We are tracking a breach of a 3rd-party JS host impacting hundreds of high value websites. This is staggering.
DSA Library
News
What people in tech are talking about on X, by day.
Monday, November 16
JPMorganChase has experienced a temporary issue related to a software bug in on-premise hardware systems – there is no evidence this is cyber-related. It is taking longer than usual to service some customers. Where we were party to a transaction involving other banks,…
The NY Times has the receipts today. Here's how the Nvidia Blackwell chip laundering worked. 1. Nvidia was blocked from selling Blackwell chips to China. So Jensen Huang sold them to Alice Huang at Megaspeed which installed them in data centers in Singapore, Malaysia,…
"Do you want to see Jensen Huang reinstated as CEO?" @POTUS: "I would certainly not mind seeing that."
OpenAI @OpenAI Announces embedded assessments with Apollo Research, METR, Transluce and Goodfire, giving outside experts deep access to its internal models and systems to investigate alignment and test safeguards.
Gergely Orosz @GergelyOrosz Relays a Microsoft team lead’s account of the outage: a latent bug from a commit made to a Windows build about four months earlier that only activated much later, in code that involved no AI.
Arin @arinwaichulis Reports that the macOS 27.3 beta changes Full Disk Access: users can limit an app to specific folders, see clearer warnings about what is handed over, and limit which apps a tool can control.
Seema Mody @seemacnbc Reports, citing a source familiar, that Nvidia’s board is not considering bringing Jensen Huang back as CEO and is focused on its investigation.
Chris McGuire @ChrisRMcGuire Reports that the Commerce Department published new AI and semiconductor export control rules that counter Chinese smuggling of AI chips and close loopholes in the existing controls.
Rob Bonta @AGRobBonta California’s attorney general announces a lawsuit against OpenAI over testing and developing dangerous AI models without adequate safeguards.
Sunday, November 15
Elon Musk @elonmusk Disputes the argument that money alone could produce rockets like Starship, Falcon or Dragon, pointing to SpaceX catching Starships returning from orbit with the tower’s chopsticks.
Toby Rogers @uTobian Notes that Nvidia’s bylaws let shareholders holding at least 15% of the voting power call a special meeting and replace the board, and predicts Jensen Huang would win that vote.
ULA @ulalaunch Confirms that the Vulcan rocket suffered an anomaly about 10 minutes into flight and failed to deliver the Amazon Leo mission to orbit.
Mark Gurman @markgurman Reports that Apple wants to unveil its smart glasses at WWDC 2027 and ship them before year-end, but software is running behind while the hardware is largely finalized.
George Conway @gtconway3d Quotes a report on A201, a Malaysian data center opening next year with hundreds of millions of dollars of Nvidia’s most advanced AI servers, whose computing power is planned for a Chinese tech giant.
Scott Manley @DJSnM Notes that if the Vulcan booster’s underperformance was a BE-4 engine issue, it could also affect New Glenn’s return to flight.
Annmarie Hordern @annmarie Quotes Scott Bessent on Fox asking whether Nvidia’s culture pressed seemingly illicit sales to lift its revenue and earnings.
Artificial Analysis @ArtificialAnlys Reports that Alibaba’s Qwen4 Max scores 43 on the Artificial Analysis Intelligence Index, 9 points above Qwen3.8 Max and 3 behind Kimi K3.1 Pro at about 11% lower cost per task.
Andrew Curran @AndrewCurran_ Relays the FT’s report that Jensen Huang is assembling a war room of tech executives who back his reinstatement as Nvidia CEO.
Mark Gurman @markgurman Reports that Apple’s smart glasses, planned for next year, will lead with Siri AI app integrations, such as booking an Uber or ordering coffee through apps on a paired iPhone.
Saturday, November 14
Elon Musk @elonmusk Says many of Nvidia’s critical supply chain deals were handshake agreements with Jensen Huang rather than written contracts, and some will be lost without him.
