Paul Ford: Hey, Rich.

Rich Ziade: Hey, Paul.

Paul: The economy’s crashing!

Rich: No, it’s not.

Paul: It is for real! It turns out that China can do AI just as well as the U.S. and our whole magical productivity future just went up in a wisp of smoke.

Rich: That doesn’t sound like bad news, necessarily. Why is the economy crashing?

Paul: Well, let’s talk about it in a minute. But for now, the official Reqless podcast. For Reqless, the podcast about how AI is changing the world of software. Boy, is it.

[intro music]

Paul: Okay, let me tell you what happened.

Rich: What happened?

Paul: This is my favorite story in…ever. There’s a company called DeepSeek.

Rich: Okay.

Paul: And you can go to deepseek.com. It’s a Chinese company. It spun out of an algorithmic hedge fund. Like, just sort of, like, deep quant stuff.

Rich: Okay.

Paul: And the founder went, you know, we could use these GPUs for something a little different.

Rich: Okay. So they had some GPUs.

Paul: They had 10,000 H100s. The good stuff.

Rich: The modern ones.

Paul: Yeah. But they can’t—

Rich: Not the very latest.

Paul: No, but pretty good. And they can’t get more, because we’ve—

Rich: Right, there’s embargoes on it, yeah. This is a Chinese company.

Paul: Yes.

Rich: Okay.

Paul: And so they started doing the same thing that ChatGPT and OpenAI do, and building large language models, and a lot more in Chinese and so on and so forth.

Rich: Okay.

Paul: They’ve released a couple different versions, and everybody’s been kind of looking at them out of the corner of their eye.

Rich: Okay.

Paul: They’re not really building, like, products—like, let’s say Anthropic is, where it spits out code and it’s got a big chat interface.

Rich: Mmm hmm. Yeah.

Paul: They’ve got an API, they’ve got a chat interface, and they’re like, “Whatever you need to do from here: Cool.”

Rich: Yeah.

Paul: They’re also very cheap.

Rich: To use it?

Paul: Very cheap to use.

Rich: Like, a fraction of—

Paul: Pennies on the dollar.

Rich: Okay.

Paul: And kind of no API limits, like, just kind of all the things, when you do stuff with AI, it slows down. It doesn’t always work the way that, like, you expect computers to work. And in this case, it’s like, “Nah, go to, have a good time. It’s cheap.” So, you know, they, they kept incrementally releasing and they got to kind of their V3, which was pretty good. I started using it. I’ve been writing about it. I’ve been using it for weeks. Because I kind of like it because it shows the seams. It’s like, it’s almost, like, it feels like something you can tinker with?

Rich: Mmm hmm.

Paul: Whereas, like, ChatGPT and—

Rich: More shrink-wrapped, yeah.

Paul: It kind of feels like they do a magic trick and then you’re like, “Hey, I’m ready to get to work.” And they’re like, “Hooold onnn, just slowing down a little bit.” [laughter] You’re like, “No, no, no.” And so DeepSeek is just, like, “Yeah, I may not be the best, but I will not stop. I will just keep going until—”

Rich: Yeah.

Paul: And so I keep learning more about this new technology by using it than I do from the other ones.

Rich: Okay.

Paul: Okay? So I’m a big fan. And then they released what’s called R1, which is their reasoning model.

Rich: Okay.

Paul: And I love it, because actually unlike ChatGPT, which is just like, “I’m going to hide a lot of the details from you,” it actually narrates its little—

Rich: It talks through its thinking.

Paul: And so what’s going on here, what you start to see is that when we say “reasoning,” what it means is that you give it a set of prompts and you run some software and somebody puts a question in, and it kind of expands the space of the question and generates enough input that it feeds back to itself and that it answers the question in kind of a “more educated” way.

Rich: Right.

Paul: But you can really see that clearly with DeepSeek. It’s much more transparent and it’s open. They put out a lot of papers. I’ve been reading interviews with the founder and he’s like, he’s like, “Look, this wasn’t the best business decision to do this. We’re an algorithmic trading fund and we decided to take our GPUs and spin out into AI.” But he’s like, “I feel that China needs to innovate. We can’t just be…”

Rich: Huh.

