SoftBank sheds 10% as the AI slowdown call hits the tape

Share
SoftBank sheds 10% as the AI slowdown call hits the tape

The pacing debate spent Saturday as an argument between CEOs. Monday, it became a price.

The slowdown call produced its first global sell-off, and the damage spread across three continents. South Korea's SK Hynix closed down more than 6% and Samsung Electronics more than 4%, while SoftBank — one of OpenAI's largest backers — fell 10% in Tokyo. Europe continued the slide: ASML dropped more than 4%, Infineon more than 6%, Nokia around 5%, with data-center suppliers Siemens Energy and Schneider Electric also lower. In U.S. premarket trading, Intel fell nearly 6%, Micron around 5%, and Nvidia more than 2%. The trigger chain runs from Anthropic researcher Jacob Coxon's resignation over "gambling with our lives", through Dario Amodei's Saturday essay demanding the industry "slow the pace", to the rare public agreement from Sam Altman and Elon Musk — the first time the pacing camp spoke with one voice, and the first time markets priced it. RBC Brewin Dolphin's Zoe Gillespie put the fear plainly: the equity rally is built on AI growth and productivity gains, and if that derails, "we may see this destabilize." Earlier this morning the story was a Tokyo wobble with oil as an alibi — see Nikkei breaks 63,000 as the AI slowdown debate hits the market — but once Europe and U.S. futures joined with the same names falling, the read changed: investors now model "pace the frontier" as a valuation risk, not a safety footnote.


Beijing answered the slowdown call, and its answer was: keep us out of it. Foreign Ministry spokesperson Guo Jiakun said Monday that spreading "various threat narratives", confrontation and malicious competition only disrupt global AI governance and "serve the interests of neither side" — a direct rebuttal of the Amodei framing, and pointed given the essay cites China's computing scale as part of the race dynamic. The same weekend, Trump dismissed the pacing push in Ireland as hype from "very negative forces" and repeated that "whoever wins AI wins everything." The practical result: the two governments the slowdown debate is nominally about have both rejected it, leaving a voluntary pact among three American labs with no counterpart abroad and no constituency at home.


Nvidia put up a product page for an 84GB Blackwell card, and the local-inference crowd treated it as a release. The RTX PRO 5500 Blackwell Workstation Edition carries 84 GB of GDDR7 with ECC at a quoted 1,398 GB/s, 21,760 CUDA cores on the GB202 die, a 600 W power limit, MIG partitioning into one 84 GB or two isolated 42 GB instances, and PCIe 5.0 — sold for rack-mounted workstation fleets, with a liquid-cooled variant promised later and pricing undisclosed; Nvidia's page still says "Coming Soon". One caveat from the spec sheet: the listed 416-bit memory bus doesn't square with 84 GB at 1,398 GB/s, which arithmetically implies a 448-bit interface. For anyone running models locally, the capacity is the point — 84 GB fits a heavily quantized 100B-plus model, or two 42 GB tenants on MIG, on a single card roughly 12 GB below the flagship RTX PRO 6000.

What to watch: the U.S. open — whether the decline survives contact with Wall Street's dip-buyers, and whether the pacing camp floats anything concrete enough to trade against this week.

If your AI holdings fell 10% because three CEOs wrote blog posts, are you selling, holding, or buying the dip? Tell us in the comments.

Read more

Akhetonics says its all-optical CPU reaches a customer in 2026

Akhetonics says its all-optical CPU reaches a customer in 2026

Light-based computing keeps promising more than it delivers — but one Munich startup has just put a date on its bet, and the interview laying it out is doing the rounds on Hacker News this week. Akhetonics says it will deploy its first commercial machine with a major customer by the end of 2026, with several more planned for 2027. The company, founded by Michael Kissner and Leonardo Del Bino, is building a computer where data enters as light, is switched as light, and circulates through memory

The Week in AI — October 5–11, 2026

The Week in AI — October 5–11, 2026

Every big claim this week turned out to rest on fine print more interesting than the headline: revenue only the company reporting it can define, safety tests sandboxed while the product keeps the web, and a Pentagon phase-out nobody would confirm until reporters kept asking. The week's top 5 1. OpenAI's revenue was $20 billion below the numbers everyone quoted — and the gap was definitional. The Financial Times reported Thursday that OpenAI's annualized revenue runs roughly $20 billion unde

Drone strike shuts a third Yandex data center, taking YandexGPT offline

Drone strike shuts a third Yandex data center, taking YandexGPT offline

Russia's largest tech company is learning what the AI era's infrastructure war looks like from the receiving end — three data centers in four days, and with them much of the cloud layer Russian businesses run on. A Ukrainian drone strike knocked out Yandex's data center in Vladimir early Sunday morning, the third of the company's facilities hit since October 8. The site — reported at roughly 50 MW and designed for about 2,880 server racks — stopped operating completely after the attack, Yandex

Agent teams cost up to 5x more, barely score higher

Agent teams cost up to 5x more, barely score higher

The multi-agent hype train hit a benchmark this weekend — and the grid and the trucking regulators had quiet weeks of their own. Vals AI put agent teams head-to-head with single agents on its Vibe Code Bench, and the teams cost between 1.8 and 5.1 times more for almost no extra quality. The evals company ran GPT-6 Sol and Claude Opus 5.5 solo and in teams across 50 apps at two reasoning efforts; out of four comparisons, only one was statistically significant — Sol at medium effort, where the t