frontpage.
newsnewestaskshowjobs

Open Source @Github

fp.

Open in hackernews

Scaling Discovery Through Test-Time Communication

https://arxiv.org/abs/2609.21032
1•marojejian•49m ago

Comments

marojejian•26m ago
This is one big aspect of the ML revolution that I'm uncertain about. Clearly the future system of how models communicate will have a huge impact on their effectiveness and the form the world society will take. The older I get the more I appreciate how the collective intelligence and society of humans matters more than the level of individual idealized intelligence.

Only working together can we build a cheap toaster: https://en.wikipedia.org/wiki/Thomas_Thwaites_(designer)

I buy the Sapiens argument that our dominance is due to being more social, than being "smart" individually.

On the other hand big dysfunctional group can hobble the effectiveness of even its best members (I bet I don't need to cite an example for you).

But the opportunity is far less constrained than this. The space of human societies is limited to agents who all have similar structures and relaitvely fixed "selves" and goals. ML agents can be as diverse as we like, can change their own nature on a dime, and can have any type of self "identify" or goals imaginable. They are not constrained by any "human nature" to only cooperate in certain ways, but can dynamically pursue any strategy at all.

Perhaps there is a natural convergence or local equilibrium with more stable and uniform identities. But I'd bet on the other side - that there will be an insanely diverse and dynamic society that will be able to coordinate in ways that we can neither compete with nor conceive of. Cryptic message boards are just the first drop in the deluge.

The Rise and Fall of Tech Worker Power

https://www.bostonreview.net/articles/the-rise-and-fall-of-tech-worker-power/
1•mooreds•27s ago•0 comments

Ask HN: Honey Pots in 'who wants to be hired'?

1•jwally•51s ago•0 comments

Betel nuts are a top mind-affecting substance. There's evidence of ancient use

https://text.npr.org/g-s1-144054
1•mooreds•53s ago•0 comments

LiveCrew – a persistent AI exec team that argues back (with receipts)

https://www.livecrew.tech/
1•shamil_b•1m ago•0 comments

Jev Creator: System One Models for Prod, Not God [video]

https://www.youtube.com/watch?v=cFx9Z3ZXca0
1•Anon84•1m ago•0 comments

Show HN: RUUN – Tor Browser with a second, non-Tor mode in the same app

https://gitlab.com/varad_7/ruun
1•opentestudox07•2m ago•0 comments

Ask HN: Which is cheaper: ChatGPT tokens or a 2nd Pro 20x subscription?

2•yesitcan•3m ago•0 comments

Untangling Lifetimes – Arena Allocators in C

https://www.dgtlgrove.com/p/untangling-lifetimes-the-arena-allocator
1•eliasebner•3m ago•0 comments

Nano Banana 2.5: Faster, Better AI Images

https://nanobanana25ai.net
1•yoxi1•4m ago•0 comments

Fast-CLI: CLI tool written in Zig for testing internet speed

https://github.com/mikkelam/fast-cli
1•hoyd•5m ago•0 comments

Sauna Jungle Expansion

https://www.saunajungle.com/
1•hubears•5m ago•0 comments

GPT-6 Sol and Luna

https://openai.com/index/introducing-gpt-6-sol-and-luna/
74•OfficialTurkey•8m ago•19 comments

Cookie started its life as a plastic bottle

https://www.acs.org/pressroom/presspacs/2026/august/this-cookie-started-its-life-as-a-plastic-bot...
1•dustfinger•9m ago•0 comments

GPT-6 Sol

https://developers.openai.com/api/docs/models/gpt-6-sol
6•sfkgtbor•10m ago•1 comments

Arguing about Arguments

https://steveklabnik.com/writing/arguing-about-arguments/
1•birdculture•10m ago•0 comments

Delaying all JavaScript doesn't make your site faster

https://michalek.blog/blog/delay-all-javascript
1•absolutmachal•11m ago•1 comments

It's Hard to Learn from Machines

https://carlkolon.com/2026/09/22/learning-from-machines/
1•cckolon•14m ago•0 comments

U.S. Climate Collection

https://usclimatecollection.org
2•Tomte•15m ago•0 comments

Prominent art therapist quoted by Vice, Forbes and others is AI-generated

https://pressgazette.co.uk/news/prominent-art-therapist-quoted-by-vice-forbes-and-others-is-ai-ge...
1•ilamont•16m ago•0 comments

Vibe code what prints money

https://buildradar.leapd.ai
2•javeedshah•16m ago•0 comments

OpenCode removed usage transparency after a billing bug was reported

https://github.com/anomalyco/opencode/issues/39822
2•bigprof•16m ago•0 comments

Reviewing Your Blog Post with an LLM

https://writethatblog.substack.com/p/reviewing-your-blog-post-llm
1•cyndunlop•18m ago•0 comments

My Weird New Hobby: Wandering Around Tokyo on Google Maps

https://ahmedhossamdev.com/writing/my-weird-new-hobby-wandering-around-tokyo/
3•ahmedhossamdev•19m ago•0 comments

U.S. Site Blocking Bill Adds VPNs to the List of Blocking Intermediaries

https://torrentfreak.com/u-s-site-blocking-bill-adds-vpns-to-the-list-of-blocking/
3•Cider9986•20m ago•0 comments

PyTauri – Tauri Bindings for Python

https://pytauri.github.io/pytauri/0.8/
1•amai•20m ago•0 comments

How Healthy Are Sardines?

https://www.nytimes.com/2026/09/14/well/eat/sardines-health-benefits-recipes.html
3•mooreds•21m ago•0 comments

Graduating into Disruption: Labor Market Outcomes for AI-Exposed College Majors

https://www.census.gov/library/working-papers/2026/adrm/CES-WP-26-56.html
1•mooreds•21m ago•0 comments

Goolsbee's Remarks at Monetary and Financial Institutions Forum Event [pdf]

https://www.chicagofed.org/-/media/publications/speeches/2026/9-21-26-omfif.pdf?sc_lang=en&hash=9...
2•mooreds•22m ago•0 comments

Porsche CEO assures staff no further job cuts planned, according to memo

https://www.reuters.com/business/world-at-work/porsche-ceo-assures-staff-no-further-job-cuts-plan...
2•frank_clover•22m ago•1 comments

Sentience, create your digital self

https://sentience.com/
2•tomaitagaki•23m ago•1 comments