untakenusername @ untakenusername @sh.itjust.works

Posts

0

Comments

544

Joined

4 mo. ago

pointless chat room.

that doesn't sound too bad

Ai haters out there, doesn't this give a valid use case for LLM?

Hopefully it actually gets into orbit this time

evolution is stupid

we could probably drive mosquitoes to extinction if we were motivated enough

Room temp superconductors, would be amazing for maglev trains and the electric grid

Basically a material that doesn't have electrical resistance and has some funky magnetic properties but works at normal temperatures

average Texas moment

This honestly isn't that bad of an idea, but theres much more pressing issues than this

If you have an amount of water that can be divided, it touches other water. Therefore water is wet unless it is a single molecule

that's brilliant

This amuses me

low-key maybe killing tons of people isn't a nice idea

yesn't

for those who don't know what this is, https://en.wikipedia.org/wiki/Klein_bottle and

I'd prefer solar to nuclear and oil, but I'd also take nuclear over oil any day

For things that require a definite, correct answer, an LLM just isn't the best tool for it. However if the task is something with many correct answers, or no correct answer, like for instance writing computer code (if its rigorously checked against its actually not that bad) or for analyzing vast amounts of text quickly, then you could make the argument that its the right tool for the job.

I think somehow incentivizing companies to use solar power to power their data centers would be a step in the right direction

I'm all for open source

but you know I was totally with you until you said kill all the billionaires

You shouldn’t trust anything the LLM tells you though, because it’s a guessing machine

You trust tons of other uncertain probability-based systems though. Like the weather forecast, we all trust that, even though it 'guesses' the future weather with some other math

Nobody talks how it highlights the success of capitalism either.

it definitely does both

OpenAI, for example, needs to be regulated with the same intensity as a much smaller company

not too long ago they went to Congress to get them to regulate the ai industry a lot more and wanted the govt to require licences to train large models. Large companies can benefit from regulations when they aren't easy for smaller competitors to follow.

And OpenAI should have no say in how they are regulated.

For sure, otherwise regulation could be made too restrictive, lowing competition

Before any model is released to the public, I want clear evidence that the LLM will tell me if it doesn’t know something, and will never hallucinate or make something up.

I think thats technically really difficult, but maybe if the output of the model was checked against preexisting sources that could happen, like what Google uses for Gemini

Every step of any deductive process needs to be citable and traceable.

I'm pretty sure this is completely impossible