zbyte64

joined 2 years ago
[–] zbyte64@awful.systems 1 points 1 day ago

Antisocial people develop an antisocial AGI promising it won't be antisocial towards a select group. Surprised Pikachu when the AGI concludes they are not in the in group

[–] zbyte64@awful.systems 0 points 3 days ago* (last edited 3 days ago)

"If the west doesn't build it then China will" is a claim to a timeline. And in the context of governments, most would assume you're talking about less than 100 years.

[–] zbyte64@awful.systems 22 points 3 days ago (3 children)

Because asking AI to do humans is still uncanny. Pixar and Ghibli was already ruined by AI...

[–] zbyte64@awful.systems 6 points 4 days ago* (last edited 4 days ago) (2 children)

Psychedelics doesn't magically make a bad person into a good person, just look at Joe Rogan. Though I am all for putting the entire admin deep into a k-hole that they will never escape from

[–] zbyte64@awful.systems 3 points 4 days ago

Yeah, it's only our fault if they die a peaceful death instead of going to trial

[–] zbyte64@awful.systems 2 points 4 days ago (1 children)

Yes, code harnesses help by providing deterministic feedback like with a language server and reduce the amount of prompting requirements. I guess I should have led with that example 😅

[–] zbyte64@awful.systems 2 points 4 days ago

The person I responded to said:

Also this is class/power issue not a gender issue.

[–] zbyte64@awful.systems 2 points 4 days ago* (last edited 4 days ago) (3 children)

No, it's not the same as copying and pasting the TODO into a prompt. Embedding the TODO in code instead of the prompt reduces tokens burned and increases accuracy because it's observing the TODO in context. Sure you can write more prompting to provide that context, but it still won't be as accurate. The less context you provide via prompting and instead provide more context through automatic deterministc feedback the better the results

[–] zbyte64@awful.systems 1 points 4 days ago* (last edited 4 days ago)

The Romans enslaved their own (and other modern day "whites") so racism isn't required to make the analogy.

[–] zbyte64@awful.systems 2 points 5 days ago (5 children)

Examples to consider:

A code base with TODOs embedded will make fewer mistakes and spend less tokens than if you attempt to direct the LLM only with prompting.

A file system gives an LLM more context than a flat file (or large prompt) with the same contents because a file system has a tree like structure and makes it less likely the LLM will ingest context it doesn't need and confuse it

Lastly consider the efficacy of providing it tools vs using agent skills which is another form of prompting. Giving an LLM a deterministic feedback loop beats tweaking your prompts every time

[–] zbyte64@awful.systems -3 points 5 days ago (1 children)

That's like saying slavery wasn't about race but economics because you had black slave owners

[–] zbyte64@awful.systems 1 points 5 days ago (2 children)

And there were Roman slaves that enjoyed a higher standard of living relative to the global norm. We have literal white supremacists argue that chattel slavery was good actually because the standard of living blacks experience now is higher than what they would have if we left them alone.

view more: next ›