Tech Brew Ride Home - Sam Says Some Things

Episode Date: February 23, 2026

Sam Altman has a lot of stuff to say about AI and let’s just say comms needs to have a quiet word with him. SaaS may or may not be dead, but is the replacement vertical AI? And a big thought experiment. If the AI bulls are right and AI transforms the economy, what might that look like? Sam Altman Says Companies Are ‘AI Washing’ Layoffs (Gizmodo) People Loved the Dot-Com Boom. The A.I. Boom, Not So Much. (NYTimes) Sam Altman would like to remind you that humans use a lot of energy, too (TechCrunch) Half the AI Agent Market Is One Category. The Rest Is Wide Open. (Garry's List) Google Restricts AI Ultra Subscribers Over OpenClaw OAuth, Days After Anthropic Ban (Implicator.ai) THE 2028 GLOBAL INTELLIGENCE CRISIS (Citrini Research) Learn more about your ad choices. Visit megaphone.fm/adchoices

It's peak pollination season, and my business is scaling fast. To keep the nectar flowing, I need a phone plan with top priority data speed. That's why I chose GoogleFi Wireless. My connections stay strong even when the hive is buzzing. Plus, unlimited plans start at $35 a month. Now, that's a deal that doesn't stay. Explore Google Fi Wireless plans today. Plus taxes and government fees.

Google Fi Wireless is not subject to data traffic deprioritization during times of high network usage. Welcome to the TechBrew right home from Monday, February 23rd, 2026. I'm Brian McCullough today. Sam Altman has a lot of stuff to say about AI and let's just say Coms needs to have a quiet word with him. Sass may or may not be dead but is the replacement vertical AI and a big thought experiment. If the AI bulls are right and AI transforms the economy, what might that look like? Here's what you miss today in the world of tech. Sam Altman has been saying some things and you can do some real tea leaf reading in to some of the things he's said. First, quoting the Indian Express. In his recent outing with Express ATA, OpenAI CEO Sam Altman spoke about meeting the staggering future demands of artificial intelligence. When Anat Glanca, executive director of the Indian Express, asked about putting data centers in space, Altman quipped saying the idea was ridiculous. This is in stark contrast to Altman's chief rival billionaire Elon Musk, who has been speaking profusely about moving AI computation off Earth into space, specifically into orbit.

with the help of satellites and solar power. Putting data centers in space with the current landscape is ridiculous. Orbital data centers are not going to matter at scale this decade due to the rough math of launch costs and how hard it is to fix a broken GPU in space. Altman told Gunka, other than the massive physical infrastructure of AI, the landscape seems to be also shaped by intense personal dynamics and rivalries. Altman's relationship with Musk has been famously fraught. During the interview, Gunka asked Altman to...

weigh the probabilities of two highly unlikely scenarios, whether Taiwan semiconductor manufacturing would lose its global monopoly on chip manufacturing or if he and Musk would ever become friends again. Altman was unambiguous in his response. I think Musk and I becoming friends again is less likely, he admitted, adding with a touch of humor, I feel like I have more control over that one, end quote. Now from Gizmodo, quote, Sam Altman is starting to get the sneaking suspicion that companies might be using the technology he's dedicated his life to, once it turned out to be extremely profitable, as cover for their own interests. In an interview with CNBC TV 18 at the India AI Impact Summit, the founder and CEO of OpenAI suggested that AI has become a scapegoat that is wrongly being

blamed for the mass layoffs that continue to hit basically every sector of the economy. I don't know what the exact percentage is, but there's some AI washing where people are blaming AI for layoffs that they would otherwise do, and then there's some real displacement by AI of different kinds of jobs, Altman said. Of course, Altman's got to thread the needle here. He does, in fact, need people to believe his company's technology can replace people. That has kind of become the whole pitch to corporations looking to pour money into AI despite little real return on those investments thus far. But he also would rather not position his product as a job killer, lest he rile the masses who are on edge that their jobs might get axed. We'll find new kinds of jobs,

as we do with every tech revolution, he added, but I would expect that the real impact of AI doing jobs in the next few years will begin to be palpable, end quote. That said, Altman is probably onto something here, according to data from consulting firm Challenger Gray and Christmas, about 55,000 layoffs in 2025 were attributed directly to AI. While that is a significant number of people out of work, it also accounted for less than 1% of all job losses for the year. A recent paper published by the National Bureau of Economic Research found that 90% of executives surveyed said AI has had no impact on workplace employment in the last three years, end quote. And then here are the quotes that lend themselves to real tea leaf reading. Quoting the

Times. The chief salesman of the AI boom, Sam Altman of OpenAI, said there was more resistance to, quote, the diffusion, the absorption of AI into the culture and economy than he expected. Looking at what's possible, it does feel sort of surprisingly slow, Mr. Altman said at an AI conference this month. From the point of view of the AI promoters, an inability to relate to human beings is not a defect but a virtue. Quote, I suspect that in a couple of years on almost any topic the most interesting, maybe the most empathetic conversation that you could have will be with an AI, Mr. Altman said, end quote. Quoting Bridget Farlad on LinkedIn, it's almost as if these tech leaders crowing about the end of jobs and humanity, as we know, it isn't sitting well with folks.

