Daily Brief Nov 1, 2026
How we checked the items below
For readers who want to dig in. What happened, in brief, is on the Today tab.
This issue draws on material from October 31 to early November 1; the Intel item comes from the October 29 earnings report, and the From the archive item comes from a SecondSource call of July 31. Last night's collection took in 220 pieces, and this issue uses 20 outside sources with checkable links; the later sections also draw on material from the past week.
1. [Today] OpenAI's target reportedly falls from US$30 billion to US$20 billion, and the US$1.25 trillion valuation was in print eight days before that report
Not proven yet： the first close's price and terms haven't been made public
How we checked
The US$20 billion and the US$11 billion each come from one newspaper. The above is TechCrunch's roundup of those two reports, not two independent sources confirming each other. The first close's price and terms haven't been made public. The US$1.25 trillion appears in CNBC, citing anonymous sources, which states that it excludes new money. The Financial Times' US$1.25 trillion is relayed on X by Rui Ma, an investment analyst with Tech Buzz China and Alpha Girl, and the Kobeissi Letter; those posts don't say whether it includes new money, and whether it is independent of CNBC can't be told from them. The US$1.4 trillion may be only a September press figure, and whether it includes new money isn't stated. OpenAI didn't respond to TechCrunch's request for comment.
[This week] (October 29 earnings) Intel's data-center segment grew 78% year on year in the third quarter, and fourth-quarter total revenue guidance of US$15 billion to US$16 billion is still below the third quarter's US$16.5 billion.
How we checked
(Intel 8-K, 2026-10-29) (third-party call transcript, 2026-10-29) (AINews, 2026-10-30)
[Today] Azeem Azhar, author of Exponential View, rebuts McKinsey: data centers compete on how fast they can get power, not on the price of power.
How we checked
The post doesn't say how those figures were calculated (McKinsey, 2026-10-31).
[This week] (preprint, arXiv 2610.39641) Two authors at Samyama AI, which builds research knowledge graphs, hand-checked 41 model rulings and found 9 fully correct.
How we checked
[This week] (preprint, arXiv 2610.39570) Leo Itoh and Junichiro Kadomoto self-report that an agent testing the same CPU design on an FPGA finished 8,000 CoreMark iterations, build time included, 9.6 times faster than single-threaded RTL software simulation.
How we checked
The result is self-reported and comes from one human-supervised single-core design.
[This week] According to the paper's abstract, Amazon's sandbox service Sandman, built for training tool-using agents, created 3.1 million isolated execution environments (sandboxes) across four training campaigns in October.
How we checked
[Looking back] (SecondSource call, July 31) SecondSource put the odds at about 85% that Claude Sonnet 5 would return to its original price after the promotion ended, and the price didn't go back.
How we checked
This was the highest estimate in that set of calls, and the mistake was reading an expiry date as evidence of pricing power.
(card)
Sources & accounting
The past 24 hours. 220 new pieces came in from October 31 to November 1: 112 academic papers, 80 social-platform posts, 12 from a second academic-paper feed (counted separately from the academic papers), 6 linked articles, 5 company and personal blog posts, 3 subscription newsletters and 2 show transcripts. Of those 220, 60 were read in full and 89 were screened out as off-topic (8 of them before fetching). Another 2 were logged with a stated reason for not reading closely, and 69 haven't been read yet.