EmbeddingGemma 2: an open, lightweight multimodal embedding model
deepmind.google·6d ago
AI Engineering
MIT Technology Review·13d ago·5 min read
The Download: climate tech companies to watch and AI’s discovery problem
TL;DR
MIT Technology Review is set to release its 2026 list of Climate Tech Companies to Watch, highlighting ten innovative firms tackling climate change. These companies are making significant advancements in areas like energy storage and nuclear power despite challenging political climates.
✦ Why It Matters
Explore the upcoming list to identify potential partners or technologies for your own climate initiatives.
Key Takeaways
- Stay informed about AI's evolving role in scientific research.
- Engage with AI tools to enhance your own discovery processes.
- Critically evaluate claims of AI-driven discoveries in your field.
- Collaborate with biologists to understand the implications of AI findings.
Full Summary
Anthropic recently announced that its AI agents identified a previously uncatalogued pattern associated with an enzyme, which they claim is reminiscent of the breakthrough that led to CRISPR gene-editing technology. However, this announcement has sparked criticism from biologists who argue that simply recognizing a pattern does not equate to a genuine scientific discovery.
Some researchers have pointed out that they had already discovered the same pattern, raising concerns about whether Anthropic's AI had learned from their prior discussions. This situation underscores the complexities of attributing discoveries to AI, as what may seem novel to an AI system can be routine or inconsequential to human experts.
The debate emphasizes the need for clarity in how we define and recognize scientific contributions made by AI technologies.