Paul: Yeah, he’s, he’s literally doing this from a point of view of like, “Hey, we’re always—we’re never 0 to 1, we’re 1 to 10.” That’s what he says in an interview.

Rich: That’s interesting.

Paul: Anyway, you know, “I just was like, if I get a bunch of smart people together and I let them publish, even if the money’s not there, we’re going to do something really interesting and impactful.”

Rich: Yeah.

Paul: Obviously he’s got a little money in the bank, right? To do this.

Rich: Sure. And I mean, I think this has geo-economic implications that probably, that pretty much seep into the political sphere.

Paul: I can’t imagine—

Rich: This is a big deal.

Paul: The Chinese government is not unhappy that this is happening.

Rich: No, no, no, it’s, it, it, it’s showing…I think it says a lot about America as much as it says about China. That’s what I want to talk about when it’s my turn to talk.

Paul: Oh, did I do too much talking?

Rich: Absolutely not.

Paul: I’m very excited.

Rich: Absolutely not.

Paul: I’m having a good time. This is, like, the most fun I’ve had in forever.

Rich: Okay, so draw the throughline into your retirement account. [laughter] What happened?

Paul: All right, so the reason we’re doing this podcast in a hurry is that this crashed the economy, and Nvidia lost $465 billion in value in, like, I don’t know, like a minute. I don’t know.

Rich: Right. It didn’t the crash—crash the economy is, like, recessions.

Paul: It’s a big economy. Everybody can relax.

Rich: Yeah.

Paul: So look, here’s, let me give you—

Rich: The mark—as we speak, and who knows, we might come roaring back for some other reason, but as we speak, the markets are down.

Paul: Well, let me, let me give you why, okay? Chinese company, not an American company. Commercial product, but it’s really cheap to use. Really cheap to make. They’re able to spin it up without buying their own nuclear reactors like Microsoft did.

Rich: Okay.

Paul: Okay? And it’s open and downloadable with open research. So—

Rich: So wait, it’s open source?

Paul: Yeah, it’s that weird kind of open source that the AI does—

Rich: Open source-y?

Paul: Yeah, like, you can kind of, you can get everything. They call it open weights. But you can’t always get all the code.

Rich: Got it.

Paul: And you can’t get all the source material that they feed in.

Rich: Okay.

Paul: But pretty transparent.

Rich: Okay.

Paul: And honestly, reproducible. Like, some people are like, “Well, maybe they have secret GPUs or like, I don’t really buy this.” But like, we’ll know, because if this, if the paper is accurately describing the processes that they use to achieve this result, other people will be able to reproduce it.

Rich: That has been the case so far, which is, like, “Hey, you think we cheated? We didn’t. Here’s how we did it.”

Paul: Yeah.

Rich: And they’re actually putting out, they’re being pretty transparent about stuff.

Paul: But let me give you a couple side effects of this new thing, which isn’t new. Everybody has known these technologies were getting cheaper and more commoditized, but it just kind of slapped America across the face.

Rich: Mmm hmm.

Paul: Also simultaneous with, like, Sam Altman going to the White House and hugging up to Trump.

Rich: $500 billion initiative called Stargate.

Paul: Yeah, it’s just, like, as—we are just looking real smug and we got slapped, and it makes for a great story. But here’s what’s up. If this is real, you don’t need as many 3D GPU chips to create your large language models. And Nvidia has been the darling of the entire market.

Rich: They’ve held the keys.

Paul: That’s right. Then Microsoft has been investing in OpenAI and Google’s been building their own stuff. And so they thought they had a really big moat.

Rich: Yeah.

Paul: But maybe they don’t have as big of a moat, which again, like, Microsoft bought Three Mile Island. Maybe that wasn’t the best purchase. If it’s actually, you ten—

Rich: You could build out a really cool Microsoft-themed water park.

Paul: [laughing] That’s right.

Rich: On Three Mile Island, okay?

Paul: Just go in there, go swimming.

Rich: “Hey, do you want to go on the PowerPoint Lazy River?”

