right?
If you want to partner with major companies in the Solana space, right?
If you want to partner with major companies in tech,
right? If you want to do, look at what Cloudflare did, where they partnered with Anthropic,
where they partnered with Stripe, where they partnered with these massive behemoths, right?
Those conversations are done privately, right? And that's just because of like human nature. I think,
you know, you guys have probably, you know, whatever you do for work, wherever you are,
you've probably at some point negotiated some type of partnership somewhere um and you probably recognize that like
those conversations are very human to human right they're very relationship driven and um you know
getting to the point of having something tangible to work on together uh between organizations is
something that takes time, something that
takes a couple iterations of conversations to really nail down, like, okay, we're excited
about each other.
What are we going to work on together?
What are we going to ship together?
How are we going to announce it together, right?
And so a lot of the last few weeks for me has been understanding, okay, like what corporations across tech, whether they're
in crypto or not, makes sense to partner with for Dark, right? And, you know, I don't have anything
to announce today, right? Because these conversations are ongoing and, you know, they'll be,
you know, partnerships will be announced publicly the same way that they are announced like, you know, there'll be, you know, partnerships will be announced publicly the same way that they are announced, like, you know, the other organization.
When, you know, the Solana ecosystem
on the kind of larger company side, right? So if you think about some of like the bigger companies
and even like, you know, Solana Foundation, Solana Labs, right? Like some of these like
major companies that have been revolving around Solana. I think the energy there for agents, for, you know, agents on chain or interacting with the Solana blockchain, I think
is really, really great. And I also think that the energy outside of crypto is, I mean, all time high
with respect to agents and, you know, the types of services that Dark will offer. Again, I've mentioned this
earlier, but my number one thing that I point to there is if you look at the conference that Stripe
is running this week in San Francisco, the work that they've been doing with Cloudflare to enable agent payments.
If you look at the, I'm super excited about this.
If you look at the keyless payment infrastructure that Coinbase just announced,
right, that allows agents to right now just pay on base,
but, you know, hopefully in the future, like on Solana.
I think those are really interesting opportunities. And so building that brand has really been, I
think, my number one focus. Because at the end of the day, the thing about a startup is that
everybody continues to ask if it's a real thing, if it's going to work, if it's going to live.
continues to ask if it's like a real thing, if it's going to work, if it's going to live.
Startups are default dead always, right? And so you have to keep pushing the startup boulder up
the hill repeatedly to convince people that this is something that's going to continue. This is
something that's going to ship product. This is something that's going to grow product after it's shipped, right? And that's just like a perpetual kind of effort as a founder, as an operator.
And so I've been focused a lot on that. I think like what I want to do, right? Just kind of like
rounding out a product conversation is really find an angle that
doesn't just make a splash, but integrates well into ongoing product efforts, right?
What I don't want to do is I don't want to like ship like an MCP marketplace that people go,
oh, cool, like it's here, right? And then like no one's using it no one cares um maybe people care for like a week or two and then
it kind of goes away right um and uh so what i'm looking for is kind of those like compounding
um compounding opportunities right and i think like you know hopefully some of these partnerships
that we're talking uh through uh work out over the next couple days to weeks and we can start
kind of showing more publicly,
like the teams that we're working with. And then it's a matter of, you know, working with those
teams to develop like agentic solutions that make a difference for people, right? And eventually,
like getting to consumers who can use smarter AI that leverages the tooling that Dark is building and, you know, kind of
complete that, like, that, that, that impact cycle. So people can really feel like the value of the
software that we built. Right. So yeah, that's kind of, you know, my roundup, I'll kind of pause
there, you know, give people a chance to jump up, ask questions,
give me their thoughts. But a lot of brand building, partnership building, in-person
conversations, a ton of travel over the past couple of weeks. You know, I mentioned this on
the first space. I'm very, very excited to be done traveling at least for a few weeks. I'm planted in
Salt Lake City at the Mountain Dow office, if you guys are familiar with Mountain Dow.
And, you know, my plan is just to wake up, sit down in front of the computer, go to the gym, walk my dog and repeat.
Right. And so I think this is like a really good time to take all of these learnings,
all of these conversations that, you know, we've had over the past couple of weeks as
we've traveled, synthesize them and turn them into tangible software products and tangible
partnerships that, you know, we can announce publicly. So I'll pause there. Really stoked
that you guys, you know, managed to stick in the space and we're getting kicked out all the time.
