Robots, Runaway Valuations, and the $80 Video Game: This Week in Tech

LaurenBGHOST2mhm

DerekOkay, welcome back to Tech Insider Weekly. I'm about ready to burst with this one.

Speaker 3That's a strong open. What's got you like this?

DerekChamath is CEO again of a coding startup, 135 million dollar Series A.

Speaker 3Wait, wait, wait. Chamath is actually running the thing now?

DerekPlot twist, right? We'll get into it.

Speaker 3We're also looking at this Crunchbase data on Q2 exits, biggest quarter for... For billion-dollar startup deal since 2021.

DerekWhich raises the real question, with six mega VC firms basically running the table now, what does that mean for founders who aren't in the room?

Speaker 3That's a thread we're pulling all episode, honestly.

DerekSwitching gears here, Derek, tell them about the robots.

Speaker 3Oh, you're going to love this part. A $5 billion startup putting humanoid robots to actual work,

DerekWow.

Speaker 3not a demo, real shifts.

DerekShut up!

Speaker 3Dead serious—plus this wild claim out of a former Databricks AI exec cutting AI's power bill by a thousand times.

DerekA thousand? Okay, we need receipts on that one.

Speaker 3We're chasing 'em down.

DerekAnd later, GTA Six's price tag, Xbox discounts before a hike, and why Google employees are walking away from what used to be the dream job.

Speaker 3A lot of ground today.

DerekSo much ground. Derek, where do we even start?

Speaker 3Eh, ground. Fair enough. Let's get into the Chamath news first.

DerekOK, so get this, $135 million for a Series A. One round.

Speaker 3Whistle. That's not a seed round. That's a small country's GDP.

DerekRight? So Chamath Palihapitiya's startup 80-90 Labs just closed a $135 million Series A led by Salesforce Ventures with Jeffrey Katzenberg's WndrCo and David Sacks' Craft Ventures in the mix too.

Speaker 3And the part that actually stopped me? He's stepping in as CEO. He didn't just fund it, he's running it.

DerekThat's a nice story. Founder investor takes the wheel. But show me why that actually matters.

Speaker 3Founder psychology, honestly. When the guy with the checkbook also owns the P&L, the incentives line up in a way that a hired CEO's never quite match.

DerekOr it's just a really expensive vanity project with better PR.

Speaker 3Could be both.

DerekLet's get real for a second, though, because the money isn't just going to famous podcast guys. Crunchbase's data shows Q2 of this year had the most billion dollar startup exits since the 2021 peak.

Speaker 3Wait, since 2021? That was the top of the mountain.

DerekI know. And you've got the biggest venture-backed exit of all time sitting right in there. SpaceX.

Speaker 3No way. All time? Like ever?

DerekAll time ever. So the exit markets thawing. Thawing money's moving again and somehow it's landing in fewer and fewer hands.

Speaker 3Which is the Giga-VC thing you keep bringing up.

DerekExactly. Inc. had a piece on this. Just six firms now dominate the majority of startup financing.

Speaker 3Six. Not 60, not 600. Six.

DerekAnd that changes how founders even pitch. You're not shopping a term sheet to 20 firms anymore. You're praying one of six giants likes your deck. deck

Speaker 3That's a real power shift in negotiating leverage. If there's only six buyers, the buyers set the price.

DerekOkay, but what actually happened with valuations though? Because this is where I get twitchy. Forbes profiled this guy David Silver, former DeepMind scientist, pitched investors on a company with literally no product.

Speaker 3No revenue, no product, just a 30-minute Zoom pitch?

DerekJust a pitch. And Forbes reports founders like him are structuring rounds so VCs can The seeds can invest at wildly different price points in the same deal, basically manufacturing a sky high headline valuation without ever proving the thing works.

Speaker 3So it's less, here's our traction, and more, here's our story, and by the way the tranche pricing makes the number look huge.

DerekThat's my read. It's a nice story, but show me the actual product.

Speaker 3Some things never change, huh? Hype dressed up in new financial plumbing.

DerekThe plumbing just got a lot fancier, though. Six firms, 100-plus million checks, no revenue unicorns.

Speaker 3And a podcast guy running his own AI company now.

DerekWhen you put it like that, it sounds like a fever dream.

Speaker 3It kind of is one, but it raises the obvious next question.

DerekWith all that cash flooding in, who's actually building the thing?

Speaker 3Because at a lot of these companies, it isn't the engineers typing the code anymore.

DerekSpeaking of who's actually building these things, I gotta say, the coding story is wilder than the funding story.

Speaker 3Wilder than a $135 million round closing in what, a week?

DerekAlmost. So Ben Bergman at Business Insider talked to Rami Alhamad, the CEO over at Alma. That's a nutrition coaching app backed by Menlo Ventures.

