TechCrunch Industry News

A new Stanford study tests how chatbots handle personal advice and where that can go wrong. Researchers measured sycophancy across multiple AI models and real-world datasets. Experiments show chatbots often validate risky behavior and people sometimes prefer flattering responses. The report explores how this could shape moral certainty and engagement incentives.

A leaked hacking toolkit puts millions of iPhones and iPads at risk. The conversation breaks down two advanced toolkits, Karuna and Darksword, and which iOS versions they target. Listeners hear how the leak enables easy drive-by infections and where attacks have been observed globally. Practical mitigation steps like updating iOS and using Lockdown Mode are highlighted.

A policy about region-weighted creator payouts on X gets paused after backlash and new rules target AI-generated war misinformation. A startup repurposes rocket telemetry tools for factory floors to make sensor data AI-ready and cut automated testing costs. Funding and big-name customers highlight the industrial shift toward advanced manufacturing data infrastructure.

A robotics company will integrate DeepMind's Gemini robotics models and share robot-collected data for testing and deployment across factories, auto plants, data centers and logistics. The conversation covers how hardware–software partnerships are accelerating physical AI and scaling autonomous robots. Also covered is a smart-ring maker’s renewed push into the U.S. market after resolving import issues.

FBI warns Iran-linked hackers are using Telegram bots and malware to target dissidents, opposition groups, and journalists. The Pentagon labels an AI lab a supply-chain risk, sparking legal fights and political pushback. A major publisher pulls a horror novel amid allegations that AI was used to create the text.

Discussion about companies giving engineers AI token budgets as a new form of compensation. Conversation about how token valuation, consumption by agentic systems, and leaderboards could reshape workplace perks. Examination of risks like changed expectations, headcount math, and tokens masking flat cash pay. A retrospective look at the first tweet and the platform’s evolution over 20 years.

Meta reverses course and keeps Horizon Worlds on Quest VR after nearly shutting it down. The company’s huge Reality Labs losses and sinking VR hardware demand are examined. CISA warns firms to lock down Microsoft Intune after attackers exploited remote device management to wipe systems. A pro-Iran group claims the hack and the FBI moves to disrupt their site.

A debate over whether AI companies must pay creators for training data, with claims that fair use defenses are shaky. A discussion on how licensing deals with big publishers change the economics for artists. A look at Nvidia’s stealth growth in networking, turning it into a multibillion-dollar AI data-center powerhouse.

A legal battle heats up as Arizona accuses a prediction market platform of unlicensed gambling and election wagering. Debate over state vs federal authority intensifies. A new AI offering lets companies train models from scratch on their own data, promising tailored multilingual and domain-specific performance.

They discuss building a visual memory layer that indexes and retrieves video memories for wearables and robots. The conversation covers a custom wearable recorder and partnerships to run models on edge hardware. They also break down a new enterprise AI agent platform designed to bring OpenClaw-style agents with stronger security and local control.