Anthropic and Citrini Research use an internal Claude model to autonomously generate protein minibinders against five scorpion neurotoxins, testing 211 designs
THE SO WHAT
Claude autonomously generating and yielding testable protein minibinders against multiple scorpion neurotoxins is another proof point that frontier models are becoming general-purpose design tools for biology, not just language. If you’re in therapeutics or bio-platforms, assume model-driven candidate generation will compress early discovery timelines and shift your edge toward data quality, wet-lab throughput, and regulatory execution.
READ THE SOURCE
MORE FROM THE WIRE
Applied AISpaceX revenue rises 161% to $14.6B as AI business overtakes Starlink
When a launch company books more revenue from AI than from Starlink, space infrastructure is becoming an AI infra business by proxy—compute and data are now as core as rockets. If you depend on GPU access, assume nontraditional providers with their own power and real estate will be credible alternatives to hyperscalers within a few years.
Applied AIWhen technology becomes part of the talent infrastructure
If AI-ready PCs and device strategy are now framed as “talent infrastructure,” IT spend is quietly becoming a retention lever, not just a cost center. HR, IT, and line leaders should be co-owning device standards and AI access, because it will show up in both engagement scores and output.
Applied AIDemis Hassabis says he expects AGI within two to three years, down from his three-to-five-year prediction earlier in 2026, and discusses AI's science potential
When a leading lab compresses its AGI timeline from 3–5 years to 2–3, your planning horizon just got pulled forward whether you believe the forecast or not. Treat this as a scenario input: stress-test your IP, regulatory, and workforce assumptions against a world where frontier capability steps up meaningfully before 2030.
Applied AIShould you get a personal AI assistant?
Personal AI assistants are shifting from novelty to a real line item in individual productivity stacks. For orgs, the risk is shadow-AI—employees will adopt their own agents unless you provide sanctioned options with clear data boundaries.