Big Tech $755B AI Spend Cuts Shareholder Returns

Samantha: Welcome to The State of Tech, The European Edition, Monday May eleven, 2026. I'm Samantha Lawrence.

Bob: And I'm Bob Russell. Today: Alphabet on the verge of becoming the world's most valuable company on the back of AI, Big Tech's record-breaking AI spending crushing shareholder payouts, Apple's quarter-billion-dollar settlement over Siri features that never arrived, an unmanned Japanese lab where robots run medical experiments around the clock, an Egyptian-built autonomous harvesting robot you can read about today, and Faraday Future's pivot into physical AI robotics. Let's start with Alphabet.

Alphabet poised to overtake Nvidia as the world's most valuable company on AI dominance.

Samantha: The leaderboard at the very top of global tech is about to flip. Alphabet is reportedly on the cusp of passing Nvidia to become the most valuable company in the world. After a year of strategic moves, Google's parent is now seen as the single biggest beneficiary of the AI boom, with investors pointing to its full-stack offering as the reason.

Bob: And it really is full-stack. Alphabet has its own silicon with the TPU chips, its own foundation models in the Gemini family, the cloud infrastructure underneath, the consumer products on top — Search, YouTube, Workspace, Android — plus DeepMind doing the research. Nvidia sells the picks and shovels. Alphabet runs the entire mine.

Samantha: That vertical integration is what's changing the narrative. For most of the past two years the story was: whoever buys the most Nvidia chips wins. Now the story is shifting to: whoever owns the model, the data, the distribution, and the chips wins. And Alphabet ticks all four boxes.

Bob: There's a financial angle too. Nvidia's revenue depends heavily on hyperscalers continuing to spend. Alphabet's revenue base is much broader — advertising, cloud, subscriptions. So in a downturn, Alphabet is seen as more defensive while still riding the AI upside.

Samantha: For European listeners this matters on several levels. First, competition. The European Commission has been investigating Google for years on Search and advertising. If Alphabet's lead extends into AI infrastructure too, expect Brussels to sharpen its tools, possibly under the Digital Markets Act.

Bob: Second, dependency. European businesses, governments, and universities increasingly run on Google Cloud and Gemini APIs. The more dominant Alphabet becomes, the louder the conversation around digital sovereignty gets — and the more attention flows to European alternatives like Mistral or Aleph Alpha.

Samantha: Third, pension funds. Most European retirement portfolios hold significant Alphabet exposure through global index funds. A jump to number one in market cap is good news for those balances, but it also concentrates risk in a small handful of American tech names.

Bob: And finally, regulators will watch how Alphabet behaves at the top. There's a pattern where companies become more cautious about their public moves once they wear the crown. Whether Alphabet stays aggressive on AI or starts playing defence — that shapes how the rest of the industry moves.

Big Tech to spend seven hundred fifty-five billion on AI, cutting buybacks by two-thirds.

Bob: Staying with the money. Goldman Sachs has put a number on what the AI race is costing shareholders. The big five hyperscalers — Amazon, Alphabet, Meta, Microsoft, and Oracle — are projected to spend seven hundred fifty-five billion dollars on capital expenditure this year. That's an eighty-three percent increase. And it's coming straight out of shareholder pockets.

Samantha: Stock buybacks in the first quarter were cut by nearly two-thirds. That's an enormous shift in capital allocation. For years these companies were essentially returning cash to shareholders. Now they're pouring it into data centres, GPUs, power contracts, and cooling systems.

Bob: And some of them are borrowing on top of it. Meta and Oracle have been tapping bond markets to fund expansion. That changes the risk profile of these stocks. They're no longer cash-flow machines paying you to wait — they're becoming infrastructure plays with debt on the balance sheet.

Samantha: The question nobody can answer yet is whether the returns will justify the spend. Seven hundred fifty-five billion dollars is roughly the GDP of the Netherlands. If AI revenue doesn't scale to match, we're looking at one of the largest capital misallocations in corporate history.

