AI Gold Rush, Google's Identity Crisis, Moon Robots, and the Gadgets That Almost Work

Lauren: Welcome back to Tech Insider Weekly. Oh, man, do we have a packed episode for you today. Derek: Seriously, Lauren put together the most chaotic little stack of stories, and I am here for it. Lauren: Okay, so get this. We're talking AI, robots, space, and a pizza robot that ran out of dough. Literally. Derek: That last one is both sad and extremely funny. Lauren: Right? So first up, Sam Altman walked into a Y Combinator room. and basically said, here's two million dollars in OpenAI tokens, everyone. Derek: Wow. For every startup in the batch? Lauren: Every single one. TechCrunch called it a mic drop offer. We'll get into why some founders are less than thrilled about that. Derek: Yeah, tokens for equity has a few strings attached, let's just say. Lauren: Then we're moving into the Google AI Search saga. Axios has been covering this hard. Google is basically rebuilding its entire search experience around AI. around AI mode, and publishers are starting to feel it. Derek: And there are some genuinely weird failure modes. The Verge caught Google's AI Overviews responding like a chatbot when you search certain words. Lauren: As in, you type the wrong thing and it just ignores what you've asked. Derek: Plot twist, search doesn't search anymore. Lauren: And then... Space. NASA handed out nearly a billion dollars in lunar rover contracts, a Colorado startup raised 30 million betting robots build the moon base first, and a UK company literally sent a drug lab to the ISS. Derek: Whew, we've got a lot to get through. Lauren: We really do. Let's get into it, starting with the AI startup gold rush. Okay, so get this. Sam Altman walks into a Y Combinator event last week and basically offers every single startup in the room $2 million, every one of them. Derek: Which sounds amazing until you hear the fine print. Lauren: Right? because TechCrunch reported it and their headline was literally, mic drop moment. YC partner Tyler Bosmeny called it that. And I read it and thought, sure, mic drop, but also, is that mic wired? Derek: So walk us through it. What's the actual offer? Lauren: So Altman offered $2 million worth of OpenAI tokens to... To every start up in the current YC class in exchange for equity, not cash, tokens, API credits to build with OpenAI's products. Derek: Hmm, so he's investing in every founder with his own product. Lauren: Yes, he's literally paying you in coupons to his store! Derek: That is-I mean the audacity is kind of impressive. Lauren: The American Bazaar covered the founder reaction. And people are calling it tokenmaxxing, which, by the way, is a real term someone said out loud in a business context and nobody stopped them. Derek: Tokenmaxxing. Love that. Very normal phrase. Lauren: Completely normal. But the concern's a real, Derek. Critics flagged vendor lock-in, equity dilution, and loss of founder control. You take those credits, you build your whole stack on OpenAI's infrastructure, now what? Derek: Right, and once you're that deep into one platform, switching costs are brutal. You're not really independent anymore. Lauren: Exactly. And the equity piece is wild, too. OpenAI gets a... The stake in potentially every company in that YC batch. Derek: In the whole class? Lauren: Wow! Derek: That's not an investment strategy, that's a land grab. Lauren: Pretty much; and it connects to something TrueBridge Capital wrote up for Forbes about how a tiny group of AI companies absorbed hundreds of billions in capital, and reshaped the whole Midas List. Derek: Oh, I read that. Google debuted at around twenty three billion, Facebook went public at a hundred and four billion. billion. Now we're talking about potential trillion dollar IPOs. The scale has just completely broken the old model. Lauren: The Midas List used to reflect who found the best founders early; now it's starting to look like a list of people who happened to be standing near the right large language model at the right time. Derek: Right. Price place. Write LLM. Lauren: Ugh, that's the whole thing. Derek: And then there's the other side of this because while Altman's handing out tokens... OpenAI keeps shipping features that just quietly erase entire startup categories. Lauren: Oh, the Viacheslav Vasipenok piece, yeah. May 15th, OpenAI dropped a personal finance experience inside ChatGPT. Pro users can connect bank accounts, credit cards, investment portfolios, all through Plaid, with Intuit integration coming. Over 12,000 financial institutions supported. Derek: 12,000. Lauren: 12,000. And how many fintech startups spent the last three years building exactly that? Derek: I mean, a lot. A lot of them. Lauren: Nodding, yeah, the piece framed it perfectly. In the time it takes most fintech founders to raise a Series A, OpenAI ate their entire lunch. Derek: So on one hand, Altman's handing out tokens to founders, and on the other hand, his platform is becoming the product those founders were trying to build. Lauren: That's the tension. And honestly, the token maxing offer starts to look different in that light. Are you getting capital or are you getting absorbed? Derek: I'm on both sides a little bit. Like $2 million in API credits is genuinely useful for an early-stage company. That's real runway. Lauren: Fair point. Derek: But you're also building your house on someone else's land, and that someone keeps expanding their yard. Lauren: Great metaphor. Yeah. And look, Zendesk just announced a hundred million dollar commitment to startups. startups too. So it's not just OpenAI playing this game. Platforms are competing for founder