Welcome to TechCrunch’s AI newsletter!
In the latest AI news, OpenAI faced the departure of co-founder John Schulman, a key figure in the development of ChatGPT, who left for rival Anthropic. His focus on AI alignment and hands-on technical work led to this decision.
Speculation arises as Schulman’s exit coincides with OpenAI president Greg Brockman’s extended leave and the company’s DevDay event format change. Progress on a successor to GPT-4o seems slow, potentially indicating challenges for OpenAI.
OpenAI’s financial struggles, with projected losses of $5 billion this year, raise questions about its future and partnerships, notably with Microsoft. A shift away from its original mission may be necessary for survival.
In other news:
News
Apple’s Writing Tools feature faces limitations, particularly in handling sensitive topics.
Google refreshes its Nest Learning Thermostat after nearly a decade.
X’s chatbot spreads misinformation, raising concerns about AI-powered platforms.
OpenAI faces a lawsuit from a YouTuber over its AI model training methods.
AI lobbying intensifies as more groups advocate for AI-related issues at the federal level.
Research paper of the week
Researchers propose a “tamper-resistant” method to preserve AI model capabilities while deterring undesirable behavior, although scalability remains a challenge.
Model of the week
Black Forest Labs’ Flux.1 models offer competitive image generation capabilities, potentially rivaling established models like Midjourney and DALL-E 3.
Grab bag
Generative AI companies, like Suno and Nvidia, adopt the fair use defense for training models on copyrighted data, raising legal and ethical concerns in the AI industry.
