

ah right, and my eyes need to be recreated because they can’t see ultraviolet


ah right, and my eyes need to be recreated because they can’t see ultraviolet


Barely. Even with the code and seeds, it’s still a struggle to do that. There’s plenty of questions from people running pytorch and tensorflow models that can’t reproduce results. Maybe you isolate enough variables that consecutive runs actually produce the same output, but the study is about commercial models. You’ll never get deterministic output from those.


You’d expect the same answer each time. It’s the same photo, the same model, the same question. But you won’t get the same answer.
I don’t know what ads show that, but anyone who knows the first thing about LLMs knows you don’t get the same answer twice.
I’d get this expectation 5 years ago when most people weren’t familiar with it, but come on… you don’t need to feed it an image 500 times to see that.


Waste of energy. It’s like asking a person to estimate a non-trivial angle. Either use a model trained for that task, or don’t bother.


Exactly, it’s only an improvement until they’re bought and we’re all in the same boat again. We need a federated forge and open standards.


tl;dr clickbaity title for an article that cites xitter and tries to sell their security solutions


yeah, that was cringy
the world’s most-followed person
most followed on his own echo chamber :clap: :clap: :clap:


they are open weight and have a whitepaper, that’s already vastly better than whatever openai and anthropic are doing


it is, but it’s initially opt-in; meaning the system won’t have AI features, but once it does, you’ll remove them by removing snaps
it’s confusing because the paragraphs they talk about opt-in and removing snaps are different, but 26.10 won’t ship with AI features in a fresh install.


why would that only affect “some” countries


at this rate, it will be


I’m interested in setting it up, are you using vs code? Which extension or editor?


watch me go back to debugging like a real engineer: copying and pasting from stack overflow


Users on annual Pro or Pro+ plans will remain on their existing plan with premium request-based pricing until their plan expires, however, model multipliers will increase on June 1 (see table).

holy shit, 9x the previous cost. which was already not great. I was on the fence about cancelling it, but thanks for making up my mind, MS


It’s still matrix multiplication. Running it on a general purpose CPU is inefficient.


Reminded me of that white christmas black mirror episode that your conscience can be virtualized and simulated so years can pass in a minute.


The standard edition of Assassin’s Creed Black Flag Resynced will be available for €59.99/$59.99
are there people out there paying full price for a 13 year old game with better textures?


On Netflix 4k streaming for example, we’re getting over 20 hours of battery life
that’s pretty sick


there have been enough reasons to ditch nova in the past years, in case someone is still using it
It was actually about 90 minutes
only if they installed bw in that time window
Otherwise yes, I agree it’d be better if the CLI was written in a non-JS/TS ecosystem. Perhaps Rust or Go. And the criticisms to
listincluding secrets are super valid.