loyalty. Derek: Which raises a bigger question. If the biggest AI players keep eating the startup space from below and VCs are just riding their coattails from above, who's actually left building something independent? Lauren: And maybe the more important question is what happens when the next big platform decides your product is just a feature? Is there any way to build something they can't just ship in an update? Ah, and speaking of a few massive players rewriting everyone else's rules, Google just did something at I/O that is genuinely hard to categorize as anything other than eating the internet alive. Derek: Which is a sentence I didn't expect to say in 2026, and yet. So Axios called it, and I think they earned the dramatic framing here, the end of the internet's golden age. And that's a big swing, but the argument holds up when you look at what Google actually announced. Lauren: How so? Derek: So AI Overview now has over a billion monthly users. The Register flagged this from Google's own search VP at I/O. Lauren: And AI Overview queries are doubling every quarter since launch, Derek: Wow. Lauren: which sounds great until you realize the whole point of AI Overview is to keep you inside Google, you never leave. Derek: Right, so the publisher ecosystem, every news site, every blog, every recipe website that depended on Google sending traffic their way, that's the thing getting hollowed out. Lauren: Slowly and without a press release about it. Derek: Oh, how convenient. Lauren: CNN had a good piece on the actual product logic here. Google's VP of search product basically said people are asking longer, harder questions now, questions that don't have a clean answer anywhere on the web. So Google is generating custom visuals, interactive graphics, mini apps, all running inside the search results page. Derek: Okay, so from a pure UX angle, I get it. You don't want to send someone to six different sites for one answer. Sure. Lauren: No, no, totally, and that's the tension, right? It's a genuinely better product in the moment, but the web it was built on doesn't survive the version of Google that's building it. Derek: I do wonder though, and I'm on both sides of this a little bit, is Google being uniquely evil here or is this just what every platform eventually does? Facebook ate social media traffic, Apple built App Store rules that crushed competitors, at some point platforms stop being pipes and start being destinations. Fascinating. Lauren: Fair, but here's what makes Google different: the web itself runs on the assumption that Google indexes it and sends people there. That contract is 25 years old. When Google breaks it, there's no alternate distribution. Publishers don't have a plan B. Derek: Yeah, that's the structural part that doesn't have a clean answer. Lauren: And then there's the weird part. Derek: Oh, there's a weird part? Lauren: The Verge caught something genuinely strange this week. This week, if you searched for the word disregard in Google, the AI Overview would respond as if you'd given it a chatbot prompt like it was trying to ignore its instructions. Derek: Oh, wait. It responded to the word disregard by disregarding the search. Lauren: Basically, Jay Peters at The Verge documented it. You'd type disregard and get this chatbot-style response that had nothing to do with the word itself. Derek: That is, that's a strange product decision to ship publicly. Lauren: I believe the technical term is a bug. Derek: Sure, sure, but is that a real thing, which is that when you wire a language model into something as consequential as search, the failure modes are weird and public and occasionally very funny. And Axios had a smart frame for the bigger strategic picture. Google's trying to disrupt its own products aggressively with AI while protecting the core business that generates tens of billions in profit. Profit? Speaker 3: Hmm! Derek: Unlike OpenAI, they have a vast empire to defend while they're doing it. Lauren: Which is its own kind of pressure: you're sprinting toward a future that might undercut your present. Derek: Exactly, and nobody's really stopping them. Lauren: So far. So far. Derek: You know it's wild to me, while Google is quietly rewriting how information works here on Earth, there are startups doing things so strange and ambitious that big tech genuinely hasn't figured out how to keep up. Not how to copy them yet. We're talking lunar robots and drugs made in orbit. Lauren: I mean, when you put it that way, it sounds completely made up. Derek: I know, and yet it's very real. That's next. Okay, switching gears completely from Google rewriting the Internet to startups literally leaving the planet. Lauren: Right, because why battle over search traffic when you can just go to the moon? Derek: Exactly. So here's where it gets good. Space Daily reported this week that a Colorado company called Lunar Outpost just closed a $30 million Series B, and their whole pitch is that robots, not astronauts, will build the moon's first permanent base. Lauren: And you know what that logic is hard to argue with. Like you send robots first, they do the dangerous, slow, unglamorous work. Derek: And by the time humans show up, there's actually a building to walk into. Lauren: Right, right. And it's not just theoretical. Aviation Week reported NASA just awarded contracts totaling nearly a billion dollars for lunar rovers, targeting deployment ahead of the Artemis IV mission in early 2028. date. Derek: A billion dollars for rovers? Lauren: Nearly a billion dollars. And they're crew capable. So this is real infrastructure spending, not just concept art. Derek: There's also this other Colorado story I