Paul: Mmm hmm. Exactly. [laughter] Just watch out because the water’s glowing?

Rich: Yeah, yeah, yeah.

Paul: But it’s okay. So, now if you offer AI services on the web, and tools, like OpenAI and Anthropic, you might need to compete on price, if DeepSeek is getting pretty good. Because—

Rich: [disbelieving] Might!?

Paul: [laughing] Because it’s, like, 1% as expensive.

Rich: [laughing] Oh, goodness gracious.

Paul: So there goes another big chunk, right?

Rich: Yeah.

Paul: And apparently, like, there’s a lot of people getting real anxious inside those orgs, because they’ve been making millions of dollars and feel really good about themselves.

Rich: Well, they haven’t been making millions. They’ve been making millions of dollars. But what they’re not, they’re not looking at their revenue right now. They’re looking at their speculative value.

Paul: No, no, no. The AI researchers who they’ve hired, these plum brains that they pulled out of big colleges.

Rich: Oh, yes, yes, yes.

Paul: So they’re all fighting with each other. We have a lot of sanctions and restrictions on China about who can, what they can do with the chips, and maybe that’s not as important. So that’s actually bad from a government, Trump administration point of view. And then the last bit is places like Facebook have been paying people lots of money to build open models and put them out, and they’re spending enormous amounts of resources to create their own LLMs and then kind of drive a wedge. And it turns out they’ve been overpaying, too. They, they can’t even give it away free correctly, apparently. [laughter] So, so this is—look, I know I’m talking and talking here, but this is like Christmas for me.

Rich: Explain that, before I say something.

Paul: Because… Tech to me is always about having more options, more opportunities, and they get pushed down all the way through society. And that’s fun. That’s why I always, always, I’m always, I just don’t love crypto because crypto is about taking all the wonderful things a computer can do and making sure people can have less of it.

Rich: [laughing] Yeah.

Paul: Right? I am a maximalist. I think everyone should have access to everything the computer can do as much as possible.

Rich: Okay.

Paul: It’s empowering and it’s a tool and it helps you think better and you should be able to do that. And I see a lot of that stuff finally really happening with this technology.

Rich: Mmm hmm.

Paul: With LLMs. And I know that people have a lot of concerns about them, but they’re going to be cheap, they’re going to be everywhere, and they’re going to help you with your geometry homework. And I think that’s good.

Rich: I think so, too.

Paul: Especially when it can run for a dollar on your phone. I think, like, a lot of the fears and concerns we have can get put away, especially if there’s good regulatory frameworks, and we have a new super tool that’s actually just a weird database that can let us do really great computer things that we couldn’t do before.

Rich: Yeah.

Paul: And it also has real limits. Instead, we’ve had all these, like, dudes telling us that like a star-baby is about to be born, if we only give them a trillion dollars. We gotta give them—Sam Altman asked for a trillion dollars and he said, “Then I’ll get you a really artificially intelligent thing.”

Rich: Yeah.

Paul: And everybody’s like, you know, “Three to five years! Three to five years!”

Rich: Mmm hmm.

Paul: And these knuckleheads are just, like, beating on it with a stick until they got it.

Rich: Yeah.

Paul: And it makes it more available to everyone.

Rich: When you say “knuckleheads”…?

Paul: I’m actually—these geniuses in China at DeepSeek.

Rich: Yeah.

Paul: And so there’s a glut. Everyone is going to get to have the software they want. That’s been my dream my whole life.

Rich: Yeah. And I think that’s gonna happen. I think that probably would have happened anyway, frankly. Like, I mean, you can, we can criticize Facebook, but they did, they actually were gonna make it more accessible.

Paul: But how much fun is it to see just a little bit of acceleration happen in a way that pisses everybody else off?

Rich: I think it’s great.

Paul: Yeah.

Rich: I think it’s great. And here’s my take on it. And it’s a little different. Look, I think lightning struck, what was it, four years ago?

Paul: Yeah.

Rich: It was, like, this all—you can trace it all back to a handful of research papers that sort of open the floodgates here.

Paul: Yeah.

Rich: Right? And then massive amounts of resources went to pretty much four companies.