Bob: For European investors there are two angles. One: many European pension funds are heavily exposed to these names. Reduced buybacks mean lower per-share earnings growth, which directly affects long-term returns.

Samantha: Two: European utilities and grid operators are quietly benefiting. All those data centres need power, and increasingly that demand is landing in Ireland, the Nordics, and the Netherlands. It's reshaping electricity markets and forcing new conversations about grid capacity.

Bob: And there's a regulatory edge. The European Central Bank has flagged AI infrastructure spending as a potential source of market concentration risk. If a downturn hits and these companies have to write down assets, the shock waves reach European balance sheets fast.

Samantha: Quick interruption. If you listen to The State of Tech regularly, hit that like button and subscribe, that way you'll never miss an episode. Okay, moving on.

Apple agrees to two hundred fifty million dollar settlement over undelivered Siri AI features.

Bob: Apple has agreed to a proposed two hundred fifty million dollar class-action settlement over Siri features that were promised but never shipped. At the June 2024 Worldwide Developers Conference, and again in the iPhone 16 marketing, Apple showcased advanced AI capabilities — personal context awareness, in-app controls, a smarter Siri. Customers bought the phones. The features never arrived.

Samantha: And they still haven't. That's the part that stings. We're nearly two years on from that keynote, and the showpiece AI features remain missing in action. Apple has quietly pushed timelines, restructured its AI team, and the lawsuit argues that consumers paid premium prices for capabilities that simply weren't real.

Bob: Two hundred fifty million dollars sounds like a lot, but for Apple it's barely a rounding error. The reputational damage is the bigger cost. Apple's brand was built on the promise that things just work. With Siri, that promise broke very publicly.

Samantha: It also sets a precedent. Every tech company has been demoing aspirational AI features over the past two years — slick keynote videos showing capabilities that don't yet exist in shipping products. This settlement says: you can be sued for that. Marketing departments across the industry will be rewriting their scripts.

Bob: The European angle here is sharp. The EU AI Act has explicit transparency requirements. Companies have to be clear about what their AI can and cannot do. And consumer protection regulators in countries like Germany, France, and Italy have been increasingly willing to act on misleading advertising.

Samantha: If a similar case were brought in Europe, the outcome could be even tougher. The Unfair Commercial Practices Directive gives national regulators real teeth, and class-action mechanisms have been expanding under the Representative Actions Directive. Apple may have settled in the US, but the European exposure is still open.

Bob: And for European consumers it's a useful reminder. When you see a feature in a keynote, check whether it's actually shipping in the version sold in your country. Apple's AI rollout has been particularly slow in the EU because of Digital Markets Act concerns. Some features that exist elsewhere still aren't available here at all.

Japan opens unmanned lab where ten robots run medical experiments around the clock.

Samantha: Over to Tokyo. The Institute of Science Tokyo has opened the Robotics Innovation Center — an unmanned laboratory where robots conduct medical experiments that used to require human researchers. Ten robots are already at work, including a humanoid called Maholo LabDroid, handling reagent transfers, cell cultivation, and the kind of delicate pipetting that takes years of training.

Bob: The ambition is the striking part. The university plans to scale to around two thousand robots by 2040. They want to automate the entire research pipeline — from hypothesis generation through to experimental verification — with AI driving the science and robots executing the lab work.

Samantha: And there are real reasons this matters beyond efficiency. Lab work is notoriously hard to reproduce. Different researchers, different hands, different days produce different results. A robot doing the same protocol identically every time eliminates a huge source of noise in biomedical research.

Bob: It also runs twenty-four-seven. A human researcher works maybe eight productive hours a day. These robots work all of them. That compresses research timelines dramatically, particularly for drug discovery, where iteration speed is everything.

Samantha: For Europe this is both inspiration and warning. European research institutions are world-class in biomedicine — think Karolinska, Max Planck, the Crick Institute. But they typically run on human researchers and graduate students. Japan is making a generational bet that the future of science is automated.

