dream_weasel

AI/llm is just a statistical word predictor.

You could perhaps make the argument that it’s a statistical token predictor, but that is about as useful as boiling down various weather services as “just statistics” or the economy as “just statistics”.

Making a language model that speaks a language is not that hard, but the world of science underlying how this is done is anything but simple. Saying it’s just statistics is ridiculously reductive like saying your response to this comment is just chemistry. Context driven tokenization, byte level byte pair encoding, RoBERTa, fine tuning methods, direct preference optimization, dataset curation and management, and curriculum learning for targeted performance and memory are things that are being developed and refined very fast (like weekly or monthly breakthroughs sometimes) and with pretty staggering performance increases. It still is not for everyone because power tools injure, but instead of saying “just a statistics engine” say what you really mean “I don’t understand it, but I believe XXX is a bad use case for LLMs.”

To a lesser extent that any company is “programming AI”. Not in the way that you mean it; curriculum learning, guard rails, and fine running are all extremely indirect. Nobody had their hands specifically in a model’s parameter space directly.

I was just yanking the other guy’s chain.

6? You were a year behind already! ;)

Sounds about right.

Improvisational jazz: “bro it’s been 84 measures of discordant shit, just resolve the damn thing already and play the root of the chord!”

edit:

Roses are red, violets are blue, some people have autism doodliotoodat mmbat goodatgooctmapanda macamapandiddle patmaksboodliodoo dimpaoacmapaway choopamadampakampa shittlybittly gampapawakombucha shoodleeooowasasaampandaweeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeeee dittlyittlybimbopadooooo oooooo oooooOoooooooOooOOoOOooooo chamimbamamdapaweeeee

Smooth jazz is better 🖕

I have a feeling one authored this thread.

Except the last 2. The last 2 go in the milk directly and fall in your mouth a little soggy when you finish drinking it.

Hey look it’s the hegelian dialectic, except the anthesis is a little weak and has some non negligible overlap with students. But then that’s why we iterate.

Jasper looks so horny in this gif I can’t take it.

Yes. The trick is not to be a cheap ass, but also not to get hustled. So I guess you have to know something about the work being done?

Edit:

Also, I keep the extra materials. I paid for them. They are mine.

See also that YOUR temperature drops overnight. Your body temp goes down about 2 degrees overnight starting at the same time your body starts releasing melatonin before bed. This means as you start to wake up you also start to warm up, so you’re partially at fault here too!

If you want a space to talk, that’s cool. If you are publishing works, sorry, that should be under through review from every angle, regardless of gender, class, nationality, or any other form of segmentation.

You describe it and I can hear my aunt (in law) saying it that way. I believe it.

Try a different medium or visit a doctor maybe?

That sounds like 20 minutes or less of reading but the Internet is telling me that for slower readers that’s up to an hour for 8k words. If it takes you an hour, maybe don’t go for quite as long in a sitting and work up to it a little, if it’s 15 to 25 mins of reading I think I would switch to something more paper-like / talk to an eye doctor about it.

Ghost Boy by Jacob Tillberg.

The rest of his music is unfortunately flat and disappointing to me.

Leaving social media helps… But if someone is wrong on Lemmy that’s a whole new problem I can’t help with.

I would say don’t sweat it too much and if you eat downvotes sometimes that’s fine.

Be sure to ask Claude to remove all comments, tests, and dry runs, then ask if there’s a good way you can make the code more complicated without increasing runtime more than 40%.

You and the guy you’re arguing with feel the same way.

One or both of you is wrong.

Turn the tables and LARP as an AI myself!

https://youraislopbores.me/

Aerospace systems. Have taught several classes of it already.

I play too much chess. I couldn’t figure out the first 4 words in the first 4 reads.