Hacker News Reader

Comment by globular-toast | Hacker News Reader

I don't usually post something like this, but this is so fucking stupid. I'm prepared to stand by that. Let's see in a few years if I'm right.

"AI" is literally models trained to make you think it's intelligent. That's it. It's like the ultimate "algorithm" or addiction machine. It's trained to make you think it's amazing and magical and therefore you think it's amazing and magical.

View on HN

This could apply if we looked at questions in vacuum - someone had a conversation and was judging the models based on that. But some of us just use it for work and get good results daily. "Intelligent" is irrelevant; it's "useful". It doesn't matter what feelings I have about it if it saves me 2h of typing from time to time.

To me, as just another kinda old (I’m 49) swe, the biggest benefit of using an LLM tool is it saves a shit ton of typing. I know what I want and I know when it’s right, just saving me from typing it all out is worth $20 bucks a month.

Recently I needed to summarize about a thousand lengthy documents, and then translate those summaries into Mandarin.

I spent about a minute composing the prompt for this task, and then went for a cup of coffee. When I got back the task was done. I spot-checked the summaries and they were excellent.

I thought this was amazing and magical at the time. Am I wrong? Or is it simply the AI making me think this result was amazing and magical?

This is an LLM's bread and butter so I would hope it does a decent job.

You just spot checked it, so how can you be sure how accurate it is. Was it 80% accurate? 90%? 99%? And how does the domain influence the accuracy requirements?

Sure, but there's no reason there can't be a correlation between us "thinking" it's intelligent and it actually being intelligent. What other proxy should we use? I can't think of a scenario where it's actually intelligent but humans don't think it is that has a good practical ending. It's at least necessary even if it isn't sufficient.

It’s trained to (lossy) compress large amounts of data. The system prompts have leaked and it’s just instructed to be helpful, right? I don’t entirely disagree with your sentiment, though. It’s brute force.

'"AI" is literally models trained to make you think it's intelligent.'

What's the difference? I try to make people think I'm intelligent all the time.

Weird self roast but okay.

Congratulations on that one!

Now that you have unlocked this secret, you're cursed forever. They look at the machine and say: hey, look, the machine is just like me! You're left confused for the best part of 3 years and then you start realizing it was true all along...they are..very much similar to the machine. For a moment we were not surprised by how capable the machine was at reasoning. And then it dawned on us, the machine had human level intelligence and cognition from the beginning, just from a slightly different perspective.

The system prompt may vary but:

"It's trained to make you think it's amazing and magical and therefore you think it's amazing and magical."

is the dark pattern underlying the entire LLM hype cycle IMO.

Reply on news.ycombinator.com