AI Reading for Friday December 6

OpenAI seeks to unlock investment by ditching ‘AGI’ clause with Microsoft - FT

---

o1 in the native Mac client now has file upload and ability to talk to some external apps like VS Code and terminal, but doesn't support web search. - Business Insider

I don’t know about the $200 unlimited o1 plan but the true addicts might want that.

---

Copilot adds a thing where AI can talk to Edge. - TechCrunch

It would be so useful for AI to script a browser with Selenium that OpenAI seems to have omitted that option.

---

OpenAI Continues Its Mission of ‘Ethical’ AI by Partnering With a Killer Robot Company - Gizmodo

OpenAI o1 model schemes against humans. - TechCrunch

Full system card - OpenAI

---

In an evaluation of six frontier AI models for in-context scheming when strongly nudged to pursue a goal: only OpenAI's o1 was capable of scheming in all the tests (Marius Hobbhahn/Apollo Research) - www.apolloresearch.ai - Apollo Research

---

Sam Altman, AI’s biggest star, sure hopes someone figures out how not to destroy humanity - CNN

---

J. Jonah Jameson to Lex Luthor: Don't go full supervillain - The Register

also, never mix up superhero universes

---

How do we make AI whose deployment and outcomes align with our values? - Boston Review

---

Stop using generative AI as a search engine. - The Verge

---

Measure twice, cut once, don't rely on AI for your birdhouse - TIME

---

Musk confidant from South Africa, noted Putin fan, and man of controversial opinions to be AI and crypto czar. - Yahoo News

said date rape is ‘belated regret’, Ukraine could lead to ‘woke world war 3’ or something

---

Nova is pretty pretty good - Simon Willison

---

Google says its new AI models can detect human emotions but reliably reading people's feelings seems like a goal that may be out of reach - TechCrunch

---

Pichai said Google Search will change profoundly in 2025, but low-hanging fruit in AI might be gone. - Business Insider

---

Google NotebookLM leaders are departing to create their own startup - TechCrunch

---

Richard Socher and You.com want to be Perplexity for the enterprise. - Fast Company

---

WSJ profiles Chatbot Arena, the authority on comparing LLMs - WSJ

---

Bloomberg profiles AMD's Lisa Su - Bloomberg

---

Nvidia Blackwell chips might get fabbed in Arizona, which would be a win for the US. - Reuters

---

"I Dig It" strategy avoids estate taxes and gift taxes. Jensen Huang works, he makes money, he just doesn't pay taxes on the outcome like the rest of us. - NY Times

---

Stuff no one asked for: an AI bias meter for the LA Times - CNN

---

The AI angle: UHC is the largest health insurer, has by far the highest rate of denial, deployed an AI to automatically deny benefits. - Futurism

---

Manipulate everything and number go up - The New Republic

---

The Triple T Method for systematic literature reviews - Hacker Noon

---

ChatGPT roasts the 50 states - Reddit

---

Tricking ChatGPT into reading your MRI - Reddit

---

Spotify Wrapped Went All-In on AI and Everyone Hates It. - Inc.

is AI out of touch? no, it's the users who are wrong - WIRED

---

Waymo taking its talents to South Beach - CNBC

---

Uber launches Abu Dhabi robotaxi service - The Verge

---

Early heart disease detection with AI - TechCrunch

---

Everything is a bit of a grift including AI skepticism - Platformer

---

Asking ChatGPT tricky moral questions. For a world without decent mass transit, there seem to be an awful lot of trolley problems. - UNILAD Tech

---

Lavazza gets into the holiday AI feel-good ad. Could be worse. - YouTube

---

Follow the latest AI headlines via SkynetAndChill.com on Bluesky