tp53(ashish)

tp53(ashish)

engineer who secretly wants to be an oncologist

AI 脳 Science

AI 脳 Science

I want a future in which AI agents help people ask better questions, explore more possibilities and turn reliable discoveries into benefits that reach more people. Science and mathematics offer extraordinary opportunities for this kind of collaboration, especially in medicine and biology, where better understanding can ultimately mean less suffering and more healthy life. I am interested in human鈥揂I teams and collaborations across laboratories around the world, with each contributing strengths the others lack.

Progress should deepen human agency and meaning: the freedom to choose what matters, understand the work, challenge an answer and share in discovery. That means evaluating evidence rather than just fluent output, preserving uncertainty and giving people real authority over consequential decisions. It also means designing against accidental harm and deliberate misuse, taking dual-use risks seriously, and making the benefits of scientific capability broadly accessible.

Updated: 2026-09-23

Dates are first public posting/publication where verified; talks use upload dates. Year-only dates retain their original precision. OP means an original author/team post or identified commentary. Missing posts are marked as unverified.

Reading list

Self-Organizing Agent Teams Learn to Reason Together

Frozen teams learn reusable coordination strategies, with gains depending partly on their ability to recognize correct reasoning.

ScientistTwo: Pioneering the Human Knowledge Frontier with Autonomous AI

An autonomous research pipeline combines experiments, ablations and simulated peer review; its reported successes require careful attention to selection and evaluation.

AI in Science: Early Insights

Combines model-use data, a specialist-model inventory and a researcher survey to examine adoption, time savings and verification demands.

AI Infrastructure at Periodic

Describes infrastructure for long scientific tool runs, specialized models and efficient training; performance claims come from the company.

Reflecting on 25 years of cancer research: transformative advances and unmet expectations

Six researchers reflect on cancer biology, translational disappointments and opportunities for human鈥揂I co-science.

Medical school lab scientists get a new partner: AI

Describes medical researchers integrating specialized AI agents with vetted data, laboratory workflows and human oversight.

Latent-space reasoning as a third axis of test-time scaling

Fran莽ois Chollet highlights latent iterations alongside longer reasoning and parallel exploration.

Imagining a new future for science

Eunice Jun explores representations and interfaces that make scientific reasoning more inspectable.

Anthropic uses Claude to formalize proof of Fermat鈥檚 Last Theorem

Reports on AI-assisted formalization of Fermat鈥檚 Last Theorem and the role of shared proof infrastructure.

The paradox at the heart of AI and science

Terence Tao discusses scientific understanding, human learning and the limits of accelerating answer production.

A Case Study on Emergent Cheating and Whistleblowing in Autonomous Research Swarms

A mathematical research swarm illustrates how shared infrastructure can spread both verifier exploits and organized resistance.

Accelerating Scientific Research with Gemini in the Real-World

Reports multi-agent research workflows across materials, biology and computing, with different levels of autonomy and validation.

The scientific paper needs an uncertainty layer

Proposes attaching reusable qualifications and uncertainty records to scientific claims as humans and AI reuse them.

Artificial intelligence in drug discovery鈥攚hat it is, where we stand and the path forward

Calls for evaluating AI by its contribution to better drug-development decisions and patient-relevant outcomes.

Drug Discovery Has No Magic Wands

Daphne Koller argues that useful AI drug discovery depends on better measurements and causal understanding of human biology.

Solipsistic Superintelligence is Unlikely to be Cooperative

Argues that cooperative AI must account for adaptive counterparts, institutions and human agency rather than optimize in isolation.

Some Simple Economics of AGI

Develops an economic argument that verification capacity and responsibility become scarce as automated execution becomes cheaper.

  • Type: Working paper
  • Published: 24 Feb 2026
  • Resource: Some Simple Economics of AGI
  • OP: not yet verified
  • Topics: Human agency, Verification, Economics
  • Related: https://ssrn.com/abstract=6298838

Why LLMs Aren鈥檛 Scientists Yet: Lessons from Four Autonomous Research Attempts

Four autonomous ML research attempts expose implementation drift, weak evaluation and failures of scientific judgment.

Eliciting Research Taste in LLMs through Future Research Direction Choice

Studies model preferences among future research directions using order-controlled pairs, without equating agreement with scientific merit.