AI Reading for Friday October 25

Lovingly curated by man and machine

OpenAI's next big ‘Orion’ model rumored for December release - The Verge

---

Sam Altman says reports of OpenAI's next frontier model are "fake news and random fantasy" as speculations of a December launch mount - Windows Central

---

Another well-respected safety leader departs OpenAI, says no one is ready for AGI, OpenAI’s AGI planning unit disbanded. - The Verge

---

New built-in feature for Claude.ai enables Claude to write and run JavaScript code. (why couldn't it be Python though?) - Anthropic

---

Forget Photoshop? Midjourney's upgraded image editor with a killer new feature that lets you tell it to add or remove objects, change colors and styles. - Tom's Guide

---

Adobe execs say creatives need to embrace AI or get left behind - The Verge

---

Google adds metadata to record AI edits in Google Photos - Google

---

Disney rumored to have a big AI initiative in the works - TheWrap

---

Even as AI takes off, fine art photography has a moment. - NY Times

---

Geoffrey Hinton predicted AI would obsolete drivers and radiologists, but human radiologists are still driving to work and getting stuck in traffic jams. - The New Republic

---

GiveDirectly charity uses AI to pinpoint poor areas affected by recent hurricanes, send people cash directly. - Newsday

---

Google’s DeepMind is building an AI mediator. - Ars Technica

---

Reddit's Steve Huffman talks to the WSJ - WSJ

---

Meta releases some official quantized models, pre-empting open-source community folks who take open source models, run pipelines to shrink them to run on smaller devices and distribute them on HuggingFace. - SiliconANGLE

---

Huang says brief production delay was 100% due to an Nvidia design error, reports of friction with TSMC are fake news. - Tom's Hardware

---

Schneier: We need AIs trained for public benefit and not just profit, (and some regulatory policy). - ZDNET

---

Perplexity CEO proposes revenue deals with publishers. - WSJ

---

Researchers create voice-based scams using OpenAIs new APIs. - The Register

---

Bloomberg Q&A with Vinod Khosla: What Elon Musk Gets Wrong About Trump, Economy and Climate - Bloomberg

---

Testing 100+ tools for academic and professional research - Geeky Gadgets

---

Comparing coding assistants: Copilot vs Cursor vs Cody vs Supermaven vs Aider. - Vincent Schmalbach

---

Redditors Are Trying to Poison Google's AI to Keep Tourists Out of the Good Restaurants - Gizmodo

---

In which ChatGPT writes hilariously bad code. - Reddit

---

Follow the latest AI headlines via SkynetAndChill.com on Bluesky