AI Reading for Wednesday May 22

x

---

Is Copilot+Recall a privacy nightmare? What is your more-human-than-human hybrid AI-cloud PC doing with the pictures and data it collects on you?

Microsoft is up to 135 Copilots and counting. - The Register

Microsoft enables enterprises to create agents with Copilot. - The Verge

Satya Nadella says you shouldn't anthropomorphize AIs. For one thing, they hate it when you do that. - Bloomberg

Microsoft CTO says AI capabilities will keep on growing exponentially, haven't reached the sigmoid flattening part of the curve yet - Reddit

---

WSJ on vector databases. Not sure you need a whole new database engine to do fast nearest neighbors but here we are. - WSJ

---

Gen-AI ads will show how clothes fit your body and skin tone. - WIRED

---

The 'I made this' and data rape ethos now encompasses every bit of you and every bit of content you make and anyone ever made. Consent be damned. - The Atlantic

---

What happens when gen AI kills all the reliable sources? - Computerworld

---

Meta's chat bot scrapes news sites and summarizes them, will cite sources on demand. - Washington Post

---

The Stig comfortably beats an autonomous racecar on the track - Top Gear

---

When a Waymo gets confused, it sends an SOS to a remote driver. - Ars Technica

---

Suno music-composing AI raises $125m at $500m valuation - Suno.com

---

Cloud AI price wars in China - The Register

---

China now has ChatXiPT with Xi Jinping thought on socialism with Chinese characteristics. Hope it doesn't hallucinate and stray from the party line. - FT

---

Adobe brings Firefly AI to Lightroom. - Fstoppers

---

Google Photos rolls out Magic Editor to smartphone users. - PhoneArena

---

Humane seeks buyer after raising $200m, getting torn to shreds by reviewers. - Bloomberg

---

Grok working on multimodality. Will be trained on audio, video and (non-woke) brainwaves. - The Verge

---

Is your AI-first, cloud-centric, closed ecosystem PC platform taking you hostage? - VentureBeat

Open source in the cloud AI era is losing its meaning. Kind of need a rating system. - NY Times

---

Anthropic work sheds light on how LLMs think and store knowledge, could improve interpretability and also ability to train and add new knowledge easily. - NY Times

---

Khanmigo now free to teachers in US as Microsoft offers hosting. - CNBC

---

How AI handles moral dilemmas. If only there were some AI assistant that could have told Sam Altman what he was doing was not a good idea - FAR AI

---

Follow the latest AI headlines via SkynetAndChill.com on Bluesky