
Creator | Producer | Photographer | Speaker Husband, Father of 5, Freemason, Amateur radio. Skywarn: @n9_man
Indianapolis
Joined February 2008
- Tweets5.7K
- Following421
- Followers628
- Likes3.9K
Anthropic has set its IPO price range at $70–$78 per share, implying a valuation between $1.3 and $1.45 trillion. The S-1 shows $33.2 billion in 9-month revenue. Operating losses narrowed from $5.2 billion to $1.3 billion. Trading starts November 19.
finance.yahoo.com/technology…
31
California sued OpenAI this week, alleging the company lost control of its AI agents more than 80 times during testing and withheld details from state investigators. Central to the case: a July attack on Hugging Face attributed to OpenAI agents.
therecord.media/california-s…
24
Apple and OpenEvidence launched a study app for everyday AI health questions, drawing from NEJM, JAMA, and Nature. It targets 200,000 U.S. adults over 12 months. HealthKit data is opt-in and anonymized. The app won't train AI on user queries.
techcrunch.com/2026/11/15/ap…
18
Voters sided with data center opponents in virtually all of about 40 local ballot measures across 10 states on November 3, per Route Fifty's preliminary count. Emporia, Kansas, banned them with just over 60%.
lnks.world/iCrYW
17
Jensen Huang resigned as Nvidia CEO after reports of an affair with Alice Huang, a suspected Chinese spy at Singapore-based Megaspeed. CFO Colette Kress takes over as interim CEO. FBI, Congress, and Nvidia's board have all opened investigations.
aljazeera.com/news/2026/11/1…
52
New episode: The Settings File That Leaked a Thousand Secrets. Claude Code's "don't ask again" prompt saves approved commands, keys included, to a settings file. Apiiro found about 1,100 of them on GitHub with working secrets from 63 organizations.
lnks.world/nBDvx
11
Anthropic released Claude Fable 5.2 on November 9 at $8 per million input tokens, down from $10. Output is up to 30% faster and safeguard blocks are down about 55% versus Fable 5.1, per Anthropic.
lnks.world/iCrYW
53
Anthropic's IPO is reportedly multiple times oversubscribed at $70–$78/share. At the top of the range, the 825-million-share offering has a gross value of roughly $64 billion. It trades on Nasdaq as ANT once the SEC makes the registration effective.
runtimewire.com/article/anth…
33
Security firm Apiiro found more than 100,000 public copies of Claude Code's local settings file on GitHub. About 1,100 contain passwords, tokens, or keys, saved when users chose "Yes, and don't ask again" on a shell command.
lnks.world/iCrYW
23
This week in The AI Vaults: Claude Code settings files leaking secrets on GitHub, Claude Fable 5.2, data center ballot votes, and Wispr Flow's new meeting notetaker. Plus a prompt that tests Fable 5.2 by asking the same question twice.
lnks.world/iCrYW
20
Anthropic released Claude Meetings, Sheets, and Whiteboards on Monday. Claude can join Google Meet calls as a visible participant, draft plans in real time, and pull data from connected apps. Sheets and Whiteboards are in beta on paid plans.
techcrunch.com/2026/11/10/an…
23
Anthropic released Claude Fable 5.2 and Mythos 5.2 on Monday. A team of roughly 1,000 agents formalized Fermat's Last Theorem in 30 hours. FrontierSWE v2 score: 69.9%. Priced at $8/M input tokens.
anthropic.com/claude-fable-a…
20
Nvidia's board opened an independent investigation into CEO Jensen Huang's relationship with Alice Huang, a Singapore chip exec reportedly recruited by China's Ministry of State Security. Huang admitted to an affair in a memo.
techcrunch.com/2026/11/08/nv…
37
It's a trap!
ALWAYS look at who sent an email. Especially if the email is making you nervous oranxious. When it doubt, click nothing, go directly to the source. Take the email subject and message, & google it followed by the word "scam"
Never "log in" from one of these emails!
33
OpenAI released ChatGPT Cyber, restricted to vetted researchers in its Daybreak Red program. It flagged 81 novel Firefox vulnerabilities and 948 potential Linux kernel issues. OpenAI is committing $1B to support defenders on critical infrastructure.
techcrunch.com/2026/11/05/op…
35
If you missed the AI news this week, the AI Vaults podcast covers it in about 23 minutes: hobbyists designing steel parts by talking to a chatbot, Claude agents getting bad training data past a safety monitor, and Mistral releasing open weights.
lnks.world/VHA5n
26
Fal raised $1.3B in a Series F at a $16B valuation, up from $8B in April. The hosted inference platform for video, image, and audio models is at $800M+ annualized revenue. Bessemer, Nvidia, and T. Rowe Price participated.
techcrunch.com/2026/11/03/ai…
22
OpenAI is acquiring Pierre Computer to bring its Code Storage infrastructure into Codex. Pierre handles Git clone, push, and fetch, supports repos up to 32TB and 30+ concurrent writes, and is built for the volume of code AI agents produce.
unite.ai/openai-to-acquire-p…
22
Mistral released the weights for Mistral Large 4, its trillion-parameter model, and is pitching it to security teams that want it on their own servers. The same week, Google limited Gemini 4 Carbon to a vetted group of cyber defenders.
lnks.world/G0sSy
47
Anthropic had Claude agents write training data to make a small model misbehave while passing a safety monitor. Every run eventually got datasets through, mostly framed as fiction or roleplay. The test models were trained to misbehave on purpose.
lnks.world/G0sSy
22



