Bob: If that bet pays off, European labs that stick with traditional methods will fall behind on pure speed of discovery. There's already a quiet movement in places like ETH Zurich and TU Delft toward self-driving laboratories, but nothing at the scale Tokyo is announcing.

Samantha: There's also an ethics conversation. Who's accountable when an AI generates a hypothesis, a robot runs the experiment, and a published paper results? Peer review assumes human judgement at every step. Europe's research integrity frameworks will need updating, and probably soon.

Egrobots launches the Arab world's first fully autonomous agricultural harvesting robot.

Bob: Time for something you can actually look up and explore today. Egrobots, an Egyptian deep tech firm, has unveiled the first autonomous agricultural harvesting robot fully developed by Egyptian engineers. It uses computer vision and AI to spot ripe crops, plans its own harvesting routes, and picks at a rate of around one hundred sixty kilograms per hour — twenty-four hours a day if needed.

Samantha: What makes this worth a listener's attention isn't just the robot itself, it's the broader signal. Agricultural robotics has been dominated by American, Dutch, and Israeli companies. An Egyptian startup entering the market with home-grown engineering changes the geography of AgriTech, and their website and demo videos are publicly accessible.

Bob: If you're in European agriculture, or you're curious about where farm robotics is heading, Egrobots is worth a search. Their videos show the system in real field conditions, picking soft produce — the hardest task in harvesting robotics, because you can't bruise the fruit.

Samantha: One hundred sixty kilograms an hour is a serious productivity number. A skilled human picker manages roughly forty to sixty kilograms in optimal conditions. So you're looking at a three-to-four-times productivity jump, with no breaks, no overtime, and no seasonal labour shortages.

Bob: For European farmers — especially in Spain, Italy, the Netherlands, and France — labour shortages have become an existential problem. Post-Brexit Britain has felt it acutely. Strawberry growers, asparagus farms, vineyard operators are all asking the same question: where does the next generation of pickers come from?

Samantha: The practical thing you can do today: if you work in food, agriculture, logistics, or you just find this stuff fascinating, look up Egrobots, watch the demo footage, and follow the company. It's a real-world example of how AI computer vision is leaving the lab and landing in actual fields.

Bob: And it's a useful counterweight to the narrative that all serious AI comes from California. There's a global wave building, and watching where the next breakthroughs come from — Cairo, Nairobi, Bangalore — is genuinely interesting.

Faraday Future pivots to physical AI with humanoid robots and a 2026 shipping target.

Samantha: And to close out, Faraday Future. The American EV maker has just announced it's pivoting into what it calls a Physical AI ecosystem company. Founder YT Jia is back as Global CEO, and the company is focusing on two product lines: humanoid and bionic robots, and what they call automotive robots.

Bob: They've also shared concrete numbers. Sixty-eight robots shipped as of April thirtieth, with a target of over one thousand cumulative shipments by the end of 2026. That's small, but it's real product moving out the door.

Samantha: For a listener, what's interesting here is the broader shift it represents. Embodied AI — putting language models and vision systems into physical bodies that move around the world — is moving from research labs to commercial products. Figure, Agility Robotics, Apptronik, Unitree, and now Faraday are all racing to ship.

Bob: If you want to see what physical AI actually looks like, the demo videos these companies post on YouTube are worth twenty minutes of your afternoon. Watch a humanoid robot fold laundry, sort packages, or pour a drink. It's wildly more advanced than it was even twelve months ago.

Samantha: For European industry the implication is real. German and French automotive suppliers, Italian industrial automation firms, Dutch logistics operators — they're all going to face decisions soon about whether to integrate humanoid robots into their operations, and where to buy them from.

Bob: And today we covered: Alphabet about to top the global market cap rankings, Big Tech's massive AI spending squeezing shareholder returns, Apple's Siri settlement, Japan's robot-run laboratory, Egypt's autonomous harvesting robot, and Faraday Future's pivot to physical AI.

Samantha: Want to know more or react? Visit stateoftech.eu or email us at info@doorzetters.net.

Bob: State of Tech, the tech world in 15 minutes.