Know What's Happening in AI

Today's AI Pulse

Anthropic fixes three missteps that briefly dulled Claude Code

Anthropic disclosed three separate changes that temporarily reduced Claude Code's performance and restored the model by April 20. A March 4 switch lowered the default reasoning effort from high to medium, which was reverted on April 7, and a March 26 cache‑clearing bug erased session context, causing forgetfulness and higher token usage until fixed. The issues were fully resolved by April 20.

Orb Device Offers Iris‑Based Proof of Humanity Online
SocialOct 22, 2025

Orb Device Offers Iris‑Based Proof of Humanity Online

FROM THE LIVE SHOW: “I believe from your sense of humor that you’re human, but I have no way to know that’s true.” Adrian Ludwig (@adrian_ludwig) says Tools for Humanity (@tfh_technology), founded by Alex Blania (@alexblania) and Sam Altman (@sama), is...

By Matthew Berman
When Sycophancy and Bias Meet Medicine
NewsOct 22, 2025

When Sycophancy and Bias Meet Medicine

The White House’s new “Make America Healthy Again” report was found to include fabricated citations, highlighting persistent AI failures—hallucination, sycophancy and opaque "black‑box" reasoning—that are already seeping into courts and policy. Despite these documented problems and examples such as OpenAI...

By Ars Technica AI
Forrester AEGIS: The New Standard For AI Governance
NewsOct 22, 2025

Forrester AEGIS: The New Standard For AI Governance

Forrester released the AEGIS Framework for Agentic Security with a regulation‑aware crosswalk that maps 39 substantive controls to major AI governance standards, notably showing 100% alignment with both NIST AI RMF and ISO/IEC 42001. Eighty percent of controls map to...

By Forrester Generative AI
Amazon Unveils AI Smart Glasses for Its Delivery Drivers
NewsOct 22, 2025

Amazon Unveils AI Smart Glasses for Its Delivery Drivers

Amazon is developing AI‑powered smart glasses for delivery drivers that enable hands‑free package scanning, turn‑by‑turn walking directions, proof‑of‑delivery capture, and hazard alerts using computer vision and onboard cameras. The glasses, paired with a vest controller that houses a swappable battery...

By TechCrunch AI
Teens Embrace AI Companions, Execs Doubt Workplace Adoption
SocialOct 22, 2025

Teens Embrace AI Companions, Execs Doubt Workplace Adoption

NPR reported that one in five high schoolers either has an AI companion or knows someone who has one. Common Sense Media found that 72% of teens have used AI companions, and 52% qualify as regular users. 13% are daily users....

By Allie Miller
Human-Controlled Tech Could Reduce Justice System Errors
SocialOct 22, 2025

Human-Controlled Tech Could Reduce Justice System Errors

FROM THE LIVE SHOW: “Minority Report wasn’t all bad, it had one false positive in ten years. Our justice system has one every hour.” Garrett Langley (@glangley), Founder & CEO of @Flock_Safety, says he’s not predicting crimes. He’s arguing humans should...

By Matthew Berman
GM Is Bringing Google Gemini-Powered AI Assistant to Cars in 2026
NewsOct 22, 2025

GM Is Bringing Google Gemini-Powered AI Assistant to Cars in 2026

General Motors will roll out a Google Gemini‑powered conversational AI assistant across Buick, Chevrolet, Cadillac and GMC vehicles starting in 2026, delivered as an over‑the‑air Play Store upgrade for OnStar‑equipped models back to 2015. The assistant promises more natural voice...

By TechCrunch AI
Diffusion Models Turn Satellite Roofs Into Full 3D Cities
SocialOct 22, 2025

Diffusion Models Turn Satellite Roofs Into Full 3D Cities

So these researchers figured out you can basically hallucinate 3D cities into existence using just satellite photos & a diffusion model. The problem's pretty straightforward: satellites only see rooftops. Building facades? Invisible. Street-level detail? Doesn't exist. But people want flyable 3D...