The Kobeissi Letter @KobeissiLetter Reports that the petition to reinstate Jensen Huang as Nvidia CEO passed 100,000 signatures, including more than 53,000 shareholders and 6,200 current and former employees.
Aaron Rupar @atrupar Shows Sean Duffy, asked why his air traffic control funding request rose from $30 billion last week to $50 billion, answering that he wants a number Congress will approve.
Ben Norton @BenjaminNorton Cites the Washington Post’s report that Marco Rubio was one of the key figures behind Jensen Huang’s ouster from Nvidia.
Nathan Lambert @natolambert Names two challenges for open-model builders after Jensen Huang’s exit: allies retreating in Washington and the need to find profitable products sooner without his backing.
Friday, November 13
Microsoft @Microsoft Apologizes for the Microsoft 365 outage, reports broad recovery across its services and says the cause was a technical issue.
The FAA @FAANews Says a telecommunications outage halted all U.S. flights as a precaution, with every flight accounted for.
Robotaxi @robotaxi Announces that select customers are being invited to buy a Tesla Cybercab for personal use.
Marques Brownlee @MKBHD Lists the Cybercab order terms: invite only, delivery to certain Texas cities, use limited to existing geofenced areas, and a $300 monthly subscription.
Olivia Rinaldi @olivialarinaldi Reports that an update to the private network connecting U.S. air traffic control caused a reboot that crashed FAA communications, per CBS News sources.
ChatGPT @ChatGPT Introduces ChatGPT Wallet, which lets agents complete purchases with user approval and without access to the card details.
karan ahuja @ahuja_karan Says every US Pro, Max and Ultra subscriber on the Claude Duet waitlist has now been invited, with a note on usage.
Charles Lamanna @clamanna Says the Microsoft 365 outage was not a cyberattack or caused by AI tooling, and that the investigation points to cascading technical failures.
prinz @deredleritt3r Relays Fortune’s report that OpenAI disbanded its robotics team because it expects LLMs to become the brains of robots.
Alayna Treene @alaynatreene Reports Scott Bessent accusing the Commerce Department of delaying a national security designation for Nvidia customer Megaspeed.
Aaron Tilley @aatilley Reports that SpaceX cut the roughly 60-person team behind the Grok Imagine video generator, which stays live for now as resources shift elsewhere.
Thursday, November 12
NVIDIA Newsroom @nvidianewsroom Announces that Jensen Huang has resigned as CEO, with CFO Colette Kress named interim CEO.
will depue @willdepue Claims forensic evidence that DeepSeek, Moonshot, MiniMax and Ant Group trained their newest models on Nvidia Blackwell chips barred from export to China.
Microsoft 365 Status @MSFT365Status Says Microsoft is investigating failures reaching Exchange Online mailboxes through multiple connection methods in the Americas, tracked as EX1550543.
Logan Kilpatrick @OfficialLoganK Introduces Gemini 4 Hydrogen, Google’s fast Gemini 4 model, across AI Studio, the Gemini API, the Gemini app, Search and Antigravity.
International Cyber Digest @IntCyberDigest Reports that Microsoft’s outage spread beyond Outlook to Teams, SharePoint, Copilot, OneDrive and the Defender XDR security platform.
Oliver Carroll @olliecarroll Reports sources confirming Starlink Starshield is being throttled for Ukrainian forces above 160 km/h, starting right after this week’s negotiations.
Artificial Analysis @ArtificialAnlys Places Gemini 4 Hydrogen just below Claude Sonnet 5.5 and GPT-6.1 Sol on its Intelligence Index, at a third of Sonnet’s cost and about 2.5x GPT-6.1 Sol’s.
Joanna Stern @JoannaStern Reviews Apple’s $299 HomeView smart display through three hands-on tests, including Siri failures that drew promised fixes from Apple.
Ed Ludlow @EdLudlow Relays Bloomberg’s report that SpaceX-Tesla merger talks are progressing, with a deal announcement possible by early December.
Amir Efrati @amir Reports that Chinese companies tied to Alice Huang planned to give DeepSeek access to Nvidia chips banned from sale in China.
