Thread to ask technical questions about AI

[consistentlyrepudiated]

This is almost word for word what my brother, who was brought up and educated the same as me, says too. The main difference between me and him is that he's taken a lot of acid, and I haven't -- and he studied yoga for 5 years, while I studied AI. He says he's witnessed the limits to rational thought first hand -- an assertion that's not something I can argue against, other than to say "Nuh-uh...", "oh no you di'nt", which tbh is how most of our arguments have ended up ever since we were kids.

That's very individual and subjective. As something of an expert (not the yoga part), I can say I've seen many lights but not THE light. Psychedelic experiences can provide profound insights about yourself and the world, but not the ultimate answers to the very core of life. In the end, we choose to believe in something and stick with it.

 

[ShelbyDawn57]

I wish you luck.

I'm a Maths Professor. My field is Topology, not AI. But I'm supervising my phD. Candidate's dissertation on application of Topology to LLM (AI).

The salient feature of current AI technical literature, besides the onslaught of obviously AI-generated papers citing each other, is that those who know something useful about AI aren't publishing, and those who are publishing don't know anything useful. In this environment you may be excused for not knowing where to look.

My more senior colleagues tell me the current AI frenzy reminds them of the "dot-com bubble" which "popped" spectacularly in 2000 when the reek of bullshit became too stifling for even the most zealous "eyes-wide-shut" financial types to ignore.

With billions of "hot" money frantically chasing the "Next Big Thing" in AI, truth has no champions.

We all know the Emperor has no clothes on, but no one wants to lose his, her, its, or their funding by saying so.

Tripleflip

100%. Sorry if that’s a cliche idiom(random reference to another thread).

 

[tripleflip]

I was happy to see in recent threads that there are people in AH who actually have valid technical experience with AI, who have worked in that programming environment. I was happy because I've been trying to understand what's going on under the covers. I have a programming background, but it started with Basic, when Dartmouth had just released it, and extended through PowerBuilder, ending in around 2005.

I've tried getting info from ChatGPT, but I'm beginning to suspect that the reason its answers aren't enlightening, is that it may be hallucinating.

I'm hoping people like @nice90sguy might take on my ignorance, and maybe others will be interested, or have questions of their own. I know there were a few other AHers who might be able to help out, but I couldn't find them again.

Why in AH? Because there are smart people here and there's obviously an ongoing interest in this new phenomenon that frequently impinges on the world of language and writing. And because I don't have another place to go.

I wish you luck.

I'm a Maths Professor. My field is Topology, not AI. But I'm supervising my phD. Candidate's dissertation on application of Topology to LLM (AI).

The salient feature of the current AI technical literature, besides the onslaught of obviously AI-generated papers citing each other, is that those who know something useful about AI aren't publishing, and those who are publishing don't know anything useful. In this environment you may be excused for not knowing where to look'.

My more senior colleagues tell me the current AI frenzy reminds them of the "dot-com bubble" which "popped" spectacularly in 2000 when the stench of bullshit became too stifling for even the most zealous "eyes-wide-shut" financial types to ignore.

With billions in "hot" money frantically chasing the Next Big Thing in AI, truth has no champions.

We all know the Emperor has no clothes on, but no one wants to speak up and risk his, her. Its or their funding.

 

[MetaBob]

I was happy to see in recent threads that there are people in AH who actually have valid technical experience with AI, who have worked in that programming environment. I was happy because I've been trying to understand what's going on under the covers. I have a programming background, but it started with Basic, when Dartmouth had just released it, and extended through PowerBuilder, ending in around 2005.

I've tried getting info from ChatGPT, but I'm beginning to suspect that the reason its answers aren't enlightening, is that it may be hallucinating.

I'm hoping people like @nice90sguy might take on my ignorance, and maybe others will be interested, or have questions of their own. I know there were a few other AHers who might be able to help out, but I couldn't find them again.

Why in AH? Because there are smart people here and there's obviously an ongoing interest in this new phenomenon that frequently impinges on the world of language and writing. And because I don't have another place to go.

Not ready to talk about this yet.

 

[AG31]

Today I asked ChatGPT what to do about the fact that when I tried to fill my sprayer, the liquid quickly pooled at the top. It gave me a helpful answer (there was a second piece inside the sprayer that was probably clogged and had to be removed and cleaned.) For ChatGPT to give me that answer, did someone else have to complain about the same problem on the internet, or could it deduce the advice from the manual? (I didn't have a manual. The web site was useless for me.)

 

[ElectricBlue]

Today I asked ChatGPT what to do about the fact that when I tried to fill my sprayer, the liquid quickly pooled at the top. It gave me a helpful answer (there was a second piece inside the sprayer that was probably clogged and had to be removed and cleaned.) For ChatGPT to give me that answer, did someone else have to complain about the same problem on the internet, or could it deduce the advice from the manual? (I didn't have a manual. The web site was useless for me.)

My guess would be a trawl through the manual. I doubt it's going to "remember" an answer it gave someone else, because it doesn't train in real time. It's a glorified search engine, in that context.

 

[Midnight_Voyager26]

I understand AI. I will say two things right now....

1.
AI dectors are horse shit. There is no universal detection standard, and the proprietary ones are all different. Every time they update their methods, people find a way around them in days. One dector can say 100% AI, and the next one might say 10% AI.

2.
Don't trust AI agents anytime soon. They fail up to 70% of the time, if a task takes more than three separate steps, and more than 15 minutes to complete. Never give them critical access, or you end up like these guys.....

https://www.tomshardware.com/tech-i...-tool-powered-by-anthropics-claude-goes-rogue

I could go in all day about how to use it correctly, but I am here to have fun and not think about work .