Paul: Yeah.

Rich: Nvidia, which is, has a stranglehold on the hardware needed to create these LLMs that power AI.

Paul: And previous to that had sort of a stranglehold on a lot of the Bitcoin market.

Rich: Correct. OpenAI, makers of ChatGPT. Anthropic, which is, I don’t even know its valuation, it’s in the, probably a trillion dollars or something crazy. Microsoft got in the mix. Obviously the big players are aligning themselves. Facebook’s investing a ton. Here’s the, here’s the beauty of this story. The beauty of this story is this: What people forget is that constraints are a gift.

Paul: Yeah, that’s real.

Rich: They end up being a challenge to our ingenuity and our ability to think creatively. And when you put constraints in front of people, sometimes they can’t overcome them. But what they do is they force you to think a certain way to solve something. And the truth is the DeepSeek team wasn’t thinking, “Let’s apply constraints.” They had constraints. They can’t get better hardware. They didn’t have unlimited funds.

Paul: Well, and there’s an interesting corollary to this. This is very real. You have to have constraints. But also this leader made sure that everybody could get access to resources. Like, so yes, they didn’t have as many resources as everybody else, but it was a very, like, Xerox PARC-style environment where, “Okay, we’re here but if you have an idea, you’re going to be able to chase it and you’re going to get the tools that you need to the best of our ability to see what you can do with your idea.” And a lot of, you know, like 23-year-old, you know, PhD-student types—

Rich: Yeah, yeah.

Paul: —are getting to play with really good toys.

Rich: Yeah, but they’re not really great, good toys. They’re, they’re okay toys.

Paul: Well they’re not as, they’re, they’re amazing, wonderful super tools. But you don’t have unlimited resources.

Rich: The problem with unlimited resources is it actually becomes a handicap, because without constraints, you’re not trying to solve—you’re not challenged. You’re not trying to solve new problems. Look, there are some people who are, like, brilliant and they’re always challenging themselves. But when you have unlimited constraints, you stop innovating, actually. There’s a few examples that come to mind. First off, this whole layer of AI startups that’s sitting on top of these LLMs, right? A lot of them are just shrink-wrapping—you could pretty much do with the, with the core products underneath. And I say that not to shit on them, but I say that because it’s easy to do that.

Paul: Mmm hmm.

Rich: That’s not hard to do. You could put an interface around a thing and call it a “medical AI engine” or whatever.

Paul: Well, the mode is just, like, a couple relatively smart product people put in…

Rich: Yeah, exactly. And so I think what is so good about this story is that we can look at this one of two ways. “They stole our idea and they did it cheap, just like China always does.”

Paul: Mmm hmm.

Rich: Right? But instead, what they should really look at it as is, “This is a challenge to us.”

Paul: Yeah.

Rich: “This is a challenge to us to see if we still have the appetite to do hard stuff.” And I don’t think we do. I’ll say that right out of the gate. I don’t think we do. There’s a documentary that came out four years ago called American Factory.

Paul: Oh boy, that’s a rough watch.

Rich: It’s a rough watch.

Paul: Yeah.

Rich: And what it shows is—

Paul: Oh, they send those poor chubby Americans to China and they just are just subsumed in that. Yeah, it’s rough.

Rich: It’s not that subsumed. It’s just that, it’s just that I think, I think culturally and I just think just the circumstances, you just don’t have everything. You just don’t have anything. And if you don’t put your—look, I’m not saying those are good work conditions and all that, but you’re like, “Shit, this is what I have to do. This is all I’ve been given. I have to work within these constraints.” And constraints are a gift, dude. There is nothing more beautiful than 130 lines of code that you can’t believe are 130 lines.

Paul: No, no, no. I mean, look, this has been my whole life. I love the—constraints are, are utterly necessary for create, for creative thought. I think there’s also, like, these employees at DeepSeek are doing pretty good. Like they have, they’re, they’re being plucked out of colleges. They’re doing—

Rich: Sure!

Paul: They’re doing well. The CEO, his name is Liang Wenfeng. He is, like, just clearly a really thoughtful leader who’s been able to get good results.

