Monday, 26 October 2026
Apollo's Chief Economist: AI Has Delivered 'Zero Financial Payoff' to Companies Outside Tech
Apollo’s Chief Economist: AI Has Delivered ‘Zero Financial Payoff’ to Companies Outside Tech
Torsten Slok, chief economist at the private-equity firm Apollo Global Management, took the S&P 500, removed Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia and Tesla, and looked at what was left — the “S&P 493.” Profit margins: flat. Revenue per worker: flat. No sign, he writes in a note TechCrunch covered on Sunday, that AI is making these companies more productive. His verdict is blunt — the technology “has delivered zero financial payoff to corporate America outside of tech” — and so is his explanation: “AI may be a great calculator, but in business, intelligence is mostly judgment.” AI, he argues, does not understand “accountability, trust, culture, unwritten rules, ethical trade-offs or fairness.” It is one note and one cut of the market, and leaving out the seven biggest spenders shapes the answer by design. But it lands in a week when Anthropic’s leaked filing showed $9 billion spent on compute in six months. The money is going in on one side of the index. The question is when it comes out on the other.
Source: TechCrunch, October 25, 2026
1,946 OpenAI Agents Read a Mouse-Brain Atlas in 32 Minutes
OpenAI, the startup PerturbAI and Scripps Research pointed 1,946 GPT-Rosalind agents at an existing whole-brain CRISPR atlas — data from millions of mouse brain cells — and gave each agent one gene disruption to explain. They looked at how the cells responded, searched the literature and wrote up hypotheses, all in parallel, in 32 minutes. Without seeing the original study’s manuscript, they matched or refined 48 of its 55 findings that the data could support. Thirty doctoral-level biologists then graded the reports and broadly agreed with the agents’ ranking of which leads were worth chasing; one, a link between the neurodegeneration gene DCTN1 and cholesterol production in neurons, scored 4.3 out of 5 for follow-up. The caveats come from the team itself: the “months to a year” a human would need is an estimate built on assumed minutes per task, the preprint is not peer reviewed, and nothing has yet been tested at the bench. Two thousand readers, thirty-two minutes. The experiments still take as long as they always did.
Source: R&D World, October 21, 2026
TechCrunch Lists 25 AI Agents That Live in Your Text Messages
The agent you text like a friend is now a crowded category. TechCrunch’s Saturday guide counts 25 of them, reachable over SMS, iMessage, WhatsApp or Telegram. Some do the small stuff — reminders, calendars, nightly family summaries. Others act: Instinct, which raised $1 billion at a $10 billion valuation in September and is still in private beta, plans trips, buys groceries and cancels subscriptions; Wajo’s agent Fo has its own email, phone number and payment card and calls businesses for you. Prices run from free to about $30 a month, and one charges $1 a day. Poke became the first AI agent on Apple’s Messages for Business in June, and its parent was bought by Cognition a month later. TechCrunch flags privacy and security concerns around agents that act for you — the same week US banks began shutting third-party agents out of customer accounts. The agents now have phone numbers. Nobody has yet said who answers when one of them gets it wrong.
Source: TechCrunch, October 24, 2026