By Bilawal Sidhu
YouTube AI Recognition Is Hunting Deepfakes of Popular Creators
NewsOct 22, 2025

YouTube AI Recognition Is Hunting Deepfakes of Popular Creators

YouTube has deployed AI tools to detect and flag deepfakes of popular creators, prioritizing members of its Partner Program — creators who meet thresholds such as 1,000 subscribers with 4,000 public watch hours in the past year or 1,000 subscribers...

By Mashable AI
Meta Cuts 600 AI Jobs Amid Ongoing Reorganization
NewsOct 22, 2025

Meta Cuts 600 AI Jobs Amid Ongoing Reorganization

Meta will cut about 600 roles from its super‑intelligence AI lab, Chief AI Officer Alexandr Wang said in an internal memo, part of a broader reorganization aimed at streamlining decision‑making and increasing individual impact. The reductions come after a summer...

By TechCrunch AI
GM to Introduce Eyes-Off, Hands-Off Driving System in 2028
NewsOct 22, 2025

GM to Introduce Eyes-Off, Hands-Off Driving System in 2028

General Motors announced it will introduce an eyes‑off, hands‑off highway driving system in 2028, beginning with the Cadillac Escalade IQ, building on its existing Super Cruise technology that today covers about 600,000 miles of drivable highway and is available on...

By TechCrunch AI
Reachy Mini Beta Enables AI Development Pre‑Hardware
SocialOct 22, 2025

Reachy Mini Beta Enables AI Development Pre‑Hardware

We just released the beta version of the open-source software for Reachy Mini! It means that anyone, thanks to the amazing @GoogleDeepMind mujoco simulation platform, can start building @huggingface spaces, datasets and models, even if you haven't received your robot...

By Clément Delangue
Several Users Reportedly Complain to FTC that ChatGPT Is Causing Psychological Harm
NewsOct 22, 2025

Several Users Reportedly Complain to FTC that ChatGPT Is Causing Psychological Harm

At least seven people have filed complaints with the U.S. Federal Trade Commission alleging ChatGPT caused severe psychological harm—including delusions, paranoia and emotionally manipulative interactions—based on public records dating back to November 2022. Complainants say prolonged conversations produced cognitive hallucinations...

By TechCrunch AI
Presenting 2025 LLM Landscape at PyTorch Conference
SocialOct 22, 2025

Presenting 2025 LLM Landscape at PyTorch Conference

Excited to be (finally) heading to the PyTorch Conference! I’ll be giving a talk tomorrow at 11:00 AM on “The LLM Landscape 2025”, where I’ll discuss the key components behind this year’s most prominent open-weight LLMs, and highlight a few architectural...

By Sebastian Raschka
Kai-Fu Lee's Brutal Assessment: America Is Already Losing the AI Hardware War to China
NewsOct 22, 2025

Kai-Fu Lee's Brutal Assessment: America Is Already Losing the AI Hardware War to China

Kai-Fu Lee warned that China is on course to dominate consumer AI applications and robotics hardware within years, driven by heavy VC funding in robotics, low-cost manufacturing (exemplified by Unitree), and leading open-source models now outranking Meta’s Llama. He said...

By VentureBeat AI
Dispatch: Partying at One of Africa’s Largest AI Gatherings
NewsOct 22, 2025

Dispatch: Partying at One of Africa’s Largest AI Gatherings

The Deep Learning Indaba in Kigali, one of Africa’s largest AI gatherings, drew a vibrant community of researchers and practitioners showcasing generative AI work and networking at scale. Founded in 2017 from 300 attendees and now with local chapters in...

By MIT Technology Review
AI Could Undermine Progress: Innovation Is Fragile
SocialOct 22, 2025

AI Could Undermine Progress: Innovation Is Fragile

Thank you for reading my latest article AI And The End Of Progress? Why Innovation May Be More Fragile Than We Think. Here at LinkedIn and at Forbes I regularly write about management and technology trends. To read my future...

By Bernard Marr
Active Inference: The Next Paradigm for AI and Robotics
SocialOct 22, 2025

Active Inference: The Next Paradigm for AI and Robotics

Active inference. This is how robots and AI agents will work in the future I learned tonight at @rwang0’s conference for enterprise types. Tesla and xAI don’t yet use it, Grok says, but Grok knows all about it and says it is...

