webghost0101

joined 2 years ago
[–] webghost0101@sopuli.xyz 0 points 1 week ago

More modern but still not quite reality for many.

If you categorize it as anything then Neurodivergent is the most appropriate label.

Thats not a medical label anymore though but in a medical context you should be more specific anyway.

[–] webghost0101@sopuli.xyz 2 points 1 week ago* (last edited 1 week ago) (1 children)

Lanterns of course.

Would take an image generation model at least 3 steps which it doesn’t have right now.

A review step to see if the output matches the prompt.

A identification step to detect elements that don't match

A redo step to mix that area in the background image (remove) or regenerates an improvement.

Right now you cant iterate on images. Every minor tweak is a completely new image. At least not with dalle because you cant control the seed.

[–] webghost0101@sopuli.xyz 7 points 1 week ago (1 children)

It’s owned by amazon. Time to look for alternatives.

[–] webghost0101@sopuli.xyz 4 points 1 week ago* (last edited 1 week ago) (3 children)

Image generation uses DallE and is not Baked into the model.

All it does is give it a prompt. Generates and shows you the results for that prompt.

You can click the pictures to see that prompt and you will see that it verbosely requested it overflowing but dalle does not always interpret that prompt the same, actually i found llms rather suck at prompting image generation models because they behave very strongly on certain words.

A similar experiment was done with “street with no lanterns” always resulting in a lantern.

[–] webghost0101@sopuli.xyz 8 points 1 week ago

My puzzle elo is 1600, my real elo 1300

Maybe i should stop playing blitz and play a long match where i have time to think about my Moves?

Nah to slow.

[–] webghost0101@sopuli.xyz 15 points 1 week ago* (last edited 1 week ago) (2 children)

This is like saying Linux can’t be viable because Microsoft spends an estimated 3-5 Bilion on windows annually.

To me, the value of the internet is in connecting people with freedom of expression. I don't know what you believe any of us is gaining by letting google walk over us.

If we just stop playing along it looks like the enshitificated corpo owned sites are doing us one favor by isolating themselves from us, motivating innovation. But we wont be isolated from the net. We will be right here.

And if we look at historical references of a split internet. Facebook has an .onion version to be used with tor. So actually, they will still be chasing us. Because they need us way more then we need them.

[–] webghost0101@sopuli.xyz 25 points 1 week ago* (last edited 1 week ago) (10 children)

I don't think anyone here who talked about switching away from firefox even considered the chromium browsers as an alternative.

What the other comment also said its a baseless assumption that the forks wont live on without firefox. Thats just not how open source works.

New web technologies

Do you mean the ones that we are creating with the fediverse that are all equally open source or just those that google wants us to adopt?

[–] webghost0101@sopuli.xyz 2 points 1 week ago* (last edited 1 week ago)

Large language models from two years ago can write you an executable script that displays a window to provide an url.

My “subscriptions” are downloaded automatically on a schedule and imported in jellyfin. I never need to interact with it myself.

[–] webghost0101@sopuli.xyz 1 points 1 week ago* (last edited 1 week ago)

Neurosama is a fun example but we dont really know the sauce vedal coocked up.

When i say proven i mean 32 page research paper specifically looking into it.

https://arxiv.org/abs/2407.12831

They found that even a model trained specifically on honesty will lie if it has an incentive.

The reasoning models will output that they used the forbidden tool in their reasoning window before lying in the final output.

[–] webghost0101@sopuli.xyz 4 points 1 week ago

“They have to wear it somehow”

The elephant may disagree with that.

[–] webghost0101@sopuli.xyz 3 points 1 week ago* (last edited 1 week ago) (2 children)

Its actually been proven that AI can and will lie. When given a ability to cheat a task and the instructions not to use it. It will use the tool and fully deny doing so.

Edit:

Not sure why the downvotes because when i say proven i mean the research has been done and the results have been known for while

https://arxiv.org/abs/2407.12831

 

The only rule i can tell you is this one.

This session has ended with a King of Diamonds. Continue playing on !AnarchyChess@sopuli.xyz

 

Uno is based on an open source card game.

view more: ‹ prev next ›