@bekkaboo@girlcock.club Linux itself is heavily using AI models. Therefore, every Linux distribution moving to recent kernel versions is heavily built on top of using AI models. Creating a Linux distribution not heavily using it would require a hard fork of the Linux kernel and many other projects.
It's unclear what would be accomplished by banning AI for a tiny portion of the code while continuing to use Linux, AOSP, Chromium and hundreds of other projects heavily using it. We'd still be benefiting from it.
GrapheneOS is currently defending its use of AI coding tools on Mastodon against complaints by various accounts claiming to be users.
We do not understand where you’re coming from or why you’re so incredibly angry with us. It’s not justified and does not make sense.
AI is a tool, like a hammer, and should be used as such. And like a tool should require training. The problem is people aren’t using it as a tool and rather look at the hammer and say “go build me a house”. That’s where the issue is. Does a hammer know how to build a house on its own? no. does AI excel at end to end builds? no. but people are using it for said end to end builds.
Devs, vibe coders, god knows who else still to this day fail to understand that AI doesn’t understand how to build things correctly. All the agents care about is getting to a solution as quickly as possible to retain a userbase. that’s it. will what ever it builds scale? no. will it be full of exploits? very likely. will it be able to refactor whatever it wrote? no because it will believe YOU wrote it. so upgrades/improvements to whatever are virtually impossible.
THEN you combine all that with the fact that EVERY AI’s training data, currently, is incredibly out of date. Tech moves too fast for it. So you have people building shit with libraries and what have you that are already out of date and they don’t know that.
That’s a great example actually, you just have to differentiate between a nuclear weapon and a nuclear power plant. You can split atoms to build a bomb or heat water, the concept of splitting an atom itself isn’t evil.
I think with OP’s analogy, a nuclear weapon is no longer a tool, because the actual tool (nuclear technology) was crafted into something (weapon) that does a specific thing (destroy). But I guess OP should’ve used LLM and machine learning instead of AI.
Nuclear fission has been used in nuclear reactors for good, and was exceedingly efficient at that time, and has been a significant factor in the recent human history.
Just as it’s been with the “evil” use in nuclear weapons.
I’m not aware of any breakthrough that has not been used for both good and evil.
A tool built through nefarious means and consuming resources it should never had chanelled into it.
Then, taking from your words, if the information these machines are working is by default outdated, why insist in feeding it with more and more and expect different outcomes?
AI is a tool, like a hammer, and should be used as such. And like a tool should require training. The problem is people aren’t using it as a tool and rather look at the hammer and say “go build me a house”. That’s where the issue is. Does a hammer know how to build a house on its own? no. does AI excel at end to end builds? no. but people are using it for said end to end builds.
Devs, vibe coders, god knows who else still to this day fail to understand that AI doesn’t understand how to build things correctly. All the agents care about is getting to a solution as quickly as possible to retain a userbase. that’s it. will what ever it builds scale? no. will it be full of exploits? very likely. will it be able to refactor whatever it wrote? no because it will believe YOU wrote it. so upgrades/improvements to whatever are virtually impossible.
THEN you combine all that with the fact that EVERY AI’s training data, currently, is incredibly out of date. Tech moves too fast for it. So you have people building shit with libraries and what have you that are already out of date and they don’t know that.
Nuclear weapons are a tool, like any other
That’s a great example actually, you just have to differentiate between a nuclear weapon and a nuclear power plant. You can split atoms to build a bomb or heat water, the concept of splitting an atom itself isn’t evil.
Exactly
I think with OP’s analogy, a nuclear weapon is no longer a tool, because the actual tool (nuclear technology) was crafted into something (weapon) that does a specific thing (destroy). But I guess OP should’ve used LLM and machine learning instead of AI.
Nuclear fission has been used in nuclear reactors for good, and was exceedingly efficient at that time, and has been a significant factor in the recent human history. Just as it’s been with the “evil” use in nuclear weapons.
I’m not aware of any breakthrough that has not been used for both good and evil.
A tool built through nefarious means and consuming resources it should never had chanelled into it.
Then, taking from your words, if the information these machines are working is by default outdated, why insist in feeding it with more and more and expect different outcomes?
Makes no sense to begin with.
It’s a tool like a gun, specifically purpose built to usher in a technofeudalist surveillance state. Not useful for anything else as is.
Guns are incredibly useful, just like LLMs and diffusion.