frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

Getting AI to write good SQL

https://cloud.google.com/blog/products/databases/techniques-for-improving-text-to-sql
216•richards•7h ago•112 comments

Wow@Home – Network of Amateur Radio Telescopes

https://phl.upr.edu/wow/outreach
32•visviva•2h ago•0 comments

Thoughts on thinking

https://dcurt.is/thinking
327•bradgessler•9h ago•223 comments

Show HN: KVSplit – Run 2-3x longer contexts on Apple Silicon

https://github.com/dipampaul17/KVSplit
207•dipampaul17•8h ago•32 comments

Postman for MCP

https://usetexture.com/##
17•andes314•2h ago•3 comments

XTool – Cross-platform Xcode replacement

https://github.com/xtool-org/xtool
28•TheWiggles•2h ago•11 comments

A Research Preview of Codex

https://openai.com/index/introducing-codex/
389•meetpateltech•13h ago•311 comments

Rustls Server-Side Performance

https://www.memorysafety.org/blog/rustls-server-perf/
70•jaas•3d ago•10 comments

MIT asks arXiv to withdraw preprint of paper on AI and scientific discovery

https://economics.mit.edu/news/assuring-accurate-research-record
263•carabiner•13h ago•142 comments

Coding agent in 94 lines of Ruby

https://radanskoric.com/articles/coding-agent-in-ruby
68•radanskoric•2d ago•21 comments

ClojureScript 1.12.42

https://clojurescript.org/news/2025-05-16-release
110•Borkdude•8h ago•20 comments

A Linux kernel developer plays with Home Assistant: general impressions

https://lwn.net/SubscriberLink/1017720/7155ecb9602e9ef2/
5•pabs3•1h ago•0 comments

I'm Peter Roberts, immigration attorney, who does work for YC and startups. AMA

192•proberts•13h ago•356 comments

WebGL Gray-Scott Explorer (2012)

http://www.mrob.com/pub/comp/xmorphia/ogl/index.html
16•joebig•3h ago•1 comments

Show HN: Visual flow-based programming for Erlang, inspired by Node-RED

https://github.com/gorenje/erlang-red
205•Towaway69•13h ago•89 comments

X X^t can be faster

https://arxiv.org/abs/2505.09814
165•robinhouston•12h ago•47 comments

MCP: An in-depth introduction

https://www.speakeasy.com/mcp/mcp-tutorial
34•ritzaco•3d ago•12 comments

The Joys of Discovering the Roman Underground

https://www.smithsonianmag.com/travel/the-joys-of-discovering-the-roman-underground-from-the-colosseum-to-whats-beneath-the-trevi-foundation-180986626/
3•ulrischa•1d ago•0 comments

They Were Identical 'Twinnies' Who Charmed Orwell, Camus and More

https://www.nytimes.com/2025/05/04/books/review/the-dazzling-paget-sisters-ariane-bankes.html
20•lermontov•5h ago•4 comments

Will AI systems perform poorly due to AI-generated material in training data?

https://cacm.acm.org/news/the-collapse-of-gpt/
65•pseudolus•4h ago•68 comments

New high-quality hash measures 71GB/s on M4

https://github.com/Nicoshev/rapidhash
15•nicoshev11•2d ago•4 comments

Show HN: Roast My Dish – AI roasts your food photos with brutal honesty

https://www.roastmydish.online/
12•romeumaleiane•3h ago•6 comments

Java at 30: Interview with James Gosling

https://thenewstack.io/java-at-30-the-genius-behind-the-code-that-changed-tech/
170•chhum•15h ago•251 comments

The first year of free-threaded Python

https://labs.quansight.org/blog/free-threaded-one-year-recap
249•rbanffy•18h ago•252 comments

Material 3 Expressive

https://design.google/library/expressive-material-design-google-research
331•meetpateltech•3d ago•513 comments

Show HN: Solidis – Tiny TS Redis client, no deps, for serverless

https://github.com/vcms-io/solidis
38•jayl-e-e•7h ago•3 comments

Behind Silicon Valley and the GOP’s campaign to ban state AI laws

https://www.bloodinthemachine.com/p/de-democratizing-ai
19•spenvo•1h ago•2 comments

The Magic Hours: The Films and Hidden Life of Terrence Malick

https://www.lrb.co.uk/the-paper/v47/n09/david-thomson/cool-tricking
53•mitchbob•10h ago•31 comments

Foundry (YC F24) Is Hiring – Founding Engineer (ML × SWE)

https://www.ycombinator.com/companies/foundry/jobs/uwi8b6I-founding-engineer-ml-x-swe
1•lakabimanil•11h ago

Returning to My Roots in Hardware

https://dancrimp.nz/2025/03/31/hardware/
69•dcrimp•3d ago•11 comments
Open in hackernews

ChatGPT can accurately estimate Body Fat Percentage from photos

https://annaleptikon.substack.com/p/can-chatgpt-accurately-estimate-body
34•Luc•9h ago

Comments

zerohp•8h ago
Methodology doesn't support the conclusion.

Try using images that it wasn't trained on.

3vidence•8h ago
I would have to agree, this feels like fake science. They don't even seem to really present a reasonable initial problem.

This seems like a man / women / skinny / fat detector.

If you just classified people into those categories I wonder if you would achieve a similar level of accuracy.

xk_id•1h ago
It would be great if people stopped submitting misleading ads for the AI industry.
MostlyStable•8h ago
So that is obviously a significant relationship between ChatGPT estimate and the Henselman Guide BFP numbers. But just being a significant relationship doesn't necessarily mean it's useful. I'd bet that BMI (which the author repeatedly mentions is outdated and not very useful), is almost certainly also significantly correlated with BFP values.

Eyeballing that plot, it looks like ChatGPT estimates are +/- 10%. I'm wondering what the cutoff for usefulness is. I wouldn't' be surprised if BMI is that level of correlated or better.

cluckindan•8h ago
The error was analyzed in the text.

Men: median signed error: +0.8%, median absolute error: 2.4%

Women: median signed error: +3.5%, median absolute error: 5.7%

MostlyStable•8h ago
That doesn't do anything to address my concerns. I was talking about the total spread, not the mean error, but in either case, a comparison to how well BMI correlated would be needed to address the issues I raised.
orionsbelt•8h ago
I just fed 4o and o4 mini a photo of my torso and gave it my height and asked it to estimate my body weight and body fat %. It was 20-40 lbs too low, so way off.
OtherShrezzing•8h ago
Given the images and text are definitely in the model training set, I'm surprised the results are this inaccurate.
ulf-77723•7h ago
Dexa is expensive, you can get a modern smart scale for 50€, the cheapest accurate method would be the Caliper tongs.

What can it be good for? Video surveillance with metadata enrichment? Profiling? Losing body fat takes time. A smart mirror that displays body fat?

BrandoElFollito•7h ago
Just by looking at the graph I know that this will be rubbish - because of the line

When you draw a line this means that your model expects a linear relationship between the variables. Otherwise this is either a feel good line, or incompetence.

donclark•6h ago
Can AI estimate psychological profile based on a photo of the face (face structure)?