Hacker News: brendanashworth

New comment by brendanashworth in "OpenAI Sells Statsig to Amplitude"

brendanashworth — Tue, 05 May 2026 21:25:32 +0000

This link gives 404 Page not found, perhaps https://amplitude.com/blog/amplitude-and-statsig-partnership is a better URL [1]

[1] https://news.ycombinator.com/item?id=48028790

Amplitude and Statsig Partnership

brendanashworth — Tue, 05 May 2026 21:23:34 +0000

Article URL: https://amplitude.com/blog/amplitude-and-statsig-partnership

Comments URL: https://news.ycombinator.com/item?id=48028790

Points: 10

# Comments: 2

An intriguing case of "exceptional resilience" against dementia

brendanashworth — Mon, 27 Apr 2026 14:59:50 +0000

Article URL: https://peterattiamd.com/heat-shock-proteins-and-alzheimers-mutation/

Comments URL: https://news.ycombinator.com/item?id=47922534

Points: 1

# Comments: 0

New comment by brendanashworth in "Recursive self-improvement is possible for apps, too"

brendanashworth — Wed, 15 Apr 2026 23:53:10 +0000

Out of the box, Opus/GPT-5.4 uses checks like `tsc --noEmit` or `python -m py_compile ...`, so compile checks on a Swift/Objective-C app probably get you pretty far. Also just setting up a Dockerfile maybe with swift:5.10-focal would give the agent the right tools to verify its own work.

Sandboxing isn't about having the perfect devex environment, most dev sandboxes (looking at codex cloud) don't actually have full verification available. Sometimes Github actions CICD is all you can get!

Recursive self-improvement is possible for apps, too

brendanashworth — Wed, 15 Apr 2026 21:27:58 +0000

Article URL: https://www.codecanary.ai/blog/recursive-self-improvement-is-possible-for-apps

Comments URL: https://news.ycombinator.com/item?id=47785490

Points: 5

# Comments: 2

New comment by brendanashworth in "Show HN: Steerling-8B, a language model that can explain any token it generates"

brendanashworth — Tue, 24 Feb 2026 03:25:20 +0000

Is there a reason people don't use SHAP [1] to interpret language models more often? The in-context attribution of outputs seems very similar.

[1] https://shap.readthedocs.io/en/latest/

Effects of Full-Fat and Fermented Dairy Products on Cardiometabolic Disease

brendanashworth — Thu, 05 Feb 2026 03:06:41 +0000

Article URL: https://www.sciencedirect.com/science/article/pii/S2161831322004367

Comments URL: https://news.ycombinator.com/item?id=46895146

Points: 5

# Comments: 1

New comment by brendanashworth in "Use of Bayesian methodology in clinical trials of drug and biological products [pdf]"

brendanashworth — Thu, 15 Jan 2026 22:48:36 +0000

If choosing a Bayesian approach in a clinical trial can reduce the number of recruited subjects, I would imagine the pharma industry is strongly incentivized to adopt it.

Use of Bayesian methodology in clinical trials of drug and biological products [pdf]

brendanashworth — Thu, 15 Jan 2026 07:36:15 +0000

Article URL: https://www.fda.gov/media/190505/download

Comments URL: https://news.ycombinator.com/item?id=46629295

Points: 71

# Comments: 22

Salesforce Acquires Qualified

brendanashworth — Thu, 18 Dec 2025 03:40:57 +0000

Article URL: https://www.salesforce.com/news/stories/salesforce-signs-definitive-agreement-to-acquire-qualified/?bc=HL

Comments URL: https://news.ycombinator.com/item?id=46308698

Points: 3

# Comments: 1

New comment by brendanashworth in "AI beyond LLMs: a wearable foundation model based on JEPA"

brendanashworth — Wed, 10 Dec 2025 19:46:58 +0000

Would a wearable model like this gain in predictive power by adding FHIR/EHR inputs? CoMET [1] has great AUC even without daily/lifestyle metrics like steps, awakenings at night, or vo2max estimates (which are present in say apple health). It would be interesting to know how much you would gain from having a holistic model that sees both inputs.

[1] https://arxiv.org/pdf/2508.12104v1