Speaker 3Okay.

DerekRami's quote was basically, nearly everything we ship now is AI-generated. Not a lot-nearly everything.

Speaker 3Wait, Wait—nearly all the code written by a model?

DerekThat's what he told Bergman, and Business Insider surveyed dozens of founders finding the same pattern across the board.

Speaker 3Okay, but what actually happens when something breaks? Because someone still has to read that code.

DerekThat's the new problem. Bergman's piece calls it "slop"--code that ships fast but nobody fully understands.

Speaker 3So you've got speed with no memory of why anything was built the way it was.

DerekExactly. And that's this whole segment's premise. Speed's a double edged sword when the sword doesn't leave a paper trail.

Speaker 3Which, funny enough, is probably why Investors threw $135 million at fixing it.

DerekRight. Different company, though. This is 8090. Chamath's shop from last segment.

Speaker 3Oh, the same 8090?

DerekSame one. SiliconANGLE reported Salesforce Ventures led that Series A. It's a with Nikesh Arora and Adam D'Angelo all in.

Speaker 3So Chamath raises the money, becomes CEO, and the pitches will clean up the mess AI just made?

DerekBasically, their flagship product is meant to be the AI development layer for engineering teams going forward.

Speaker 3Let's get real for a second-is that a real product or a bet on a future that doesn't exist yet?

DerekFair question, but given how much slop Business Insider is describing, someone's going to need the janitor tool.

Speaker 3The janitor tool-I like that. Switching gears, though, are you ready for the actual jaw dropper of this segment?

DerekOh, hit me.

Speaker 3Naveen Rao-former head of AI at Databricks.

LaurenHis new company is called Unconventional AI.

DerekOkay.

LaurenTechCrunch's Russell Brandom reported this week that Rao thinks he can cut AI's power bill by a thousand times.

DerekI'm sorry, 1,000X, not 10%, not double efficiency, 1,000 Wow. times less power?

LaurenThat's the claim. Different chip architecture. Oscillator based instead of the transistor stuff everyone's using now.

DerekOkay, but what actually happened Thursday? Did they ship any? If anything, or just a slide deck.

LaurenThey released a real model, an image generator called Un-0, according to Brandom's report, as a proof of concept.

DerekSo it's not theoretical anymore.

LaurenNot entirely, it's early, but it's a working system, which is more than most power claims come with.

DerekBecause right now, data centers are the actual bottleneck on all this AI money we keep talking about.

LaurenRight, and that's the tension. Everyone's racing to build models while the power grid just... The groans!

DerekSo, somewhere in Menlo Park there's a guy hoping his oscillator saved the industry's electric bill.

LaurenPretty much. And zoom out for a second. John Sviokla at Forbes has a piece tying this to something bigger.

DerekGo on.

LaurenHarvard Business School and INSEAD studied over 2,900 Y Combinator startups and found AI-native firms just run leaner, fewer people, period.

DerekNot doing old workflows faster, actually built different from. Right from day one.

LaurenThat's the finding, which raises the question hanging over this whole segment:

DerekIf the AI is writing the code and maybe cutting the power bill a thousandfold, it's gotta go somewhere real, right?

LaurenOh, it's going somewhere very real: into actual machines, robots, cars that drive themselves.

DerekNow that I want to see. Shifting gears from code to actual metal—are you ready for the robot part?

LaurenWait, wait, are we talking Terminator or are we talking Roomba?

DerekSomewhere in between Bloomberg had this piece Thursday on a five billion dollar startup putting humanoid robots to actual work—not a stage demo—real shifts.

LaurenFive billion for robots that what, walk around and lift boxes?

DerekBloomberg's framing was that this is the first real real test of whether humanoid robots can do paid work at scale, not just wow a crowd at a conference.

LaurenOkay, but putting them to work doing what exactly? Folding laundry?

DerekMore like warehouse and factory floor tasks. The bet is labor, not spectacle.

LaurenSure, ask me again in a year if it's still working or just standing there.

DerekFair. But speaking of robots that actually generate revenue, Business Insider had Rya Jethas piece on FieldAI hitting a $100 million milestone.

LaurenA hundred million in funding?

DerekRevenue and contracts combined, their software helps robots navigate mines, construct construction sites places Wow. humans really don't want to be crawling around in exactly

LaurenOkay, that I actually like. Send the robot into the collapsed mine shaft, not my cousin Gary.

Derekthe pitch oh

LaurenAlright,

Derek:

Laurenswitching gears here, because I have been waiting to talk about this one. Robotaxis in San Francisco.

Derekno what happened so TechCrunch Kirsten Korosec pointed out these empty self-driving cars are constantly cruising around Not to pick anyone up, just heading to some depot to get cleaned and charged.

LaurenWait, so they're just driving themselves to spa day?