can't stop thinking about. Denver Gazette covered a startup called Katalyst Space Technologies, and they built a robotic spacecraft in eight months to go rescue a dying NASA telescope. Lauren: Wait, eight months? Derek: Eight months from a napkin sketch to an actual spacecraft called Link. Link, the telescope the Neil Gehrels Swift Observatory was going to burn up in Earth's atmosphere later this year. Lauren: That's honestly kind of heroic. Derek: Kind of heroic! And it's such a good example of small startups doing things big institutions move too slowly to do themselves. Lauren: Okay, but I have to bring up the story that made me do a genuine double take this week. Derek: Oh yeah? Lauren: Oh, the Guardian reported that a UK startup called BioOrbit launched a box the size of a microwave onto a SpaceX flight on May 15th. Derek: a microwave. Lauren: A microwave headed to the International Space Station, and inside they're growing ultra-pure protein crystals using microgravity to make self-injected cancer drugs. Derek: So wait, the idea is that you literally cannot make this medicine cleanly on Earth? Lauren: That's exactly it. Yet gravity messes with how the crystals form; in orbit you get a purity you just can't replicate down here. Derek: So the ISS is also a pharmaceutical lab now? Lauren: The unit stays up for about six weeks, then comes back down. BioOrbit thinks this could scale into a treatment that saves millions of lives. Derek: I mean, if it works, that's extraordinary. The barrier to entry is just, you know, orbital mechanics. Lauren: Minor detail? Derek: Minor detail. And yet here we are, watching startups clear that bar while the rest of us are still arguing about pizza delivery algorithms. Lauren: Speaking of which, the pizza robot story coming up is going to hit very differently after all this. Derek: Deadpan from drugs and orbit to dough on the floor. That's this show. From space pharmacies to pizza robots. You know what? This episode has range. That's the pivot I needed. OK, so Picnic, the Seattle startup that raised over $50 million to build pizza-making robots, partners with Domino's, and then... Lauren: runs out of dough. Derek: D'oh! Lauren: I'm sorry, I had to. GeekWire and Futurism both covered it. The company went insolvent, liquidated everything, sold off the IP to some mystery buyer nobody's named yet. Derek: A mystery... Lauren: Buyer for robot pizza arms-that's a sentence, right? Derek: And here's the thing: ten years they spent a decade on this-the robot could assemble a pizza, that part worked. Lauren: So what killed it? Derek: Honestly, the economics of food service are brutal, labor's cheap relative to capital equipment costs, and restaurants don't have the margins to absorb a hardware bet that size. The robot worked, the business didn't. Lauren: That's the recurring lesson, isn't it?--the technology clear He clears the bar and then reality shows up. Derek: Hundred percent. Lauren: Switching gears, because not all hardware is dying, smart glasses, Derek, smart glasses are actually having a moment. Derek: I genuinely did not have that on my 2026 bingo card. Lauren: So TechCrunch got hands-on time with Google's Android XR glasses at I/O. In-lens display, Gemini overlay, real-time translation directly in your field of view, and they said... head almost there. Derek: Almost there is doing a lot of heavy lifting. Lauren: It is. But Xreal, Google's partner on this, their CEO told TechCrunch they think the industry has finally hit its turning point after a decade of being, quote, a financial black hole. Derek: Which, to be fair, it absolutely was. Remember Google Glass? That was 2013. Lauren: We don't speak of Google Glass. Derek: Fair. But okay, if the glasses are the hopeful story. What's the creepy story? Lauren: Oh, I'm so glad you asked. Amazon's Bee Wearable. Derek: The wrist thing? Lauren: The wrist thing. TechCrunch tried it. It records, transcribes, and summarizes your conversations all day. Syncs with your calendar. Very useful if you're forgetful. Derek: And very useful if you're a surveillance device? Lauren: The reviewer literally said they felt both intrigued and slightly creeped out. Doubt. That's the quote. That's the whole product review in seven words. Derek: I mean, intrigued and slightly creeped out is also my review of this entire era of AI wearables. Lauren: Right? Like, at some point we just normalized wearing something that listens to everything we say and summarizes it for a trillion dollar company. Derek: At least the pizza robot didn't know your schedule. Lauren: That's the bar now. That's where we are. Okay, that was a packed one, from Sam Altman handing out tokens like coupons to his own store to Google quietly walling off the open web. Derek, what's your one takeaway? Derek: Honestly, the platforms are winning. Whether it's OpenAI absorbing startups its own product is replacing, or Google making sure you never leave their garden, the gravity is getting stronger. Founders are building inside someone else's orbit now. Lauren: Love that; and somewhere in the middle of all this a pizza robot quietly ran out of dough. Love that. Derek: Classic. Lauren: With warmth, the through line this week? Power is consolidating fast and the window to build independently is shrinking, worth paying attention. Derek: Yep, Speaker 3: Mm pay-hmm. Derek: attention. Lauren: Smiling, if you liked today's episode, subscribe wherever you listen and leave us a review-it genuinely helps. Really helps. Got a founder story or topic we should tackle? Tag us on social. Derek: New episodes every Wednesday. Thanks for listening, everyone. Lauren: Warmly, see you next week.