I'll give people a chance to jump up and request speaker spots.
All right. Some of the, some of the questions in the conference.
Okay. Let's tackle some of the questions in the comments because I love this stuff.
I was thinking building a startup in public like this,
I think it's interesting because you get the benefits of distribution, right? Like there's
more people paying attention early on. But the cost that you pay is that you spend time
that answering more publicly, more transparently
would effectively destroy those opportunities.
And so someone asked in the comments,
like, hey, how do we attain escape velocity
with dark in the short term?
Price go up equals market making,
equals revenue, equals more dollars for tech revenue equals more dollars for tech equals vitality
for token and vitality for like dark research, et cetera.
Tell us to buy often who's going to start stepping in.
And so like, and the example was like a Solana Foundation going to start stepping in, right?
Let me be really honest with you guys.
If Solana Foundation and I, and we were having conversations, right?
And I just like went on a Twitter space and I was like, yo, like,
that would kill the deal immediately, right?
That would kill any kind of conversation immediately, right?
Nobody wants to be blown up on the spot like that.
And so like the reality with those types of conversations
is that it needs to be presented correctly, right? Now, you know, I'm here for presenting it as soon as the opportunity, like, arises, right? But I need to respect the partners that we're having conversations with, right?
I need to respect the partners that we're having conversations with, right?
If they're not ready to publicize what they're working on,
maybe if they haven't fully formed their thoughts around it
or they're working on tech or products that aren't completely finished, right?
These are all things that we need to consider and keep internally
so that we can foster healthy partnerships. Right.
In terms of like attaining escape velocity with dark in the short term,
I'm not going to comment on that directly,
but I will say that like price going up equals market making equals revenue
equals more dollars for tech equals more vitality for the token vitality for
dark products is not exactly how
it happens, right? I think the thing that drives revenue, right, which is like I'm looking at this
like four kind of category like sentence, right? The thing that drives revenue is product that is usable and payable for, right? Which we
don't have today, right? So, you know, I was like talking to friends a couple days ago and I,
you know, I said something funny. I said, look, like, you know, people, a lot of people will like
ship product with no Stripe payment button and then they have no revenue, right? Because if you
don't give people a way to pay you, they won't pay you. They they have no revenue, right? Because if you don't give people
a way to pay you, they won't pay you. They can't pay you, right? And this stuff sounds like really,
really obvious, but, and it is, right? It's like, it's like blaringly obvious. But when you look at
the way that people build products across the industry, they're making a lot of these like
really obvious mistakes. You have to give people a way to pay you, a way to pay for product if you're going to have revenue, right? Market making does not give a project revenue.
In fact, my honest take, and I've gotten so many DMs from all types of market makers in the space
in the space since Dark has launched, right?
since Dark has launched, right? It's not my first time dealing with market makers.
It's not my first time dealing with market makers.
I ran one, right, over the last couple of years.
I ran a trading firm that, you know, like,
has a common archetype to be a market maker.
My perspective on market makers in the crypto space
is that effectively none of them who do deals with
companies are super healthy. Now, it's not to say that they're not valuable to work with,
but I don't think we're at a point where spending a lot of time with market makers makes sense.
Why? Because it's like a super new token, right? It's a super new product, super new company.
What's much more important is to stabilize the product foundation, pick up users, pick up
partnerships, do real work, right? Instead of like focusing on financial engineering.
I think financial engineering has, you know,
and this is like people fall into this trap in crypto
where they will spend time on the financial engineering side
because they get so caught up in where the price of their token is
on a daily basis that everything else falls apart.
And then what happens is you effectively become this like shell corporation that is just facilitating
financial engineering efforts. And the people that always lose are the users in retail and the people
that always win are, you know, the hedge fund boys
that understand how the game works and know how to structure these deals and structure liquidity
in a way that works out for them, right? I really don't have an interest in that. And so, you know,
not spending my time on market making conversations. In terms of more dollars for tech,
market making conversations. In terms of more dollars for tech, definitely I think this is like
really interesting problem. The price going up does not provide more dollars for the tech, right?
As you guys know, I own a very small portion of the token. The token is structured to live alongside
the company in a way that I think is like, you know, humbly put, like has potential to be
really innovative, kind of front ran this Believe app, you know, kind of explosion of
tokens that launched alongside like software products.
And frankly, I think opens up conversation for a massive, you know, landscape of discussion around,
I just said it opens up a conversation for a discussion, phenomenal Edgar.