Rich: Other than the big five or six companies that are just, like, are pretty much always the names that come up when it comes to AI right now, name me another one that’s just blowing you out of your chair. Name me one.

Paul: Oh, I know, I know.

Rich: There isn’t any. And I’m not saying they’re not doing interesting things. My point is this: This incredible opening occurred, like, three years ago. Right? And I think and then you marry that with just blank check, and you’re just not going to see a lot of, like, new invention come out on top of that. Maybe you will here and there. I don’t know.

Paul: I see it differently, which is that, what you’re asking for is like, where’s the great product? Okay? Where’s the thing? And a great product would be like an iPhone, the thing that brings us all together, and makes it—

Rich: No, let me ask you this, because I don’t think DeepSeek’s a great product. I think it’s a copycat. They just did it much cheaper, which is impressive in and of itself.

Paul: Yeah, but everybody’s borrowing the same ideas. I actually don’t—

Rich: Let me ask you this question.

Paul: I don’t even think it’s a copycat.

Rich: Why wasn’t this a team in Indiana?

Paul: For a few reasons, actually. If we, if we look at it—okay, if you want to be structural about it, right? This is a quantitative hedge fund. Okay?

Rich: There’s a lot of quantitative hedge funds in the U.S.—

Paul: Not in Indiana.

Rich: Fine. In New York. Why wasn’t this a team in New York?

Paul: Because—

Rich: Or California.

Paul: This guy went, “You know what? I’m a little tired of China having to eat it. I’m feeling the sting. I’m as smart as these knuckleheads I keep meeting from New York.”

Rich: Uh huh.

Paul: “And I think it’s probably time for us to show what we’re made of. And I think we can. And I happen to have many millions of dollars in the bank and a really big chip on my shoulder.”

Rich: Let’s give it a whirl.

Paul: Let’s give it a whirl. I’m going to go get some of the smartest PhDs I can. And I’m going to say, “Fellas! Ladies! Let’s do it.” And I’m going to see what comes out of it. And what’s great is that instead of, “I want what comes out of it to be, you know, this whole new thing and AGI and blah, blah, blah.” And this guy does believe in AGI. He thinks they’re working towards it. But what’s great is he said, “All right, look, they’re telling us we can’t have the new toys? So let’s show them what we got.”

Rich: Yeah.

Paul: And I—yeah. Do you think that that makes people stay up a little later at night? Yeah, absolutely.

Rich: I guess, let me put it differently. If I had gone to VCs in the middle of this frenzy of, you know, just AI mania and said, “Hey, I think I can just do whatever they are doing. It’s going to be a little worse, but it’ll be a lot cheaper.”

Paul: Yeah.

Rich: That’s not interesting. “I’m going to upend radiology” is much more interesting. Because Americans tend to not run back towards finding efficiency. They tend to run out towards new value.

Paul: Well, VC culture in particular.

Rich: And new opportunity. No, but, yeah, I agree, I agree.

Paul: But, like, the Xerox PARC idea was that the computer should cost as much as a house, because eventually it will cost as much as a computer.

Rich: Right.

Paul: So when you’re, when you’re experimenting and learning, you should just be burning cash. And that’s that ethos.

Rich: Yeah, that’s an ethos. Right? And I also think, look, I mean—

Paul: But these are—you know what’s funny? These are quantitative hedge-fund guys. When you think about what they actually do, they look for microsecond transactions and algorithms that can be executed in such a way that they make one penny a zillion times.

Rich: [laughing] Yeah.

Paul: Right? And so when you actually think about what, where this came out of, everybody’s like, “Wow, that’s wild.”

Rich: Yeah, a hedge fund.

Paul: “They had a bunch of GPUs—”

Rich: Yeah, Yeah.

Paul: But it actually makes perfect sense. They’re like, “Wait a minute. We always win on just transacting faster. So let’s do it over here.”

Rich: I gotta say—

Paul: Unless it’s all a wacky conspiracy, which people are saying, it’s just a…

Rich: I’ve heard this, by the way. We should say it out loud.

Paul: Yeah.

