[mathieuh]

It's the same in a few subreddits I frequent as well. Loads of accounts, all a couple of days old, all making the same or similar comments in a tone of voice that is completely different from the tone people usually use in those subreddits. If these incredibly-obvious no-content non-comments are the best that LLMs can do I think our jobs are safe.

[haswell]

> If these incredibly-obvious no-content non-comments are the best that LLMs can do I think our jobs are safe

I think we have to look at these as an early iteration. These bots are only going to get better and better.

[apwell23]

ah the classic " This is just the beginning" argument driving the hype. Have you considered that this might all that we will ever get?

chatgpt-4o after much hype, billions of dollars and years later didn't improve much in terms of core intelligence of its responses.

I guess thats not enough for "AGI is coming because we have chatbot" crowd.

[Workaccount2]

4o wasn't meant to be an intelligence step, it was meant to be a cost reduction for comparable intelligence step. It's cheaper and faster than 4 while being sort of as smart.

GPT5 will be the test of whether or not things are still going upwards.

[haswell]

We don’t need anything nearing AGI for these bots to become nearly indistinguishable from humans and to have increasing impact on communities.

I think it’s utterly fascinating how quickly people have lost sight of current reality, either by believing AGI must be around the corner, or by being sure it absolutely isn’t. Assuming 4o is the most advanced model we ever have, the point still stands. Many of the poorly built bots would blend in much more effectively with iteration on their prompting and design and by more effectively using existing services.

I personally think we’re decades away from AGI (if we ever get there), but that has nothing to do with this.

[lupusreal]

> "We don’t need anything nearing AGI for these bots to become nearly indistinguishable from humans"

More and more I think AGI is shifting to mean "machine soul" or some nebulous wishy-washy unfalsifiable religious/philosophy shit like that. What meaning could AGI have which is empirically verifiable, other than being indistinguishable from people? Humans I presume are supposed to have "general intelligence", so if there isn't a detectable difference between the machine and humans, how is that not artificial general intelligence?

[haswell]

Language is just a variable resolution abstraction on top of non-linguistic thinking and information processing.

I think it’s a mistake to equate the language ability of these models with general intelligence, even if the language produced is excellent.

I think some people see AGI that way (some nebulous unfalsifiable thing), but that’s not what I’m arguing. I think there’s a strong case to be made that at a minimum, AGI’s core “knowledge” will need to be made up of far more than just a lossy textual representation of the world.

[apwell23]

> there isn't a detectable difference between the machine and humans

Humans come in all levels of intelligence. Yes there might be a human somewhere that believes that they should eat rocks or add glue to pizza. But as a generality we can assume that adult "humans" know not to eat rocks.

[lupusreal]

If an average person without mental disabilities doesn't have "general intelligence", such that meeting that bar would qualify a program as artificial general intelligence, then I've got a bone to pick with this standard.

[apwell23]

> bots to become nearly indistinguishable from humans

curious, why do you this is going to happen. Whats the thought process here to come to that conclusion.

[haswell]

I don’t just think this will happen; I think it’s already happening. Harder to pull off here, but already extremely prevalent on Reddit and other popular social spaces. And part of the reason it works is that a significant subset of human commenters write rather poorly and are not hard to beat, to say nothing of the evolving tools.

But don’t take my word for it. Go spend some time experimenting with the state of the art and run some of your own tests with it. Assuming you yourself are not one of these bots, what you find might surprise you.

[apwell23]

> I think it’s already happening

This was happening way before LLMs though.

[tester457]

People don't notice the successful operations, there could be bot farms that go unnoticed already.

[j-bos]

Chatgpt's context configurations already allow susers to get reposnses mimicing specific speech styles. So no, this is not all we will get because better already exists, these fools just didn't use it.

[apwell23]

can you create an example in this post and see how many upvotes it gets.

[j-bos]

Pretty sure that'd violate the rules. But you can search on twitter for some examples contexts to use. Or experiment yourself.

[lukas099]

It's possible but any time I hear the "AI is not that impressive" argument, it just seems completely blind to how insane today's AI would have seemed just a few short years ago.

[apwell23]

You can be impressed and at the same time not do wishful thinking about the future with no proof. They are not mutually exclusive.

ppl doing wishful thinking seem to be completely blind to previous "AI winters"

[lukas099]

Like I said I acknowledge the possibility that this won’t go further. I just think it’s crazy how fast people’s expectations have gone up

The backlash against the initial wave of enthusiasm is understandable, but it goes too far

[pndy]

I've seen bots accounts influx mid pandemic and shortly after Russian invasion on Ukraine praising particular countries, politicians - some of these were registered within hours apart. Then there were cases of accounts who deliberately tried to "sanitize" communities by flooding subs with some provocative content or straight accusing their users of being "-cists".

reddit despite of the advancing enshittification is still a big platform and a daily routine for many so it's not surprising that its being used by agents of various countries or communities to deploy their propaganda.

[moffkalast]

I think it's actually worse, it's so much worse because... survivorship bias. Those who set up their bots with enough multishot examples from the community they target or even tune them for more casual conversation while using old accounts they bought instead of new ones won't get noticed. Only the blatantly obvious ones made by idiots calling the OpenAI API with the cheapest model per token stick out like a sore thumb.