frontpage.
newsnewestaskshowjobs

Made with ♥ by @iamnishanth

Open Source @Github

fp.

Open in hackernews

Robots can now learn to use tools just by watching us

https://techxplore.com/news/2025-08-robots-tools.html
29•geox•6h ago

Comments

neom•5h ago
Shouldn't they be able to learn to use tools better than us? https://deepmind.google/discover/blog/genie-3-a-new-frontier...
pixl97•5h ago
Depends how you define that, but the biggest issue is fine motor control. Millions of years of evolution has made us a highly optimized machine.
ck2•5h ago
Someday, maybe 100 years from now, someone is going to figure out a minimal codeset to "bootstrap" machine-learning for a wide range of devices so they basically teach themselves slowly but steadily and then each other

and at first it's going to be absolute amazing, followed by absolutely terrifying

The robots teaching themselves soccer (football) still sticks in my mind - at first they were falling over but after many iterations they had learned to pass the ball sideways and backwards to yet move it forwards to the goal

ivape•4h ago
It’s interesting that China has already thrown a Robot Olympics. We’re laughing about it, but I suspect it will get better and better every year. When China decides winning Gold in everything is a national imperative, watch out.
squigz•4h ago
Why is it terrifying to you?
ck2•3h ago
Watch the Animatrix

(no really, this is not a joke answer, it has some great philosophy among the entertainment)

idiotsecant•2h ago
We are the apex predator. Once there is another entity capable of general purpose learning that is no longer the case. It's an undeniable existential threat.
bamboozled•59m ago
Ever seen Terminator?
pessimizer•4h ago
I'm surprised that you think this it going to be 100 years from now. I think 20 minutes from now would be a better estimate. Or 20 minutes ago; I don't know what people are doing alone in their nasty little labs and notebooks.
MichaelRazum•4h ago
This looks like a nice application to get inductive bias into the model. But I think right now there is no solution to get fine grained motor skills besides tele operating and doing behavior cloning. And even then it is far from perfect..
FirmwareBurner•4h ago
But can a robot learn to drink 3 beers before lunch?
Tade0•3h ago
"Can you?"

In my corner of the world the accepted form is a 100ml bottle of vodka, so hopefully it's flexible in this regard.

Animats•4h ago
This is huge. There's been so little progress on manipulation over the years. The 1960s Stanford AI lab videos don't look much worse than current work.

Part of the trick here seems to be using Gaussian splatting as an intermediate form for vision data. That's a new idea. Gaussian splatting is a brute-force process that doesn't try to generate a full 3D model reconstruction or object recognition. It can be applied to any visual scene where at least two viewpoints are available.

This just might get robotic manipulation unstuck. Good target for VC funding.

How can AI ID a cat?

https://www.quantamagazine.org/how-can-ai-id-a-cat-an-illustrated-guide-20250430/
38•sonabinu•3d ago•6 comments

Line scan camera image processing for train photography

https://daniel.lawrence.lu/blog/y2025m09d21/
102•dllu•5h ago•18 comments

A 2k-year-old sun hat worn by a Roman soldier in Egypt

https://www.smithsonianmag.com/smart-news/a-2000-year-old-sun-hat-worn-by-a-roman-soldier-in-egyp...
30•sensiquest•2h ago•6 comments

Static sites with Python, uv, Caddy, and Docker

https://nkantar.com/blog/2025/08/static-python-uv-caddy-docker/
50•indigodaddy•1d ago•31 comments

Manim: Animation engine for explanatory math videos

https://github.com/3b1b/manim
350•pykello•14h ago•61 comments

Acronis True Image costs performance when not used

https://randomascii.wordpress.com/2025/05/26/acronis-true-image-costs-performance-when-not-used/
43•juanviera23•3d ago•8 comments

Librebox: An open source, Roblox-compatible game engine

https://github.com/librebox-devs/librebox-demo
200•libreboxdevs•10h ago•56 comments