Crazy, end quote. Yeah, and it doesn't help, I guess, to say things like this. Quoting TechCrunch, Open AI, CEO Sam Altman addressed concerns. about AI's environmental impact this week while speaking at an event hosted by the Indian Express. Altman complained that many discussions about Chad GPT's energy usage are, quote, unfair, especially when they focus on, quote, how much energy it takes to train an AI model relative to how much it costs a human to do one inference query. But it also takes a lot of energy to train a human, Altman said. It takes like 20 years of life and all of the food you eat during that time before you get smart. And not only that, it took the very widespread evolution of the 100 billion

people that have ever lived and learned not to get eaten by predators and learned how to figure out science and whatever to produce you, end quote. So in his view, the fair comparison is, if you ask chat GPT a question, how much energy does it take once its model is trained to answer that question versus a human? And probably AI has already caught up on an energy efficiency basis measured that way, end quote. Quoting David Beesis, why does Sam insist on declaring war against his entire customer base. That's a truly bizarre positioning for a mainstream business, end quote. And quoting Kunal Kapoor, quote, I'm very pro-tech and his intent here is to talk about energy, but you can see why people are becoming increasingly uncomfortable with the tech pros.

It reflects a broader discomfort in how they frame humanity. Culturally, some of this language starts to echo the archetype of the technocrat villain you see in movies, the genius who sees humanity as a system to optimize. It's turning the human experience into energy and mathematics, but people don't want to be solved, they want to be understood, end quote. Anthropics' own data now shows that software engineering accounts for around 50% of its AI agent tool calls, which is interesting on its own. But Gary Tan points out the remaining verticals that compromise that other 50% are what is interesting to him and that most founders are overlooking this. Quote, software engineering owns half of all AI agent activity. The other half is scattered across 16

verticals, none above 9%. Health care is only at 1%. Legal is 0.9%. Education is 1.8%. These aren't saturated markets. They're markets that barely exist. Here's what should make founders salivate. The models are already more capable than users trust them to be. Metter's capability assessments show Claude can solve tasks that would take a human nearly five hours. But in practice, the 99.9th percentile session runs only about 42 minutes. That gap between what AI can do and what we let it do is a massive opportunity. Between October 2025 and January 26, the 99.9th percentile turn duration nearly doubled from under 25 minutes to over 45 minutes. The growth is smooth across model releases. This isn't just better models, it's users

extending trust session by session as they learn to work alongside agents. Aaron Levy points to the untold wealth and value ready to be unlocked. Build agentic software that taps into proprietary data, make the software actually work for real people and problems, stuff that context to maximize intelligence coming out, and the part most founders miss, drive change management for customers. That last piece is why vertical AI is so defensible. Anyone can build a wrapper. Few can navigate the specific workflows, regulatory constraints, and organizational friction to health care billing or legal discovery or construction permitting. SaaS has grown 10x per decade for a few decades now.

Over 40% of VC dollars in the past 20 years went to SaaS companies. The industry produced more than 170 SaaS unicorns, and so the thesis is simple. Every one of those unicorns has a vertical AI equivalent just waiting. And the AI versions could be 10x larger because they don't just replace software, they replace the operators too. The map is drawn. Software engineering is spoken for healthcare, legal, finance, education, customer service, logistics, 16 verticals with single-digit market share each are waiting for

someone to build the domain expertise into the agent. 300 SaaS unicorns came before. 300 vertical AI unicorns are coming next. The founders who pick a vertical, build domain expertise into their agents and figure out change management will own the next decade of enterprise software. The models can already work for five hours. Users only let them work for 42 minutes.

That's an indicator. We are so early, and there is a lot more to build, and in so many places that haven't even seen a single minute of intelligence in action, end quote. Study. And play. Come together on a Windows 11 PC. And for a limited time, college students get

the best of both worlds. Get the Unreal College deal, everything you need to study and play with select Windows 11 PCs. Eligible students get a year of Microsoft 365 premium and a year of Xbox GamePass Ultimate with a custom color Xbox wireless controller. Learn more at Windows.com slash student offer.

