Recent submissions

Comment · 1 day ago
wow. Naive perhaps but I would have thought that binance and coinbase would have insurance funds equal to the exposure of their hot wallets.

idk to me that just seems intuitively like something you'd need to have to be credible? Cold is a different story...but how can hot not be covered?

Comment · 1 day ago
I'll never not cheer for Mistral.

The more countries are in this game the better for us all.

Much like other stuff when there isn't enough competition the consumer gets wrecked

Comment · 2 days ago
Fair point. Not entirely convinced a modern car plant could be retrofitted for tanks/military but I guess even 4x4s have value
Comment · 2 days ago
Meanwhile my paid enterprise plan has Gemini 3.1 pro in chat.

Thats on a 22 bucks plan.

idk wtf Google is doing with their plans but really need to make an effort to put some distance between myself and that circus

Comment · 2 days ago
I do wonder what the US game plan is here long term. This sort of protectionism can work temporarily but isn’t really sustainable.

And it’s not like GM and friends are catching up. If anything the gap is widening rapidly.

Comment · 2 days ago
The picture has a pilot included - suggests this is a pretty small plane with very stubby wing.
Comment · 3 days ago
10m seems quite high for ground effect?
Comment · 3 days ago
Now if only I could sign up for a pro plan…

Unfortunately their plan structure is on par with their model naming.

The plans that allow for own domain and the plans that allow for sane AI use are mutually exclusive

Comment · 3 days ago
> If enough of us do this, it will work.

Tech scene can’t even muster enough collective consensus to slow a „this is potentially dangerous“ scenario let alone a vague eroding of humanity.

Still I’m all for people’s right to strike for whatever reason they like

Comment · 3 days ago
State resources and armed officers taking a political stance like this is how democracy withers
Comment · 3 days ago
> GLM-5.3 (max) score highest at 32, and Claude Haiku 5.5 (max) outscores all of them

That seems improbable

Comment · 3 days ago
How do you determine confidence levels? logprobs?
Comment · 3 days ago
Comment · 4 days ago
Notice the linguistic thing while trying to dictate prompts. I can apparently type and listen to podcasts easily but not talk with podcasts on headphones. Brain just nopes
Comment · 4 days ago
Same as the work produced by bad employees. Gets absorbed, fixed, ignored or thrown out.

It’s not like companies never shipped broken shit before AI

Live from the official Hacker News API

A modern reader for Hacker News · SvelteKit + Skeleton