Welcome back to the Neural Net!
In today’s edition: We explore the risks and rewards of the AI bubble and, apropos of nothing, highlight what OpenAI’s been up to. Pure coincidence.
▼
The Street
note: stock data as of last market close
▼
🫧 Bezos Thinks We’re in a Bubble, but the Good Kind
It’s official: even Jeff Bezos thinks we’re in a bubble.
At Italian Tech Week, he said that in bubbles, money flows fast and anything with “AI” gets funded. In moments of collective excitement, investors struggle to tell which ideas will last and which will burn out, a dynamic he says is “probably happening today” in AI.
Meta’s Mark Zuckerberg seems to be guilty of this, saying he’d rather “misspend a couple hundred billion dollars” than show up late to superintelligence.
OpenAI’s Sam Altman agrees. Back in August, he acknowledged the frenzy, calling the AI market a bubble and noting that “smart people often get a little too excited about a kernel of truth”, much like during the dot-com era.
Risky Business
Investors are hoping the dizzying sums of money they’re pouring into the tech pays off in a big way, but these investments carry significant risk. Here are a few that could cause the bubble to pop:
It Never Works — AGI has been “five years away” since 1970. If it never materializes, those trillion-dollar data centers go to zero.
It Takes Too Long — Chips age faster than milk. If AGI’s payoff is more than 5 years out, today’s hardware may be obsolete before it earns a dime.
It Costs Too Much — OpenAI’s $500 billion valuation equals ~38X revenue, more than dot-com bust Cisco at its peak.
You Pick a Loser — Remember Netscape vs IE? Neither won. The “AI winner” could be anybody, even somebody not yet around. Choose wisely!
Too Much Competition — A dozen chatbots, same features, same costs. Margins could collapse before profits appear.
The Real-World Strain — Data-center buildouts are sucking up power, metals, and machinery—potentially crowding out the rest of the economy and fueling inflation.
The Silver Lining: Bubbles That Build
So what if we are in a bubble, is that such a bad thing? Bezos sees this frenzy as an industrial bubble, one that actually produces useful infrastructure.
“This is not a financial bubble,” he said. “Industrial bubbles can even be good, because when the dust settles and you see who are the winners, societies benefit.”
▼
In Partnership With Kojo
You don’t need thousands of ads, just better ones.
Most platforms push you to make thousands of ads. Kojo helps you make better ones. We turn your social data into proven ideas, predict which will perform, and send them to real human creators in seconds. Less waste, more certainty, and ads that actually work.
▼
Heard in the Server Room
OpenAI just sent AMD stock soaring 23% after announcing a deal that could give Sam Altman’s company up to a 10% stake in the chipmaker. The agreement includes a rollout of 6 gigawatts of AMD GPUs, one of the largest GPU deployments to date, to help OpenAI overcome the computing shortages slowing new product launches. It’s a huge win for AMD, which finally gets a seat at the AI hardware table long dominated by Nvidia. But it also shows how tightly the AI boom is looping in on itself, with the same handful of companies swapping chips, equity, and billions in cash to keep the frenzy going.
OpenAI and Jony Ive are reportedly struggling to design their screenless AI gadget, which was supposed to feel futuristic but currently sounds more like a nosy roommate. The Financial Times reports that the palm-sized device listens and watches its surroundings, yet the team can’t quite stop it from talking when it shouldn’t or staying silent when it should. Questions around privacy, personality, and basic usefulness have delayed the 2026 launch. Turns out even for Jony Ive, figuring out how to make an AI helpful without being creepy might be harder than designing the iPhone.
Continuing on the OpenAI front (because three headlines a week keeps Sam Altman employed), the company revived their Dev Day event for the first time since its 2023 edition. The day was equal parts ambition and agenda: new models, developer toolkits, and big hints that the company still sees devices as the next AI frontier. Sam Altman and Jony Ive both showed up, underlining how hardware and software are being woven ever more tightly in OpenAI’s roadmap. We’ll dig into the announcements, the implications, and what they mean for your stack in the next edition, so stay tuned!
▼
In Partnership With Intercom
Startups get Intercom 90% off and Fin AI agent free for 1 year
Join Intercom’s Startup Program to receive a 90% discount, plus Fin free for 1 year.
Get a direct line to your customers with the only complete AI-first customer service solution.
It’s like having a full-time human support agent free for an entire year.
▼
Chasing AI Dreams: Why Europes’s Heading West
For Europe’s AI founders, the road to Silicon Valley is starting to look like a one-way ticket. Despite Europe’s ambitions to become a global AI hub, its startups are increasingly packing their laptops and heading west. The reason? Money moves faster (and bigger) in America. One London founder spent months chasing funds at home, then landed $500K within a week of arriving in San Francisco.
By the numbers, why the West wins:
💸 $160B vs. $20B: U.S. AI and machine-learning startups raised about eight times more than their European counterparts in 2025.
🇺🇸 71% of Europe’s AI funding came from U.S. investors this year, up from 57.5% in 2024, showing just how American-led the “European” AI boom really is.
⏱️ Weeks, not months: U.S. venture deals often close in a matter of weeks, while European founders report multi-month due-diligence slogs.
💰 Valuations double across the Atlantic: Startups like FLock.io say U.S. investors valued them at ~2× higher than European VCs.
As long as capital, ambition, and now talent keep flowing west, Europe’s dream of leading in AI may remain just that—a dream.
▼
That’s it for today! Have a great week, we’ll be back Friday.