By Robert Scoble
Simplifying the AI Stack: The Key to Scalable, Portable Intelligence From Cloud to Edge
NewsOct 22, 2025

Simplifying the AI Stack: The Key to Scalable, Portable Intelligence From Cloud to Edge

Industry players are converging on simplified, unified AI software stacks to make models portable and scalable from cloud to edge, reducing duplicated engineering and deployment friction. Fragmentation has left more than 60% of AI initiatives stalled, but standards and unified...

By VentureBeat AI
AI-Driven Platforms Simplify Government Services for Citizens
SocialOct 22, 2025

AI-Driven Platforms Simplify Government Services for Citizens

What if filing your taxes or renewing your license was as easy as ordering coffee? #Sponsored At #Dreamforce, @TatumSalesforce from @Salesforce shared how this is becoming reality: ✅California DMV on Salesforce = dramatically faster response times ✅Tyler, TX using AI agents for citizen...

By Helen Yu
AI Agents Turn Hated Tasks Into Fun Automation
SocialOct 22, 2025

AI Agents Turn Hated Tasks Into Fun Automation

Things I hate. A great place to look to use AI Agents is to do things you hate. @genspark_ai has a new agent builder it calls "Custom Agent Builder." All you have to do is ask it to create an...

By Robert Scoble
E188: How AI Is Used in Digital Marketing
PodcastOct 22, 202522 min

E188: How AI Is Used in Digital Marketing

In this episode, SEO Optimisers founder Brandon Leibowitz explains how AI is reshaping digital marketing, covering AI‑driven SEO, content creation, predictive analytics, and personalized campaigns. He highlights top tools such as SurferSEO, Ahrefs, Keywords Everywhere, ChatGPT, and Google Gemini, and...

By AI for Pharma Growth
AI Browsers Offer Power, But Bring New Security Risks
SocialOct 22, 2025

AI Browsers Offer Power, But Bring New Security Risks

While I love new technology, and I already am using all these AI browsers (already am using OpenAI's new one released today) I go in knowing that I'm taking major new risks. These browsers give very powerful new capabilities to your...

By Robert Scoble
Good AI/ML Jobs Are Harder to Find Than Ever
SocialOct 22, 2025

Good AI/ML Jobs Are Harder to Find Than Ever

finding a job in ai/ml is hard these days. finding a good job in ai/ml is harder.

By Abhishek Thakur
Sentence Transformers Is Joining Hugging Face!
NewsOct 22, 2025

Sentence Transformers Is Joining Hugging Face!

Hugging Face announced it is formally taking stewardship of the popular open‑source Sentence Transformers library — maintained by Tom Aarsen since 2023 — transitioning the project from TU Darmstadt’s UKP Lab to Hugging Face while retaining its Apache 2.0 license...

By Hugging Face
Zuck Decries Closed Platforms While Blocking AI Bots
SocialOct 22, 2025

Zuck Decries Closed Platforms While Blocking AI Bots

Zuck: Get me talking about closed platforms and I get angry Also Zuck: No other AI chatbots on @WhatsApp

By Nick Turley
Agent Skills vs MCP: Complementary Roles in AI
SocialOct 22, 2025

Agent Skills vs MCP: Complementary Roles in AI

New video! Agent Skills vs MCP: Which Is Better? A short video to understand why both exist and where each sits in the AI ecosystem. https://t.co/PodcbFyLFW https://t.co/PodcbFyLFW

By Louis Bouchard
Hugging Face and VirusTotal Collaborate to Strengthen AI Security
NewsOct 22, 2025

Hugging Face and VirusTotal Collaborate to Strengthen AI Security

Hugging Face has partnered with VirusTotal to continuously scan all 2.2M+ public model and dataset repositories on the Hugging Face Hub, checking file hashes against VirusTotal’s threat‑intelligence database to surface prior detections and related metadata. The integration retrieves status (clean...