It opens up a discussion for like a bunch of topics within like, what does the token look like?
What is the token good for, right? How does the token exist in conjunction with stuff that's built, right?
And I think there's, I mean, there's a ton of space to innovate here.
And it's important because a bunch of the ways that tokens have existed in the past have failed, right?
Like, if you guys look at crypto, all of these down-only tokens,
no wonder everyone is super stressed about getting out as soon as they possibly can right because everyone
in crypto has learned that if you invest in something you know you're playing this you know
uh fucking musical chairs game right where you need to get out before the chair gets pulled from
from out underneath you right uh and i just don't think that that's where the industry is going to end.
I think there's more that we can do there.
But it, you know, it creates new questions, right?
How does funding for the product development happen?
How does that exist?
How is founder ownership managed, right?
Where are growth incentives, right? Where are growth incentives, right? You know, I think, I think there's,
you know, I have a ton of thoughts on that and definitely happy to like talk more about it,
but I mean, there's like a ton of open questions there. So, you know, just going through this,
like, this is a long process of going through this, like price goes up equals market making
equals revenue equals more money for the tech kind of comment, right? Which I think is like,
to be clear, right? Like I'm not bagging on the comment itself. I think the comment itself is
something that like a lot of people are thinking about, a lot of people are talking about in
crypto. They're trying to figure out like, okay, like what makes a good outcome for me, right?
As a builder, as a founder, as a holder, right? And, you know, I think, I think, you know, kind of these points
that there's this illusion of, like, combined forces on the financial engineering side,
which are actually just destructive forces. I think, I think that illusion, hopefully,
you know, we can be a small part of breaking down because I'd much prefer to build something that illusion, hopefully, you know, we can be a small part of breaking down because
I'd much prefer to build something that's, that's, you know, healthier for the ecosystem
and all, all, all of the parties involved than something that like ultimately, you know,
creates more, more destruction.
So I'll, I'll pause there.
I'll read through the comments.
Someone's like, you dropping a new CA?
No, dude, I, we just did that.
Okay. um someone's like you dropping a new ca no dude i we just did that um okay who's got a question who's itching
i did tell someone i jump on a call with them at three but they're also in this space. So I think they'll forgive me.
What else do you guys want to hear about?
Drop me some comments, jump on, ask some questions.
We can just sit here and vibe to the Twitter music if that's what you want to do. Everyone's like, we just want a little bit of meditation.
From a timeline perspective, I think expect to see some product launching later this week.
My hunch, and I'm figuring this out literally today because I need to figure out how to
package some things, my hunch is that we will at least have a foundation for remote MCP servers.
Ultimately, I think that shipping something
that makes it easy for people to use MCP servers
is still valuable.
I'm also looking at MCP bundles.
Those are my two kind of go-to-markets this week.
In parallel, hoping I can announce
some partnerships this week, but it really depends on kind of
some other parties that I'm talking to
and I don't wanna push them too much.
All right, Breed AI, I'm gonna bring you up here.
Yo, Breed, can you hear me?
Yeah, Edgar, thanks for your time.
My question is, who is your target audience?
And is the target audience the token holder?
Are they the ideal user?
Or are these two separate camps?
Yeah, that's a great question.
So our target audience is developers.
I think the second part of your question,
I think I'd largely answer that token holders
and our target users are two separate camps, right?
Not to say that developers like, developers are, you know, can't be holders.
And I think what's interesting is we've spoken to more developers
that are excited about what we're building.
They've naturally, like, become token holders, right?
Because, again, like, people want to,
tokens are a great way for people to be involved
with something that they're excited about, right?
And they could be excited about that for different reasons.
In terms of like who we're building for, we're really building a software product that will be used by developers who are building AI-enabled products, right?
building AI-enabled products, right? This could be people who are building agents,
you know, like we've seen with like Eliza. This could be people who are building enterprise
agents. I think this is like a really fascinating space that crypto doesn't talk about. But, you
know, I have a few friends who are working at companies, building agents for logistics,
for construction, right?
Not construction, but warehouses,
like those types of things, robotics, right?
There's interesting operational plays there.
This could also be,
I think this is another interesting emerging
kind of vertical applications that
are not agentic, but leverage LLMs and AI tooling.
So, for example, there is someone in the ecosystem building a new type of wallet, right? That is AI powered and will be leveraging, you know, MCP tooling like in the background,
right?
But ultimately, like who we're selling to directly is developers.
