LLM backlog
2026-05-27 17:52:16.185771+02 by Dan Lyke 0 comments
J.P. Morgan, modeling what the buildout would need to earn to clear a ten percent return on current capital expenditure, arrived at roughly six hundred fifty billion dollars per year in AI-sector revenue — the equivalent of thirty-five dollars per month, in perpetuity, from every iPhone user on earth. The current run-rate is about twenty-five billion. The gap is twenty-six-fold.
Via.
Fortune: Uber burned through its entire 2026 AI budget in four months. Now its COO is questioning whether it’s worth it (paywalled), summarized by Dare Obasanjo @carnage4life@mas.to:
Uber's COO recently voiced an uncomfortable truth that many in the tech industry are yet to admit.
He said while Uber's engineers are consuming massive amounts of AI tokens, it is incredibly difficult to prove that this spending actually benefits the end user or improves the product proportionally.
Elizabeth Ayer @elizayer@mastodon.social has some screen shots of Google Gemini A/B testing:
Google is user-testing a yes vs no answer to "does X increase the risk of Y"???
I'm sure this is a hard problem in risk communication... which is exactly why A/B testing is a poor choice.
Entrepreneur: LinkedIn Is Fighting Back Against AI Slop — and AI Comments, via root @ segfault cult @segfaultcult.com
i've tasked ai with figuring out what to do with linkedin.
somehow we came to the same conclusion - just shut it down.
Pavel Samsonov: AI is not the future of software development, but the last dying gasp of the past:
If you skipped over Garris’s and Bjarnason’s essays, I strongly recommend going back to read them — even the uncomfortable parts — and sit with their conclusions for a bit. Because what they add up to is that the “golden age” of tech was not in 2014, or even in 2005. By the time Design Thinking, Agile development, DevOps, etc came on to the scene, it was already too late: practitioners had given up the picturing relation that gave them real leverage.