Rich: I mean, did they really do it with 1/100th the resources? I have no idea. We can’t confirm that. They did share out their approach in a very transparent way, which seems to indicate—and I think that’s why the markets are cratering—that they did, they used some slick tricks—

Paul: Richard, the markets are dumb.

Rich: Fine.

Paul: These—people were like, “Unlimited money. Number go up.”

Rich: I mean, this was coming anyway.

Paul: Then there was a whole weekend, and then, you know—and then the futures start to slip and everybody’s like, “Oh, I guess in Nvidia’s dumb now.”

Rich: Yes, that’s right.

Paul: I don’t buy a single bit of any of it.

Rich: I’m happy—look, man. I’m happy, I do think this, I do think AI is real. I do think there’s incredible innovation ahead. I’m happy this happened because I see it, I see it—this is, this is a dare. It’s almost like a challenge.

Paul: Let me tell you something. This is important, because I have been using DeepSeek nonstop for weeks, and I use it for hours a day. And what I do is I keep trying—I have now tried to solve the exact same problem about 3,000 times using different approaches.

Rich: Okay.

Paul: Okay? And what I do is I’ve now, I’ve now got four—I’m trying to build an API, classic software, I’m trying to turn a schema into an API.

Rich: Yeah.

Paul: And I have a set of prompts. And I now have gotten to a point where I can get the scaffolding set up, I can get the schema converted, and I’m very, very close to the part where it actually is working and implementable. If I go in and give it an hour, like, an hour of my time to fix the code, I will absolutely have what I need that is working.

Rich: Okay.

Paul: But what I’m practicing is can I get the machine to do it for me?

Rich: Right. Which is the hard part.

Paul: That gap will never be at zero. It will never be at zero, because I will always want to look—

Rich: Okay, so wait, let’s put a pin in DeepSeek, blah, blah, blah. You’re saying something different here, and I want to use what you’re saying right now—

Paul: I’m going to bring it back to DeepSeek again.

Rich: Well, no, no. But I think this deserves its own podcast.

Paul: Fine.

Rich: This is about people.

Paul: Yes.

Rich: And how people are going to be interacting with these incredible tools.

Paul: Here’s what we’ve got. We’ve got the American narrative, which is, okay, we’re just kind of doing this until we get it to be fully intelligent, and then it will take all the jobs.

Rich: I don’t know what that means. I want to have another podcast.

Paul: Okay, so what I’m saying is, like, I just think we’re always going to push it a little further down the road, because this is a database, it’s not an intelligence. And we’re humans, we’re monkeys. We’re going to keep hitting it with a stick and wanting it to do something just a little bit more.

Rich: Yeah.

Paul: There’s a strong argument to be made that—see, here’s what’s happened. I want this thing to build an app for me, and now I’ve got it to the point where it’s building the backend for the app, and it’s like 90% of the way there, and I would have to take it the remaining 10%.

Rich: Okay.

Paul: If I was a normal human being, I would say, “Wow, what a miracle. This is amazing. I couldn’t do this before. Now I can do weeks of work in minutes, and then I just have to take it that last little bit.” But what do I do instead? I practice over and over, trying to close up that last 10%.

Rich: Mmm hmm.

Paul: What happens when I close it up? I think I’ll get pretty close. I’m going to move on to the UI. I’m going to say, “Well, now it does that. Can I have it build me the frontend?” And we’re doing a little bit of that at work, which we’ve been talking about. Aboard is going to have some of those functions as well. But it’s—you’re just never going to get fully there because you’re always going to start to see another gap. Everybody’s like, “Oh, it’s going to do the miracle for you.” But no, you’re going to see one little thing.

Rich: Yeah.

Paul: And DeeSeek makes that very apparent. And you know why? Because you can hit the API infinitely, you can really, really start to see it because you’re not waiting for it all the time.

Rich: Yeah. Okay. Is—well, I don’t want to ask this question. I’ll ask it on another podcast. Like, I’m trying to draw a dotted line between what you’re saying is, like, there’s always that last leap and AGI. Is AGI going to take care of the last leap?

Paul: No, we are.

Rich: Why not?