Sounds good.
Thanks.
Cool. Sounds good. Thanks. All right. Who else do we got? I got time for maybe one more question, and then I'll probably jump off.
Okay, I'll ask myself a question.
So something that I've been like, you know, kind of reflecting on.
I get this question a lot.
AI agents like the first time around in crypto got like super hyped up and then blew up
because people like ultimately recognize that,
you know, they couldn't really do much.
What's going to be different this time, right?
The thing I like about building AI tooling
is that you can make a bet on kind of this broader thesis,
orient a project around a broader thesis,
and then figure out the specifics later.
And this is something that, like, I think when I was, you know,
first starting companies, I would spend a lot of time thinking about,
like, the specific idea that I was excited about, right?
Oh, I just need to sit here and
figure out what am I going to build exactly and how is it going to work exactly. I can't get
started until I figure all those pieces out. Over time, and for those that are less familiar,
I've been an entrepreneur for a little over five years now. I've had a number of companies across
tech. I've had some acquisitions that I'm happy with. I've learned a ton of lessons along the
way as well. One of the things that I've learned is that the way to figure out the answer to
what to build and how to build is to get started without having all the answers.
And this is something if you guys have been following Double Zero, Austin Federa, one
of the co-founders over there, was talking about this the other day.
He was highlighting his view on part of why investors were so excited about double zero is because he would get on calls and highlight all of the unsolved questions that were in front of them.
And I think there's a ton of value to that approach.
The beautiful thing about having a broad perspective that's strongly held is you can say, okay, our view is that AI tooling is going to enhance AI capabilities, AI intelligence.
That's very, very broad.
The specifics don't need to be figured out in order for you to say, okay, like, I think there will be a more tangible opportunity for progress here. If you look at AI Agents 1.0, right, the reason I think
they fell flat is because ultimately they couldn't do a lot. And they couldn't do a lot because they
didn't have tools to use, right? Or the tools that they did have that they tried to build for them
were ultimately like very painful to use.
And so at a high level,
I'm really excited about this tooling kind of chapter for AI.
I also think it nicely aligns with,
I think we've seen LLMs
from an intelligence perspective
have started to plateau.
LLMs are still getting smarter, Gemini's getting smarter, LLAMA's getting smarter,
OpenAI shipping some interesting reasoning products,
but it's nothing compared to the differences
between GPT-3 and GPT-4, right?
We're just starting to see that, you know, throwing more compute at the current transformer architecture is getting us less.
The more interesting opportunities for improvement are around that core intelligence.
AI tooling is one of them.
one of them. Performant inference is another one, the cheaper inference, right? So, you know, I'm
Performant inference is another one, the cheaper inference.
excited about this, like, broad kind of direction of AI tooling. What I think is important in order
to see it through, right, is to build up a supply of AI tools. So what I expect Dark to be doing for the next couple months, right,
is to be building up a supply of tooling and to kind of be showcasing it, right?
Hey, here's interesting stuff you can work with as a developer, right?
And then, you know, the other side of that, right,
to get usage to solve this cold start problem, okay, right?
We've built some cool tooling. No one's using it. How do we get people to to solve this cold start problem okay right we've we've built some cool
tooling no one's using it how do we get people to use it well in order for people to use it like
there need it needs to exist right so we'll build the tools first right and then the second side of
that and this is something that we've started investing in already um will be incentivizing
developers to leverage our tools right and that And that will mean like, you know,
distribution and awareness to developers, that'll mean investing in hackathons, you know,
launching bounties, launching events for developers, right? Those types of things,
right? And that's something that like, you know, if you guys are familiar with my background with
MountainDAO, right? MountainDAO is a flourishing
developer community in Solana I spent a lot of time with everyone across Solana that you know
has you know spent time organizing hackathons spent time incentivizing developers like I know
all those people I know that pattern I think it's something that like we as an organization at Dark
have you know I won't say in the bag but like we as an organization at Dark have, you know,
I won't say in the bag, but like we definitely have the expertise to execute on it. And so I
think it's going to go like that, right? Like the starting point is, okay, why is it different this
time? Why is tooling going to meaningfully impact agents? Well, it's because agents faltered the first time due to like a lack of capability,
a lack of additional tooling. And then how will we do it? We'll need to build up tooling supply
first and then incentivize developers to build with that tooling supply to complete, you know,
the cold start problem. All right, Neuron, I'll let you up and then I think that'll be all for the space.