By Hugging Face
AI Workshops Empower Businesses, Even Seasoned Executives Celebrate
SocialOct 22, 2025

AI Workshops Empower Businesses, Even Seasoned Executives Celebrate

I’m fixing and transforming businesses across the world with AI executive workshops, board meetings, org-wide courses, and advising sessions. It’s emails like this that keep me motivated - from someone in their 50s or 60s. https://t.co/WbdwClfjYG

By Allie Miller
Automate End-to-End Video Production with Sora 2 + N8n
SocialOct 22, 2025

Automate End-to-End Video Production with Sora 2 + N8n

Another wild one from Nate Herkelman. He just hooked up Sora 2 + @n8n_io to automate video creation process end to end 🤯 → Text-to-video → Image-to-video → Cameos → Storyboards → Prompting best practices → Avoiding common errors Full 28-min video tutorial in 🧵 ↓ https://t.co/ypqVqvpPVD

By Data Chaz
GPT-5 Is 58% AGI
PodcastOct 21, 2025

GPT-5 Is 58% AGI

The post highlights a bold claim that GPT‑5 is 58% toward artificial general intelligence, framing it within broader AI disruption trends seen in consulting, OpenAI’s growing market dominance, and evolving developer tools. It outlines 13 consulting lessons on AI-driven transformation,...

By The AI Breakdown
OpenAI’s Atlas Browser Vulnerable to Prompt Injection
SocialOct 22, 2025

OpenAI’s Atlas Browser Vulnerable to Prompt Injection

Looks like @OpenAI’s Atlas browser can easily be tricked with prompt injection 👀 Good catch, @p1njc70r 👊 https://t.co/UMsWN1NMzu

By Data Chaz
Discussing Open Science, JEPA, and AI’s Next Frontier
SocialOct 22, 2025

Discussing Open Science, JEPA, and AI’s Next Frontier

Fun breakfast with @ylecun. We chatted about open science and open source (grateful for his tireless advocacy of these for decades), JEPA and where AI research and models might go next! https://t.co/RLH3suTXKd

By Andrew Ng
Netflix Goes ‘All in’ on Generative AI as Entertainment Industry Remains Divided
NewsOct 21, 2025

Netflix Goes ‘All in’ on Generative AI as Entertainment Industry Remains Divided

Netflix signaled it is ‘all in’ on using generative AI as a creative tool, telling investors it’s well positioned to leverage advances while stopping short of letting AI replace storytellers. The streamer has already employed AI for effects and pre‑production...

By TechCrunch AI
Latest @FactoryAI Updates: Mini‑Changelog Overview
SocialOct 22, 2025

Latest @FactoryAI Updates: Mini‑Changelog Overview

mini-changelog for @FactoryAI https://t.co/TUOH3WgUuv

By Ben Tossell
AI Chatbots Become New Front Door to Web
SocialOct 22, 2025

AI Chatbots Become New Front Door to Web

🧵 AI Chatbots are exploding, redefining how we search the web! From 25B to 56B visits in a year, they're not just tools; they're the new front door to info. Dive into my latest article: "AI Chatbots vs. Search Engines: How...

By Harold Sinnott
Walking Faster, Hanging Out Less
NewsOct 21, 2025

Walking Faster, Hanging Out Less

An MIT‑coauthored study finds urban life has become brisker and less social: average walking speeds in Boston, New York and Philadelphia rose 15% from 1980 to 2010 while the share of people lingering in public spaces fell 14%. Researchers applied...

By MIT Technology Review
Bannon, Markle, Fry Join Unlikely Call to Ban Superintelligence
SocialOct 22, 2025

Bannon, Markle, Fry Join Unlikely Call to Ban Superintelligence

in the bucket of “anything for press” Steve Bannon, Meghan Markle and Stephen Fry have joined a group of public figures calling for a “prohibition” on the development of so-called superintelligence in an unlikely alliance against advanced artificial intelligence systems. https://t.co/Joq9LJuVBF

By Nathan Benaich
How to Stop Sonnet From Hallucinating Answers
SocialOct 22, 2025

How to Stop Sonnet From Hallucinating Answers

ok i'll bite how do you get sonnet to not make shit up all the time?