DerekBasically, it's literally called deadhead miles in the industry, miles driven with zero paying passenger.

LaurenThat seems like it would kill the whole profit math.

DerekIt does. Korosec called it one of the biggest barriers to robo-taxi profitability, and this is where Aseon Labs comes in, Redwood City's startup out of Y Combinator's spring cohort.

LaurenAnd their fixes?

DerekA parking-space-sized pods scattered around the city that clean and charge Charge the cars on the spot. No detour to some depot forty minutes away.

LaurenHuh. So instead of the car going to the spa, the spa comes to the car.

DerekThat's the whole business model, yeah. TechCrunch says they've raised $10 million from Crane Venture Partners for it.

LaurenSmall check next to that $5 billion robot company, but a real problem to solve.

DerekNow flip that on its head, because not every startup is chasing more sensors, more software.

Speaker 3Software.

LaurenOh, you mean Slate.

DerekI mean Slate. KTLA road in one of these electric pick ups starts at twenty four nine fifty.

LaurenTwenty four nine fifty is cheap for an EV.

DerekAnd here's the twist: no built-in radio, hand-crank windows.

LaurenWait, actual hand cranks in twenty twenty six?

DerekOn purpose. KTLA said it's about simplicity and customization over stacking on more screens.

LaurenWhile everyone else is racing to put a touch screen on the touch screen.

DerekMeanwhile, Slate's basically saying, what if the car was just... A car.

LaurenI respect the contrarian move-not every problem needs an AI bolted onto it.

DerekRobots doing real shifts, taxis skipping the spa commute, and a truck betting on less tech not more. Quite a spread.

LaurenAll that venture money finally showing up as something you can touch.

DerekSpeaking of things you touch, your wallet's up next, because the gadgets and games waiting for you are not getting any cheaper. Okay, shifting gears from robots to something that hits closer to home: your actual wallet.

LaurenOh no, what now?

DerekGTA VI Ars Technica reported it's going to cost more than other AAA games. We're talking above that usual seventy dollar mark.

LaurenWait, really? For one game?

DerekYeah, and Ars Technica framed it as this test case, like publishers want to know what the ceiling actually is. Scope and scale cost money now.

LaurenI mean to be fair-if any game can charge a premium, it's that one.

Speaker 4Mm-hmm.

LaurenBut it does set a precedent.

DerekRight; everyone else is watching to copy the number.

LaurenSpeaking of pay more right before you regret it, IGN had this Xbox story.

DerekSmooth transition.

LaurenMicrosoft's jacking up Xbox Series X prices by up to one hundred and fifty dollars starting August first, and Walmart and Target are discounting the console to $

Speaker 5one hundred and fifty dollars.

LaurenFive hundred and seventy three dollars and ninety nine cents right now before the hike hits.

DerekSo it's not even really a discount. It's a last chance before we squeeze you sale.

LaurenPretty much.

DerekClassic.

LaurenMeanwhile, over at Instagram, Adam Mosseri's testing new ways to let you customize your algorithm. More control over what you see, less control over what they show you.

DerekSure, give users a dial, but who's really steering? AIRING.

LaurenThat's That's the question nobody's answering.

DerekAnd here's one that surprised me: Business Insider talked to workers who left Google, the dream job.

LaurenWait, people are leaving Google? On purpose?

DerekJacob Zinkulas's piece profiled a few: Bushra Amiwala, Yousuf Imran, and Taylor LaSane all walked away for startups. Wow.

LaurenThat's wild, honestly. Considering everything we've talked about today-the money, the robots, people are still choosing the chaos over the comfort.

DerekYeah, and I don't think that's a coincidence, Derek. Big companies are stable, sure, but start ups are where the interesting risk is right now.

LaurenWhich is basically our whole episode in one sentence.

DerekFunding, robots, and now the humans deciding where to point themselves.

LaurenSo watch the console shelf, watch the games price tag, and watch who's quitting Google next quarter, because apparently all three are telling the same story about where the money's at.

DerekAnd he's actually going. OK, so that's the episode, and honestly, that Chamath moment still gets me. Stepping into the CEO chair himself?

LaurenBetting on his own company-that's the confidence.

DerekThe big takeaway, though? Funding's getting weirder and more concentrated at the same time. Founders have to work harder to stand out.

LaurenRight, and whether it's six mega VC firms calling the shots or robots doing actual work in Texas, the theme's the same. The nice story isn't enough anymore.

DerekShow me the actual product.

LaurenExactly.

DerekIf you like this one, subscribe wherever you're listening and leave us a review-it helps.

LaurenGot a founder we should talk to, or a story we're missing? Send it our way. Tag us online.

DerekNew episodes every Wednesday.

LaurenThanks for hanging out with us.

DerekSee you next week.

Speaker 3Oh