November 16, 2026
Stories from this day
-
Tech · 14 min readAnthropic IPO price above $78 — a 20% rule lets bankers go higher without a new amendment
The rule — a change of up to 20% in the maximum aggregate offering price goes into the final prospectus with no new amendment. At the same share count that is $93.60 a share
-
Tech · 13 min readChatGPT Wallet — the agent pays without seeing your card, and only US Pro users get it
What it is — a dot shops, a human approves every purchase with a biometric check, and a network-issued token pays on the merchant's own site (announced November 13, 2026)
-
Tech · 15 min readEdge0-145B specs — 145 billion parameters in 29GB? The download is 106GB
The arithmetic — fitting all 145 billion parameters in 29GB would take 1.6 bits each. The actual files are a 40GB expert pool plus a 51GB embedding table, about 106GB in total
-
Tech · 11 min readMicrosoft Oracle deal wind-down — three-year-old Nvidia chips go back, $80 billion in rentals reviewed
Reported — Microsoft did not renew its 2023 Oracle server rental for Bing and has been switching it off for months; the reason given is older Nvidia chips (Bloomberg, November 6, citing people familiar)
-
Economy · 14 min readWhy Korea's export boom isn't reaching consumers — chip jobs are 1.9% of employment, 10.6% in Taiwan
The three reasons — IT manufacturing is 1.9% of Korean jobs (10.6% in Taiwan), gains flow to high earners, and retail stock trading cooled after the market correction
-
Tech · 14 min readSamsung AI data center push — OpenAI and Anthropic talks are 'early', the 2031 revenue figure a broker's estimate
Reported — five Samsung affiliates would bundle construction, operation, batteries, chips, cooling and floating facilities into one AI data center offering (Financial Times, November 15, 2026)
-
Economy · 18 min readNvidia earnings date: November 17, 2026 — first call without Huang, $108 billion bar
Schedule — release at about 1:20 p.m. Pacific on November 17, 2026, call at 2 p.m. Pacific; 6:20 a.m. and 7 a.m. on November 18 in Seoul
-
Tech · 20 min readGemini 4 Hydrogen pricing — free, but you cannot pick it; the API costs twice 3.8 Flash
Price — $1.50 input and $7.50 output per million tokens, $0.075 cached, batch at half. Twice the current 3.8 Flash rate
-
Tech · 16 min readCVE-2026-126253 FortiOS patch versions — 7.6.8 and 7.4.13 shipped 19 days before the advisory
Versions — FortiOS 8.0.2, 7.6.8, 7.4.13, 7.2.14 or later. 7.0 has no fix and must migrate. Builds shipped October 22; the advisory followed on November 10
-
Tech · 19 min readWhat is llama.cpp — the engine under Ollama and LM Studio, and what 1.0 promises
Meaning — a C/C++ engine that runs AI models stored as GGUF files on CPUs and GPUs; Ollama and LM Studio run on it