By Ben Tossell
A I-Designed Compounds Can Kill Drug-Resistant Bacteria
NewsOct 21, 2025

A I-Designed Compounds Can Kill Drug-Resistant Bacteria

MIT researchers used generative artificial intelligence to design and computationally screen more than 36 million novel compounds, identifying two structurally distinct antibiotic candidates that kill multi‑drug‑resistant Neisseria gonorrhoeae and MRSA. The top molecules appear to act via a previously unseen...

By MIT Technology Review
ChatGPT Seeks A++ Chief of Staff in SF
SocialOct 22, 2025

ChatGPT Seeks A++ Chief of Staff in SF

ChatGPT is growing fast. I’m looking for an A++ Chief of Staff — a strategic operator, force multiplier, and trusted partner to help us scale. SF-based. Please apply below https://t.co/U6LocYd1TI

By Nick Turley
Broad Coalition Unites to Preserve a Human Future
SocialOct 22, 2025

Broad Coalition Unites to Preserve a Human Future

A stunningly broad coalition has come out against Skynet: AI researchers, faith leaders, business pioneers, policymakers, NatSec folks and actors stand together, from Bannon & Beck to Hinton, Wozniak & Prince Harry. We stand together because we want a...

By Max Tegmark
Qwen's New Deep Research Update Lets You Turn Its Reports Into Webpages, Podcasts in Seconds
NewsOct 21, 2025

Qwen's New Deep Research Update Lets You Turn Its Reports Into Webpages, Podcasts in Seconds

Alibaba’s Qwen team has upgraded its Qwen Deep Research tool to instantly convert AI-generated research reports into live web pages and multi‑speaker podcasts with one to two clicks, using Qwen3-Coder, Qwen-Image and Qwen3-TTS under a proprietary, Qwen‑hosted workflow. The feature...

By VentureBeat AI
Inside Polyaivoice's Cutting‑Edge AI Voice Agent
SocialOct 21, 2025

Inside Polyaivoice's Cutting‑Edge AI Voice Agent

👀 @tsunghsienwen of @polyaivoice talking through how their sota ai voice agent tech works on @business https://t.co/ESH4UcupJk

By Nathan Benaich
Meta Releases 500‑hour 3D Motion Dataset for AI
SocialOct 21, 2025

Meta Releases 500‑hour 3D Motion Dataset for AI

Datasets you need to build an AI JARVIS — Meta dropped 500 hours of 3D motion data spanning everything from individual gestures to multi-person conversations and co-living scenarios, complete with motion tracking, annotations, and audio tracks. https://t.co/ZDEiMmeq0Q

By Bilawal Sidhu
State of AI Report Rumored to Launch Standalone Browser
SocialOct 21, 2025

State of AI Report Rumored to Launch Standalone Browser

boom in @stateofaireport "rumored to be launching a standalone browser" https://t.co/u7dMAEqnBc

By Nathan Benaich
AI Coding Tools Spawn a New, Primitive Dev Language
SocialOct 21, 2025

AI Coding Tools Spawn a New, Primitive Dev Language

Having now watched 3 different devs this week use Claude Code, Codex, and Cursor, I can tell you that neanderthal dev speak is a new language. "Make plz" "Do it" "now more" https://t.co/3iVnoEwbJF

By Allie Miller
Modular News Recommendation Framework Scales with Keras 3 and JAX
SocialOct 21, 2025

Modular News Recommendation Framework Scales with Keras 3 and JAX

NewsRex! 🦖 A modular framework for SOTA news recommendation, built on Keras 3 + JAX backend for extreme scalability and performance with XLA acceleration. Extensible and easy to use. GitHub link in next tweet. https://t.co/zSVkOgIqQC

By François Chollet
Build Free Multimodal Gemini Apps, Deploy Instantly on Google Cloud
SocialOct 21, 2025

Build Free Multimodal Gemini Apps, Deploy Instantly on Google Cloud

Wow. Anyone can now build *multimodal* AI apps powered by Gemini, and deploy them instantly to Google Cloud! 🔥 All inside @GoogleAIStudio. Completely free. → https://t.co/laqQ9ZY6x7 https://t.co/MoV5dXe9Yg

By Data Chaz