Kuinox
Kuinox on Hacker News
I answered about tokens being apparently something not measurable, I do agree that there is a lack of transparency on the subscription usage.
> "your usage will be measured in tokens, which we do not define, nor give you a way to measure them
They have an open source tokenizer on GitHub, they indicate which tokenizer they use per model, and they have a simple to use website that tokenize text and visualize it.
I dont read the argument as specifically AI being bad at the physical world but more that they can't be creative, physical world or not
> It’s not dead. Some companies use it while I believe the majority of use cases are addressed by open-weight rivals.
The quality of the answers on my datasets are in another league compared to the benchmarks.
I have a mixed language dataset, Jev answers only lack some accuracy, the open weight rivals output just look like random noise
This is the same thing as "but AI can only repeat what it already saw" but the goalpost moved again.
> Not in rural Europe it isn't.
> There is still no widespread charging infra.
You can charge your car at home on a simple reinforced 230V socket.
Even city car which should be bad for this purpose have at least 200km of range.
AI are excellent at ingesting tons of data and getting something relevant from what you asked.
I ask it about linux behavior and it will dig through hundreds of patches & mails and will return me informations sourced with links to the patches, mails containing it's affirmations.
Codex found the root cause:
> The site’s graph incorrectly uses 32 texture bindings, exceeding its WebGPU device’s limit of 16.
> Firefox 153 mishandles that failed rendering operation and leaves the compositor waiting
So a firefox + website bug
With an updated firefox, the website works except the mindmap on the top right
> Uncaptured WebGPU error: Too many bindings of type SampledTextures in Stage ShaderStages(FRAGMENT), limit is 16, count was 32. Check the limit `max_sampled_textures_per_shader_stage` passed to `Adapter::request_device`
Works on phone but not on my linux desktop.
I cant select text nor click links on this page with firefox.
There is more to the browser, it can give you a remote desktop access to the browser.
So it's not completely headless.
But when I tried it, the remote control completely froze after a few seconds.
It's incredibly slow for a single core, that's my reference point.
1.4 billions requests, 258 160 cpu hours.
That's 1.5 requests per second ?
I'm starting to believe, the issue is more that their software is not well optimized.
Claude did the changes, I didn't do anything, I think it's was mostly recompiling with flags.
I did a similar thing recently.
I run kube in a VM, and it was ressource constrained.
I hooked an old samsung to my pc and have let claude handle everything.
I only had to unlock the bootloader, flash the image claude uploaded.
It rebuilt postmarketos multiples times to enable various flag, and now there is a kube node running on it.
It doesn't rotate on firefox on my linux machine, but does on firefox on my windows.
edit: it looks like my firefox linux have a "prefers-reduced-motion: reduce" set...
So I discovered that setting animation to instant on kde, tell the browser to reduce motions...
I jumped directly to insane and completed it on the first try. It's too easy.
No, the value of scrapping the page is long gone and the scrapper would have moved on.
The users would still access the page by waiting a bit.
The difficulty is increased when the server get heavy traffic.
I just generated some docs for a lib I'm writing and it says:
> 9 % machine likely human-written
The stackoverflow moderation is the reason I do not post on it.
You have middle party with no competence on the technology trying to do useless moderation.
Instead I directly go on the project github page and ask the question directly to the mainteners.
Oh it's because the bench is lying.
You need to pass each level without failing, if you fail a level, it count as you "lost" the minigame.
The fact it start to get a score means it managed to get a 100% score on one of the minigame.
How so, are you purely speculating or you found a hole in the zero knowledge proof system some countries are implementing ?
> the lag time between hitting the button and getting your money back can be weeks.
Nope, it's as soon as the bank is aware. In your rush to correct me, you failed to understand that there are simply better customers protection here which make a product such as a credit card mostly useless.
I always find it funny that americans think the button to oppose a payement doesn't exists on my banking app because I have a debit card and not a credit card.
yes the ps3 is now considered a retro console.
Yes this is kinda my point.
Instead of having a few projects/org, it's a constellation of packages too small, it's impossible to know who you depend on when adding a dependency.
The biggest problem is not software but culture, not at npm, but in the js ecosystem.
The js ecosystem is simply a juicy targets, the attack surface is enormous.
The attacker can make their attack more sophisticated,
there will always be a maintainer that can seed the worm spread.
Meanwhile in the nuget ecosystem is way smaller and have way less mainteners involved for a single given dependency.
nuget have targets, and allow to run code on build, it doesn't have this problem because there is less dependencies.