Paul: We are. But the leap will keep showing up. No matter what you do. You’ll be like, “We’re done.” And it’ll be like, “Well, no, actually, there’s this one thing I wanted to do, and it won’t quite do it yet.”

Rich: Yeah.

Paul: It’s great. It’s a wonderful technology. Everybody should learn it. But this fantasy—and the market is crashing because it’s cheap and available? The market should be blowing up with excitement that suddenly this empowering thing that everybody told us is the super miracle technology that’s going to guide the future?

Rich: Yeah.

Paul: The fact that it’s cheap and there’s less moat should drive an unbelievable amount of global innovation.

Rich: Right, right, right, right, right.

Paul: But instead, everybody just wet their pants.

Rich: Yeah. Yeah. Well, no, I think money moves around. Money doesn’t—

Paul: I know it does.

Rich: And I think it’s been so concentrated in, like, five companies, and I think that’s getting reset.

Paul: But this was always on the radar. If you’d listened to this podcast and read our newsletter.

Rich: Yes.

Paul: Which, God knows, we don’t have any secret information, we’re reading the same API documentation as every other idiot out there.

Rich: That’s true.

Paul: You would have heard us say over and over, “This will become cheap and commoditized and run locally on your machine.”

Rich: That’s happening.

Paul: Geez, boy, anybody could figure that out. Everybody knew it. And suddenly, what it took was a global geopolitical shift [laughter] for everyone to be like, “Oh, whoa, if the Chinese do it, it won’t be under our terms.”

Rich: Look, here’s—I want to close this thought.

Paul: Yeah, close it. Finish this.

Rich: No, no, no. I will close it. Like, I’m not saying go gather your quant friends and do, do what DeepSeek did. Most don’t have quant friends.

Paul: I have…

Rich: Most quants don’t have friends.

Paul: I have a lot of quant friends.

Rich: You have a lot of quant friends.

Paul: I love my quants.

Rich: Here’s what, here’s the challenge I think it puts out there, because I think this is where there’s even more opportunity. I think this, this thing is, is, is truly, truly a platform, not a product.

Paul: Yes!

Rich: The challenge out there is can you create differentiating value that just makes people go, “Whoa!” Right? On top of this platform. And I think there are opportunities to do that. Okay?

Paul: You want to know one of the most exciting and disruptive technologies of the last 20 years?

Rich: What?

Paul: Postgres. I’m going to tell you what it is because you know what it is, but many people don’t. It’s an open-source database.

Rich: Yes. That’s right.

Paul: Okay? You want to do database stuff, you go ahead, you go type in “Postgres”—

Rich: It’s available to the whole world. It’s free.

Paul: You can have—it and it does everything you need to do with your data.

Rich: Does everything.

Paul: And I would say trillions and trillions of dollars of the world’s economy are now managed this way.

Rich: Oracle still in business. Microsoft SQL is still a product. But Postgres is here and it’s created immense value on top of it.

Paul: That’s right.

Rich: It’s a great point. Great closing point.

Paul: That’s the future.

Rich: So go innovate on top of this! Go ride it.

Paul: Yeah, that’s what we’re doing.

Rich: That’s what we’re doing.

Paul: We’re talking to all kinds of customers and people are in touch all the time.

Rich: Yeah.

Paul: It’s really cool. You can build a really great, thriving community on top of these technologies, if you’re open-hearted and want to actually learn.

Rich: Yes.

Paul: So you should get in touch with us!

Rich: [laughing] Hello@aboard.com. We are building a software-creation platform, a solution-delivery platform.

Paul: An engine.

Rich: An engine, on top of—

Paul: On top of this platform.

Rich: All this amazing stuff.

Paul: That’s right.

Rich: Check us out at aboard.com. We’ve got big, big exciting things cooking so we can’t wait to share those with you.

Paul: We really do. And we’re if you’re interested in—if your organization is trying to figure out like, “I gotta get AI into this, they told me, but I gotta like, I gotta deliver some value in here and build some software, and what’s going on?” Absolutely get in touch. Just ping us. hello@aboard.com. We’re here to help.

Rich: Have a lovely week.

Paul: Bye!

[outro music]