Statistics, medicine, humanity

Thoughts, essays, etc by Samuel Weisenthal

Math is everywhere in medicine

10/12/2026
Outcome-based evaluations

10/1/2026
LLMs are estimators

9/13/2026
What’s in a name?

9/13/2026
Math and medicine, follow up

8/3/2026
Medical chatbots that draw upon trusted literature still need inaccuracy disclaimers

7/28/2026
Math and medicine

7/22/2026
AUC of 1, sensitivity of 0.70?

7/13/2026
Journals

7/13/2026
Misinformation, groupthink, and the statistician

7/5/2026
D-dimer: not bad, just misunderstood

6/15/2026
The information mismatch

6/11/2026
Likelihoodist inference

6/2/2026
First, do no harm?

6/1/2026
UWorld is not a hobby

5/31/2026
Reasoning, clauses, and terminology

5/27/2026
What shared decision making is not

5/15/2026
On the AI Lie

5/14/2026
The AI Lie

5/13/2026
Utility: difficult to measure, easy to set…

5/6/2026
Imitation, AGI, and medical decision making

5/5/2026
Metaphysics?

5/2/2026
Generative LLMs and medical triage decision making

4/22/2026
A consensus on causality?

4/9/2026
Subjective hypocrisy

3/29/2026
Why work with a statistician?

3/25/2026
The match

3/15/2026
Automation vs. invention

3/13/2026
Action-conditioned world models

3/13/2026
IWD26

3/8/2026
Safe use of generative LLMs in medicine

2/10/2026
Ouroboros

2/6/2026
Prediction models and counterfactuals

1/25/2026
Fixed-length

1/19/2026
Policies, adjustment, decisions, sharks

1/3/2026
On the acceptable range of model calibration

12/12/2025
GPT-5 and medical decisions

12/1/2025
Decision analysis and medical research

11/9/2025
Politics again

11/2/2025
Decision analysis and insurance

10/19/2025
Image processing, chatbots, and “AI”

9/28/2025
Decision analysis and the pharmaceutical industry

8/25/2025
Should chatbots make medical decisions?

7/3/2025
Rejoice on error

6/23/2025
Standing together

6/14/2025
Ode to an appendix

6/1/2025
Keep the real data in the loop

5/25/2025
Time-evolving reports

5/16/2025
Support for reporters

5/11/2025
Scroll Breaker

5/9/2025