Give me a second.
All right, can you hear me, Niran?
Yeah, hello. First, thank you very much for all your transparency in the last weeks.
It's very interesting to follow your ideas and to follow how you build up that.
And I think that finally, I hope people will understand that Dark is now much more than a game.
Initially, it was more view at the game. You can play, you can look on the planet, you can look for the basic people.
It was just a new game. We will play, we will win, but I think now what you explain is really
much more than that. You have a very, very good idea, I think. So I hope you will develop that
and you will continue that and I think
Dark will be very huge in the future so thank you for that. Yeah yeah of course
yeah I mean I you know I can touch on the game piece this is a great point and
I'm surprised we haven't talked about it this far on the space but the game, the space game that we, you know, started focusing on at Dark
initially was always supposed to be like a short-term marketing play, right?
The idea was, all right, we've, you know, captured virality, right?
Dark kind of popped on the internet.
How can we capture usage of anything dark, right? And grow
the brand in a way that is aligned with the longer term vision, but really focus on the short term,
right? And what can we do in the span of, you know, the initial goal was something like 96 hours,
what can we build in 96 hours to get people using something dark? And I want to be super clear about this. I think we miserably failed,
right? And we miserably failed because it took much longer than 96 hours to ship this game,
right? Three weeks in, we were still working on it. It's technically still not live, right?
You're still working on it. It's technically still not live. Right. And like what that did is it it it turned that effort into an unintended
conversation right which was like oh like is dark building games now right well it's like well we've
been working on games for three weeks and that's all we've been talking about so it makes sense
that people ask that question right um look do i do i wish that we shipped it faster? Absolutely. Right. Like, no question. I generally like in the kind of person that has a pretty high bar for myself. I don't meet that bar very frequently. Right. Like, I kind of try to keep my expectations for myself pretty high.
I also think that like startups in order to succeed, generally need to have extremely rapid execution, right?
Like it's less about like my personal bar for myself or like what I'm happy with, blah, blah, blah.
It's more about like what does it take to win, objectively speaking, right?
If I look back on why we didn't ship the game in the 96 hours that we had time boxed, right? I really think it comes down to
a scope problem. And, you know, there's a ton of excuses that I could make, right?
I could talk about how, like, you know, we started working with certain info that we thought would be
faster and it turned out to be slower. And, you know, then we built from scratch and we
said, hey, let's build this, this ecosystem. And, you know, this, this space universe. And
initially the idea was to have like a single player mode, but then we found out that people
were a lot more excited about multiplayer mode. Like there's all these kinds of folds that you
can dig into. Right. But at the end of the day, the reality is super simple. Right.
We had a plan for a short term marketing play that would onboard users to dark products.
And we blew that timeline. Right. Now, you know, other things that are that are relevant here, just as I'm thinking about it, right?
The payment button, right, that I brought up early in the space, right?
My thesis that, like, people can't pay you if you don't give them a button to pay you, right?
Who knows how much revenue Dart could have picked up early on if we had shipped something in 96 hours with a payment
button, right? That's going to be an open question forever because we never shipped the payment
button, right? So like the top line was zero. Had we had a payment button, we could have picked up
revenue. We could have picked up users. Maybe we wouldn't have, right? Maybe it would have flopped
and people didn't want to use it. People didn't want to play it. But my point is that you literally can't answer these questions unless you execute on some of these key procedural
pieces. How does that change how we're thinking about things in the future? It was a very stark
reminder for me. I've been building software for 15 years. I've been building companies for a third of that time officially and my whole life
unofficially. And yet, like some of these basic lessons, right, keep coming back. You kind of,
you know, there are things that happen that kind of refresh your, you know, remind you that these
things are important to focus on. The game, building the game, missing the timeline
that we had set for ourselves was a very stark reminder
for me, right?
That it's super, super important to focus on minimal scope,
on maximum velocity to cutting releases of products
that are usable and to adding some kind of,
whether it's payment or some kind of clear usage,
key performance indicator, KPI, some kind of KPI to that product to get an understanding of like,
okay, this was a success, right? So I hope that that's what you guys see from the team this week.
We're definitely very excited to take these learnings from the game,
to take these fresh learnings and funnel them into like shipping product, cutting releases,
moving with velocity as this week goes on. Some things that I'll say that, you know,
I think were really exciting from kind of the game like journey, right? We put 5,000 people on the wait list, right?
I think it's pretty cool.
