AI-assisted review finds code sharing rare in clinical prediction model research
EricTopol · x · 2026-10-09
Eric Topol highlights two items:
- Immunity and aging: A new @NatureAging study shows anti-cancer immunity declines with age, underscoring the role of the thymus and naive T cells — especially in middle-aged men with lower androgen availability (extrapolated from animal models).
- AI-assisted scoping review (Nature Medicine, Oct 2026; Sounack, Celi et al.): quantifies code-sharing practices in clinical prediction model research, finding analytical code availability remains limited despite its importance for reproducibility and generalizability. Results will inform the TRIPOD-Code reporting guideline on code availability.
More from AGI Musings
- You may just be the loading screen: on the obsolescence paradox of humans in the AI era — YogeshMalik · 2026-10-09
- Guardian: AI in ads could deny Britain its next Ridley Scott or Guy Ritchie — nordicinst · 2026-10-09
- Veteran PM: One Hour of Vibe Coding Beats Weeks of Corporate Dev, Bureaucracy Is the Real Blocker — NickPassig · 2026-10-09
- Eric Topol and global team publish paper on ethical principles for medical AI — EricTopol · 2026-10-09
- Rereading a 1999 malware newsletter: AI safety discourse may age just as fast — sebkrier · 2026-10-09
- Mathematicians urged to feed AI all their data: 'Our egos should not stop progress' — iruletheworldmo · 2026-10-09