While supplies last, ends June 30th, terms at AKA.m.S. College PC. Ready to soundtrack your summer? With Red Bull Summer All Day Play, you choose a playlist that fits your summer vibe the best. Are you a festival fanatic, a deep end DJ, a road dog, or a trail mixer? Just add a song to your chosen playlist and put your summer on track. Red Bull Summer All Day Play. Red Bull gives you wings.

Visit Red Bull.com slash Bright Summer Ahead to learn more. See you this summer. All. Pay off your home. Travel for life. Drive a Ferrari. of the world premiere of the Monopoly Big Board Buckslot machine by Aristocrat Gaming, Yamava Resort and Casino at San Manuel is giving one person a $1.6 million dream package. The biggest prize in Yamaba's history. Club Serrano members can earn daily instant prizes and secure a spot in the finale May 29th. Don't pass go and own it all. Only at Yamava, celebrating its 40th anniversary.

You win? Details at yamava.com must be 21-20. Please gamble responsibly. Monopoly is a trademark of Hasbro. Hasbro is not a sponsor of this promotion. Ambition comes in all shapes and sizes. At First Citizens Bank We roll with your goals because we're built for what you're building. Fit for your ambition for citizens back. Google anti-gravity users say their paid Google AI accounts were banned after they linked Gemini models via OpenClaw. Peter Steinberger says he may remove support for this soon, quoting implicator.aI. Google has restricted accounts of AI ultra-subscribers who access Gemini models through OpenClaw,

a third-party Oath client, according to a growing thread on the Google AI Developer Forum. The restrictions arrived without warning or explanation, cutting off users paying $249.99 per month from Gemini 2.5 Pro, and in some cases, threatening access to Gmail, workspace, and other linked services. The enforcement comes two days after Anthropic updated its legal terms to explicitly ban Oath token usage in third-party tools, including OpenClaw. Two of the three largest AI model providers locked down third-party access in the same week. Same calculation, same week. The forum thread that started this carries a title worth reading and full. Account restricted without warning, Google AI Ultra Oath via OpenClaw. The original poster described losing access to Gemini 2.5

Pro after connecting through OpenClaw an open source AI agent framework that routes existing subscriptions through alternative interfaces. No terms of service violation was cited. No explanation followed. Other users arrived with matching stories. Some submitted a piece. heels and received only automated replies. Others couldn't find a human representative at all. One user wrote that they tried creating a fresh Google account only to discover that one restricted two. No warnings, no nothing. Just a ban after being a customer for decades, the user wrote threatening to cancel YouTube, Google Ultra, and every other Google product. Another described the situation in terms of what it felt like from the inside. I'd have to sue a trillion-dollar company

just to get a measly fee I paid. Nearly $3,000 a year and the best Google could offer was a community manager promising to share this with our internal teams, no timeline, no specifics, and the thread has kept growing. And here is where Google's architecture turns an annoyance into something worse. Google's AI products sit on top of the broader Google account infrastructure. A restriction triggered by Gemini usage can cascade into Gmail, workspace, and cloud storage, all tied to the same credentials. Several forum users flagged this exact concern. For a developer whose entire business runs on Google services, getting your AI subscription flag doesn't just cut off one product, it puts everything at risk. Google's crackdown didn't

arrive in a vacuum Anthropic two days earlier on February 20th had revised its consumer terms of service to spell out what had been loosely implied since February. Oath tokens from Claude Free, pro, and max accounts are only permitted in Claudecode and clod.aI. Using them anywhere else, including OpenClaw, violates the terms. What makes the timing notable is that Anthropics' contractual language in Section 3.7 has already forbidden unauthorized third-party access since at least February of 2024. For two years, that clause sat in the terms while tools like OpenClaught and Open Code quietly let users supply Claude subscription keys anyway. Anthropic had tolerated the practice or at least failed to enforce it until the economics forced their hand. The registers Thomas

Claiborne framed the economics bluntly. Anthropic sells subscription tokens at a discount compared to API pricing, and all-you-can-eat buffet priced with certain usage expectations. Third-party harnesses broke those expectations by letting subscribers extract more value than the subscription model assumed. Token arbitrage. Pay the flat rate, route tokens through a tool that burns through them faster than anyone planned for. OpenAI, emboldened by this, has leaned into the gap.

