Sunday, March 1, 2026
FDA trades monkeys for machines
Hey builders, today brings us more news in AI. The FDA is ditching monkeys for AI models and human organoids in drug testing (bold move), while Switzerland just dropped a fully open AI model with completely transparent training data and architecture—no black boxes here. Meanwhile, chatbot interviewers are quietly taking over hiring screens across the industry. Would you trust an AI to decide if you're worth hiring?
Top Stories
The FDA is replacing animal testing requirements with AI computational models, human organoids, and real-world data to accelerate drug development, improve safety predictions, and reduce costs. This paradigm shift in drug evaluation begins immediately and could eventually spare thousands of laboratory animals while delivering treatments to patients faster.
Article content unavailable due to security verification blocking access; title suggests research on AI chatbot adoption in recruitment and hiring processes.
artificialintelligence-news.com
Switzerland releases Apertus, a fully transparent open-source AI model (8B and 70B parameters) trained on 15 trillion tokens in 1,000+ languages, with complete public access to training data, weights, and architecture. The initiative positions AI as public infrastructure while demonstrating that powerful generative models can maintain full openness and regulatory compliance.
A highly-viewed Twitter post from Zed announcing a user-requested feature related to agent infrastructure standardization, though the full technical details are not included in the provided content.
Keep Reading
Industry Voices
Krishna Rao
Chief Financial Officer at Anthropic
Rare insider view into how AI labs actually fund multi-billion dollar compute infrastructure and navigate existential cash burn.
Fidji Simo
CEO of Applications at OpenAI
Steers how billions of people will actually interact with AI through ChatGPT's product evolution and enterprise rollout strategy.
Dean Leitersdorf
Decart
Building real-time AI world simulation engines that could remake gaming, film production, and synthetic training data generation.
Martin Jaggi
Professor of Machine Learning at EPFL
Pioneering federated learning and decentralized AI training methods that could break Big Tech's monopoly on model development.
Dharmesh Shah
HubSpot co-founder investing heavily in AI tooling with boots-on-the-ground perspective on what actually drives enterprise adoption.
Enjoyed this issue?
Get daily AI intel delivered to your inbox. No fluff, just the stories that matter.