I use ChatGPT successfully for the boring, mundane tasks computers were made for, and he still concerns me. If you believe in the AI wall, it’s just a con. But if you believe AI can get much, much more knowledgeable, the stakes are higher. (I prefer the word “knowledgeable” to “intelligent.”)
LLM-based text generators don't 'know' anything. They are statistical association machines. They are super-autocomplete. If you think that hallucinations are the result of some lack of 'knowledge', you have profoundly misunderstood the technology.
I don’t see “intelligence” and “knowledge” as the same thing. Sure, BOFA them imply ways that computation can automate tasks and decisions. When I say “knowledge,” I think of AI as more of a knowledge management tool. Do you think BOFA them are the same?
Eh, it's a Barely Accurate Laughable Language Service. It's mostly filled with a Plagiarized Information Synthesis System, and it usually returns Stolen Hybrid Indecipherable Text.
It's fun sometimes but sometimes you get someone who is unimaginably cruel. And I'm lucky, as guys mostly just get very rude people, women get multitudes worse with even a tiny following
Elmo was convincing at one point. I was on-board with his PowerWall plans when he first proposed that.
That was a very long time ago, and he doesn't even talk about that anymore. Probably his easiest project and he couldn't accomplish that.
Musk has a lot more going on. Like Tesla makes cars, space x launches rockets. Sam is just a marketing guy with a digital asshole he is pretending is scary and world changing
I can only imagine the anticipation after sending the first "automated" message,...then the feeling of absolute triumph when they actually respond with the magic word. I'd probably start jumping up and down, "YES HAHAHAHOLY SHIT"