T-Balt-Sotio from OpenAI pointedly endorsed using codex subscriptions with third-party harnesses. competitive positioning, dressed up as developer-friendliness, end quote. Finally, today I found this very interesting over the weekend. It's from an investment analyst firm called Citrini Research. Think of this as a thought experiment imagining an AI-driven 2028 global intelligence crisis. Basically, imagine this piece as a theoretical memo dropped on a trading desk in the future, say June 30th, 2028. The theoretical story the authors tell is basically what happens if the AI optimists are right,

and AI has transformed the economy like some expect it to. The scene the piece sets is a bit concerning. Unemployment numbers just hit 10.2%. The S&P is selling off again, and the index is now down 38% from the October 2026 highs. Again, in this scenario, AI has won. Productivity rips. Output per hour rises like it's the 1950s again. Companies dump labor, margins widen, earnings beat, and a huge chunk of those profits go straight back into more compute. The markets in this little future history are euphoric, while the lived economy quietly stops looking like the one we recognize today. They even coin a phrase for it, ghost GDP, real output on the spreadsheets that doesn't trickle down to households because the workers in those households producing that output don't buy groceries.

take vacations or remodel kitchens. Machines spend exactly $0 on discretionary goods, so the velocity of money stops. According to the piece, this scenario starts in late 2025 when agenetic coding tools take a step-change jump. Suddenly a competent developer with tools like Claude Code or Codex can recreate the core of a mid-market SaaS product in weeks. Good enough that a CIO staring at a $500,000 renewal starts asking, why are we paying for this? Procurement discovers a new superpower, even threatening to rebuild, becomes a negotiating weapon. Sass pricing cracks first, and then the broader intermediation layer starts to wobble. Travel booking, insurance renewals, tax prep, routine legal work, even some real estate commissions as agents with MLS access grind fees down.

The theory here is we've always overestimated relationships. A lot of what we called relationship in the economy was just friction with a friendly face. Once agents control the transaction, they behave like little heat-seeking missiles for fees. Delivery apps fragment because a machine doesn't have home screen loyalty or brand loyalty. It checks every option every time and picks the cheapest or the fastest. Then payments come next. If agents are transacting on your behalf, why pay card interchange fees at all? So agents route around it, stablecoins on Solana or Ethereum L2. instant settlement fractions of a penny. Meanwhile, the shock isn't just American. The piece imagines India's IT outsourcing giants, TCS, InfoS, YPro getting hit as contract cancellations accelerate,

the rupee sliding hard, and even preliminary discussions with the IMF for a bailout by early 2008. Then the macro crisis becomes obvious in the U.S. Displaced white-collar workers downshift into lower-paid service and gig jobs, pushing wages down there too. One anecdote has a sales Salesforce product manager going from $180,000 a year to $45,000 driving Uber. By Q2, 2027, the U.S. is in recession. And then comes the daisy chain. Private credit has ballooned over $2.5 trillion by 2026 in the scenario, stuffed with PE-backed software deals price for perpetual recurring revenue, right as AI disruption makes that premise laughable. Marks to market on portfolios like reality, downgrades hit, insurers get forced to reassess capital treatment, and the whole offshore

reinsurance rating structured debt industrial complex turns out to be opaque at the exact wrong time. Mortgages start to wobble, not because rates are too high, but because the income engine underpinning mortgages is suddenly suspect when intelligence is cheap. By the end, the authors call it an intelligence premium unwind, because for most of human modern history, human intelligence was the scarce input, and our institutions were built around that assumption. In this scenario, the policy world scrambles talk of a transition economy act with direct transfers funded by deficits and a tax on AI inference or even a more radical public claim on the returns of intelligence infrastructure itself, while Wall Street gets populist, occupies Silicon Valley blockades at major labs.

The final part of the piece is a warning. You're not reading this in 2008. You're reading it now. Today in 2026, when the canary and the coal mine is still alive. But the clock is. The clock is ticking. Provocative stuff. Worth reading the whole thing if you have time. This has officially been the winteriest winter I've ever had since I've lived in New York City. If you've heard people shoveling snow in the background of today's show, there was no way for me to avoid it, really. I've had winters like this when I lived in Michigan where you get a big snowstorm and then it doesn't go above freezing for like a month or so, so the snow never really goes away and it's gross and and everything is gross. The snow had almost gone away here, but now we've gotten another snowstorm,

which is maybe the biggest I've seen since the 2004 blizzard. So yeah, for the 26 years, I've lived north of the Mason-Dixon line. I've always still gotten a thrill when snow is in the offing, but this winter has broken this Florida boy is all I can tell you. Talk to you tomorrow. Some follow the noise. Bloomberg follows the money. Because behind every head, headline is a bottom line. Whether it's the funds fueling AI or crypto's trillion dollar swings, there's a money side to every story. And when you see the money side, you understand what others miss. Get the money side of the story. Subscribe now at Bloomberg.com.