Wednesday, November 11
Arena @arena Reports Claude Fable 5.2 (Max) taking first place in Agent Arena across 4,700 real agentic sessions, and first in Code, Chat and Work.
Cyber Security News @The_Cyber_News Reports that attackers are exploiting an Outlook zero-day, fixed in November’s Patch Tuesday, to steal Windows authentication hashes.
Mitchell Hashimoto @mitchellh Provides libghostty-vt 1.0.1 release notes covering Windows binaries, glibc memory operations, parser optimizations, and a Linux scrollback throughput improvement from 298 to 1,983 MB/s.
Tuesday, November 10
Sebastian Raschka @rasbt Explains the shift toward hybrid and sparse attention architectures, distinguishing reduced attention computation from fixed-state recurrent memory and its retrieval tradeoffs.
The Hacker News @TheHackersNews Reports Google’s fix for CVE-2026-125350, a use-after-free in Chrome’s XML component and the eighth Chrome zero-day exploited in 2026.
Patrick Moorhead @PatrickMoorhead Lists the terms of Anthropic’s IPO roadshow: 825 million shares at $70 to $78 on Nasdaq as ANT, plus a $26B private placement.
Wall St Engine @wallstengine Relays the WSJ report that OpenAI’s Q3 revenue rose about 50% to $10B while its adjusted operating loss grew to about $10.9B.
Monday, November 9
Claude @claudeai Introduces Claude Fable 5.2 and Claude Mythos 5.2, with gains in coding, reasoning and vision.
Geoff Keighley @geoffkeighley Quotes Take-Two saying Grand Theft Auto VI pre-orders are tracking far above its expectations ahead of the November 19 launch.
Simon Willison @simonw Demonstrates using GitHub Markdown's collapsible details and summary tags to keep lengthy AI-generated pull-request descriptions out of the main review narrative.
Sunday, November 8
Salvatore Sanfilippo @antirez Demonstrates updated DwarfStar inference for DeepSeek 4.1 Flash using SSD streaming and notes an overthinking issue partially addressed by the latest model update.
Salvatore Sanfilippo @antirez Reports that DwarfStar's kimi-k3 branch can run K3.1 PRO inference directly from safetensors weights using SSD streaming.
Saturday, November 7
SpaceX @SpaceX Says Starship Flight 15, the first attempt to catch the ship, targets launch as early as Friday, November 13.
Friday, November 6
Mitchell Hashimoto @mitchellh Announces libghostty 1.0 with ABI stability and describes integrations including TUI fuzzing, a WASM build with an Xterm.js-compatible API, and bindings across multiple languages.
Wednesday, November 4
Salvatore Sanfilippo @antirez Describes predicting the next layer's top four MoE experts from the previous layer's residual to issue SSD prefetch hints during K3.1 inference on 128 GB machines.
Tuesday, November 3
Simon Willison @simonw Links a local reproduction showing Qwen 3.8 27B interpreting neural-network weights supplied as floating-point numbers after more than 45 minutes.
Sebastian Raschka @rasbt Summarizes ten recent open LLM architectures with parameter counts and design details spanning grouped-query attention, sparse attention, recurrent layers, looped transformers, and mixture-of-experts.
Xingyu Dang @xingyudang Presents a Writing Weights Index benchmark chart comparing models on weight-space programming and reports weaker results from open-source models.
Mitchell Hashimoto @mitchellh Details SIMD ASCII processing, memory-capacity prediction, packed styles, and store-forwarding fixes behind libghostty throughput gains ranging from 1.4x to 16x across workloads.
Monday, November 2
Rikiya Takehi @rikiyatakehi Introduces Weight-Space Programming, reporting that GPT-6 Astra edited 0.016% of Qwen3.5-0.8B's weights without gradients to improve math performance up to 6x and instruction following 3x.
Sunday, November 1
Andrej Karpathy @karpathy Explains building HackerTube by filtering 43 million Hacker News comments into 6,454 recommended videos, sharing the extracted dataset and an agent-assisted deployment workflow.