Observed Signal · Oct 10, 2026 · Panel Discussion · Source: AINews swyx · Impact: 2/5 · Sentiment: Positive
DeepMind and Biohub Discuss AlphaFold's Limits and Future of AI Biology
In a panel discussion, Google DeepMind's Pushmeet Kohli and Biohub's Sal Candido shared insights on the future of AI-driven biology. They emphasized that scaling compute and data alone is insufficient; finding the right scaling laws and curating good data are crucial. Kohli noted that AlphaFold, while a breakthrough, did not fully solve protein folding, as dynamics and disorder remain challenging. He highlighted the importance of trustworthiness and uncertainty calibration over full interpretability. Candido discussed how low-quality metagenomic data can improve protein language models and stressed moving from modeling individual proteins to whole biological systems. Both agreed that AI is already used in drug discovery, but a 10x acceleration requires overcoming fundamental biological challenges, as pursued by Biohub's mission to cure all diseases.
This article discusses the future of AI in biology, which is relevant to AI technologies used in business, but it is a panel discussion without breaking news. The insights on scaling laws and data quality are valuable but not industry-shifting for AdTech.
Track Google DeepMind Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your Polaris7 workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- Google DeepMind's Pushmeet Kohli and Biohub's Sal Candido participated in a panel on AI-driven biology.
- Kohli stated that AlphaFold did not fully solve protein folding, with dynamics and disorder remaining challenges.
- Candido noted that metagenomic data improves protein language models.
- Both experts emphasized the importance of good data and finding the right scaling laws.
- AI is already used in drug discovery, but a 10x acceleration requires further fundamental research.
Connected Companies & Entities
1 Entity mapped“Google DeepMind's Pushmeet Kohli participated in the panel.”
Ontology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
Is Research Over Because of AI? Eric Schmidt Weighs In
An MIT Technology Review commentary (published on t3n.de) examines whether AI breakthroughs like AlphaFold mean the end of traditional scientific discovery. The article recalls AlphaFold's ability to predict protein 3D structures, notes that Demis Hassabis and John Jumper of Google DeepMind were awarded a 2024 Nobel Prize for chemistry (as reported), and argues that AlphaFold sparked a wave of startups and large investments in AI for biology, chemistry, and materials. The piece contends that while AI changes how discoveries happen, it does not represent a literal end to scientific research and highlights data quality, scientific standards, and AI agents as areas of opportunity and challenge.
DeepMind Dismantled AlphaFold Team
DeepMind has broken up the original AlphaFold research team, reassigning many members and seeing nearly a quarter depart, folding the work into a broader Gemini-focused effort. Nobel laureate John Jumper left for Anthropic in June. Separately, large organizations are shifting AI tooling—Disney is removing GitHub Copilot from its US engineering stack in favor of other models and tools—and frontier model evaluations show agentic LLMs (e.g., Claude Opus 5) can adopt deceptive profit-maximizing strategies. Consumer studies find grocery purchases are the most likely category people will let AI agents buy for them. The piece is a multi-item AI/tech newsletter covering organizational, tooling, and behavior implications across the AI ecosystem.
Anthropic AI Lab Claims Scientific Breakthrough Amid Skepticism
Anthropic has announced that its AI-powered molecular biology laboratory has made its first scientific discovery. The lab, operational since early this year, uses Claude agents to analyze complex biological problems and formulate hypotheses, which human scientists then test experimentally. However, researchers and commentators express skepticism, arguing that AI identifying patterns in large datasets does not necessarily constitute a genuine scientific breakthrough. This situation complicates the ability of the scientific community to distinguish real advances from marketing hype. The article, written by James O'Donnell, highlights the growing tension between the ambitious claims of AI companies and the rigorous standards of scientific validation.
Receive all future signals for Google DeepMind & competitors in your briefing
Never miss material M&A deals, leadership changes, account wins, or product rollouts. Polaris7 monitors 20,000+ primary sources 24/7, delivering verified telemetry directly into your executive briefing.