These ones have been worrying about AI Notkilleveryoneism since well before LLMs were a thing anyone was making money with, and also holy hell entities can be so dumb sometimes, the spirit of the people who asked Babbage whether, if the wrong numbers were put in, the right answer would come out. Superhuman intelligence is a threat, but ka wow the systems that exist are supercharging the natural stupidity.
Agreed. It’s a bit of both. There’s been people talking about the “alignment problem” in the right fora for like 20 years. And, if we ever have consciousness / sentience that is created by us, it will be something that it would be nice to have an answer to.
But, the most relevant AI issues today are not that. It’s about liability-shifting and other externalization by the powerful, misunderstanding and misuse by the well-meaning, and various cognitive hazards. These are not marketing, these are what the companies don’t want you to think about. There also lock-in and labor issues that the model providers don’t want to talk about, though these do have somewhat-known mitigations. You get plenty of people that are primarily concerned with these issues being dismissed a “AI doomers” which in, I think, wrong.
Yeah absolutely, these ones focused in trying to study alignment because it seemed like the most compelling problem for various reasons, but these ones had an expectation that there would be beings out fighting the immediate issues, just as there are beings with climate change as their cause area, or biosafety. People who decide their cause area is the only one that matters are being silly at best.
Incidentally, consciousness / sentience are kind of side notes, properties intelligent systems are capable, but not a requirement to be smart enough to tile the light cone with processors grinding out ever higher prime numbers.
Superhuman Intelligence might be a threat if any of the companies developing AI right now had any way that their word guessing programs could possibly simulate consciousness. The best argument I’ve heard so far that they could is “no one understands how LLMs work, and no one understands how the human brain works, so you can’t prove they’re not the same.”
Talking about superhuman intelligence now, and fretting and worrying about what we’re going to do in the future where AI is smart enough to be dangerous, ignores all of the ways AI is dangerous now, like giving wronf answers, making it almost impossible to find useful information on the internet because sites that have it are buried under AI-generated SEO slop that’s wrong half the time, consuming way more electricity and water than municipalities can afford to supply them with, the datacenter heat island effect, carbon emissions and other forms of deadly pollution from brand new power plants for AI datacenters, to say nothing of all the jobs that are being destroyed by bosses that think their customers won’t notice the change in product quality.
Por que no los dos?
These ones have been worrying about AI Notkilleveryoneism since well before LLMs were a thing anyone was making money with, and also holy hell entities can be so dumb sometimes, the spirit of the people who asked Babbage whether, if the wrong numbers were put in, the right answer would come out. Superhuman intelligence is a threat, but ka wow the systems that exist are supercharging the natural stupidity.
Agreed. It’s a bit of both. There’s been people talking about the “alignment problem” in the right fora for like 20 years. And, if we ever have consciousness / sentience that is created by us, it will be something that it would be nice to have an answer to.
But, the most relevant AI issues today are not that. It’s about liability-shifting and other externalization by the powerful, misunderstanding and misuse by the well-meaning, and various cognitive hazards. These are not marketing, these are what the companies don’t want you to think about. There also lock-in and labor issues that the model providers don’t want to talk about, though these do have somewhat-known mitigations. You get plenty of people that are primarily concerned with these issues being dismissed a “AI doomers” which in, I think, wrong.
Yeah absolutely, these ones focused in trying to study alignment because it seemed like the most compelling problem for various reasons, but these ones had an expectation that there would be beings out fighting the immediate issues, just as there are beings with climate change as their cause area, or biosafety. People who decide their cause area is the only one that matters are being silly at best.
Incidentally, consciousness / sentience are kind of side notes, properties intelligent systems are capable, but not a requirement to be smart enough to tile the light cone with processors grinding out ever higher prime numbers.
Superhuman Intelligence might be a threat if any of the companies developing AI right now had any way that their word guessing programs could possibly simulate consciousness. The best argument I’ve heard so far that they could is “no one understands how LLMs work, and no one understands how the human brain works, so you can’t prove they’re not the same.”
Talking about superhuman intelligence now, and fretting and worrying about what we’re going to do in the future where AI is smart enough to be dangerous, ignores all of the ways AI is dangerous now, like giving wronf answers, making it almost impossible to find useful information on the internet because sites that have it are buried under AI-generated SEO slop that’s wrong half the time, consuming way more electricity and water than municipalities can afford to supply them with, the datacenter heat island effect, carbon emissions and other forms of deadly pollution from brand new power plants for AI datacenters, to say nothing of all the jobs that are being destroyed by bosses that think their customers won’t notice the change in product quality.