Motion (YC W20) Is Hiring Senior Software Engineers

https://jobs.ashbyhq.com/motion/7355e80d-dab2-4ba1-89cc-a0197e08a83c?utm_source=hn
1•ethanyu94•1h ago

RFC 9839 and Bad Unicode

https://www.tbray.org/ongoing/When/202x/2025/08/14/RFC9839
198•Bogdanp•9h ago•95 comments

Children of the Geissler Tube

https://www.hopefulmons.com/p/children-of-the-geissler-tube
25•paulkrush•3d ago•2 comments

Optimizing FizzBuzz in Rust

https://github.com/nrposner/fizzcrate
15•Bogdanp•2h ago•4 comments

Writing Speed-of-Light Flash Attention for 5090 in CUDA C++

https://gau-nernst.github.io/fa-5090/
118•dsr12•9h ago•17 comments

Texas Instruments' $60B U.S. project, the next iPhone chips fabric

https://www.cnbc.com/2025/08/22/apple-will-make-chips-at-texas-instruments-60-billion-us-project....
78•giuliomagnifico•1d ago•53 comments

Optimizing our way through Metroid

https://antithesis.com/blog/2025/metroid/
87•eatonphil•1d ago•8 comments

A simple way to generate random points on a sphere

https://www.johndcook.com/blog/2025/05/06/random-points-on-a-sphere/
48•piinbinary•4d ago•35 comments

Exploring EXIF (2023)

https://hturan.com/writing/exploring-exif
24•jxmorris12•2d ago•3 comments

Practical approach for streaming UI from LLMs

https://www.timetler.com/2025/08/19/unlocking-rich-ui-components-in-ai/
38•scranglis•4d ago•2 comments

I made a floppy disk from scratch

https://kottke.org/25/08/i-made-a-floppy-disk-from-scratch
167•bookofjoe•11h ago•67 comments

The Amiga games and demo scene collection

https://amiga.vision/
7•doener•1h ago•0 comments

450× Faster Joins with Index Condition Pushdown

https://readyset.io/blog/optimizing-straddled-joins-in-readyset-from-hash-joins-to-index-conditio...
86•marceloaltmann•4d ago•28 comments

Recreationally overengineering my location history

https://overengineer.dev/blog/2025/08/19/overengineering-location-history/
15•kickofline•3d ago•2 comments

Romhack.ing's Internet Archive Mirror No Longer Available

https://romhack.ing/database/news/entry/DW8BKnRHSEqaGDwXTiKjMw
5•pharrington•1h ago•0 comments

Rethinking the Linux cloud stack for confidential VMs

https://lwn.net/Articles/1030818/
102•Bogdanp•10h ago•41 comments

Reproducing prospect theory with 'differentiable decision theories'

https://www.science.org/doi/full/10.1126/science.abe2629
15•timshell•2d ago•4 comments

Ergonomic errors in Rust: write fast, debug with ease, handle precisely

https://gmcgoldr.github.io/2025/08/21/stackerror.html
3•garrinm•1d ago•2 comments

Stepanov's biggest blunder? The curious case of adjacent difference

https://mmapped.blog/posts/43-stepanovs-biggest-blunder
17•signa11•2d ago•1 comments

Lightning declines over shipping lanes following regulation of sulfur emissions

https://theconversation.com/the-world-regulated-sulfur-in-ship-fuels-and-the-lightning-stopped-24...
204•lentoutcry•4d ago•52 comments

Developer's block

https://underlap.org/developers-block/
166•todsacerdoti•12h ago•89 comments

Determinants and causal effects of admission to selective private colleges [pdf] (2023)

https://www.nber.org/system/files/working_papers/w31492/w31492.pdf
21•EvgeniyZh•11h ago•22 comments

WebR – R in the Browser

https://docs.r-wasm.org/webr/latest/
128•sieste•4d ago•27 comments