Dark has done effectively no marketing, right?
Other than launching the token
or, you know, building alongside this token
that had been launched.
From a distribution, from like a medium perspective,
Dark has advertised entirely on Twitter, right?
With no paid advertising, right?
It's been purely organic.
So, you know, 5,000 people on the wait list,
that's pretty good.
What's even more exciting is we gave people a way
to sign up for the game with their Twitter, right?
And that would link their Twitter to like a user account on dark, right?
That gave us a little bit more of an opportunity to say like, okay, these are probably, you know,
real users, right? You can't just like throw websites in there or throw emails into like a
wait list and, you know, kind of get away with maybe faking numbers or, you know, whatever, if you thought you were going to get something for it and decided it's Sybil. We have about 700 spaceships in our space game, right? So
700 signups. It's pretty decent. It's a pretty good conversion, right? 5,700, especially for
the complete like lack of like marketing effort that I think we put towards game signups once they went live because of our decision to pivot more publicly back to the AI infra.
And that tells me that Dark continues to have a pretty strong brand that carries.
pretty strong brand that carries, right? People are paying attention. Half of those tweet,
People are paying attention.
half of those signups, half of the 700 signups that we saw, maybe like a little less than half,
maybe like 35% or so happened after any tweeting about the game purely off of the pinned
any tweeting about the game purely off of the pinned post on the Dark Research X, right?
Which means that people are going to the Dark X profile, they're clicking on pinned posts,
and they're converting, right? That's huge. Now, the next step, right, is to ship product that,
is to ship product that actually does things.
That's absolutely key.
And convert our target users, developers,
to using and paying for those products.
But I am pretty excited about the foundation that we have.
I have started from significantly worse positions in the past.
So feeling pretty grateful.
It's quite clear.
It's better to take time to build something.
If you launch the game in the 96 hours
and after you have a lot of bugs, a lot of problems,
it will be the worst situation.
So better to take the time to make the good things, I think.
Yeah, I'll say there's two sides to this.
I both agree and disagree with you.
I agree with you that I think what's really important to figure out
when you're building product is which features
and conversely, like which bugs really matter to your users
And conversely, which bugs really matter to your users
and which ones don't.
and which ones don't.
And this is like, if you guys follow Nikita Beer,
when he talks about building consumer products,
he loves capturing attention on the internet.
I mean, that's what he's one of the best people
in the world at.
And he'll write, you can't fuck up your first time, right?
You launch something and you fuck it up,
like no one will ever use your product again, right?
One, that's just kind of not true as a whole, in my opinion.
But two, I think what's really important to dissect there
is that doing well or fucking up
is relevant on core features that your users care about.
And what I see developers do a lot that delays shipping velocity, right, is they focus on features that don't matter.
And they go, oh, we can't fuck these up, right?
Because if we blow these and we have bugs, here, like, people are never going to use us.
They're going to hate us, right? Because if we blow these and we have bugs here, like people are never going to use us, they're going to hate us, right? And the answer is there's a whole host of auxiliary features in
any software experience that can have a lot of bugs. And most startups have a ton of bugs,
and it's appropriate for them to have a ton of bugs. Like that's the path towards winning,
right? But it's those few key features for your target users
that need to feel flawless in order for you to win,
and you can't mess those up, right?
And so on one hand, it's important to take time
and architect your approach the right way.
On the other hand, I think you can overdo that as well.
So, you know, like many things in this life, it's a balance.
Yeah, that's true.
It's true.
You're right also.
After you have to,
don't take many months to make it
because people will maybe think
that it will never happen.
So of course it's kind of balance,
but I think you're doing great for the moment.
So continue like this.
Yeah.
I appreciate it. All right, moment. So continue like this. Yeah.
Appreciate it.
All right, guys.
I got to jump.
Look out for more updates throughout the week.
I mentioned this last time and then I kind of just got derailed with travel. But I think I'm going to do one of these at least once a week.
It's Tuesday.
Maybe we'll talk in the community on X about the best time of the week to do one of these spaces.
But I think like kind of like a general product talk space, I think it's pretty fun.
Helps me kind of collect some of my thoughts, you know, synthesize them for you guys.
So let's let's figure out a way to keep these conversations going and I'll see you guys online.
All right, fam.
Let's put the music back on.
We'll have some exit music.
Check this out.
Why does it not feel like it's doing it?
Oh, it's just the starting thing?
Okay. All right. i'll see you guys later
