3 ms·
> I'm unsure where this expectation of 100% absolute correctness comes from. It's a computer. That's why. Change the concept slightly: would you use a calculat
by dylan604 3y ago
> I'm unsure where this expectation of 100% absolute correctness comes from.
It's a computer. That's why. Change the concept slightly: would you use a calculator if you had to wonder if the answer was correct or maybe it just made it up? Most people feel the same way about any computer based anything. I personally feel these inaccuracies/hallucinations/whatevs are only allowing them to be one rung up from practical jokes. Like I honestly feel the devs are fucking with us.
- bostonpete 3y agoSpeech to text is often wrong too. So is autocorrect. And object detection. Computers don't have to be 100% correct in order to be useful, as long as we don't put too much faith in them.
- dylan604 3y agoYour caveat is not the norm though, as everyone is putting a lot of faith in them. So, that's part of the problem. I've talked with people that aren't developers, but they are otherwise smart individuals that have absolutely not considered that the info is not correct. The readers here are a bit too close to the subject, and sometimes I think it is easy to forget that the vast majority of the population do not truly understand what is happening.
- dr_dshiv 3y agoNah, I don’t think anything has the potential to build critical thinking like LLMs en masse. I only worry that they will get better. It’s when they are 99.9% correct we should worry.
- clumpthump 3y agoCall me old fashioned, but I would absolutely like to see autocorrect turned off in many contexts. I much prefer to read messages with 30% more transparent errors rather than any increase in opaque errors. I can tell what someone meant if I see "elephent in the room", but not "element in the room" (not an actual example, autocorrect would likely get that one right).
- llbeansandrice 3y agoPeople put too much faith in conspiracy theories they find on YT, TikTok, FB, Twitter, etc. What you're claiming is already not the norm. People already put too much faith into all kinds of things.
- kamikaz1k 3y agoOkay, but search is done on a computer, and like the person you’re replying to said, we accept close enough. I don’t necessarily disagree with your interpretation, but there’s a revealed preference thing going on. The number of non-tech ppl I’ve heard directly reference ChatGPT now is absolutely shocking.
- bitvoid 3y ago> The number of non-tech ppl I've heard directly reference ChatGPT now is absolutely shocking. The problem is that a lot of those people will take ChatGPT output at face value. They are wholly unaware that of its inaccuracies or that it hallucinates. I've seen it too many times in the relatively short amount of time that ChatGPT has been around.
- bongodongobob 3y agoSo what? People do this with Facebook news too. That's a people problem, not an LLM problem.
- janalsncm 3y agoIf we rewind a little bit to the mid to late 2010s, filter bubbles, recommendation systems and unreliable news being spread on social media was a big problem. It was a simpler time, but we never really solved the problem. Point is, I don’t see the existence of other problems as an excuse for LLM hallucination, and writing it off as a “people problem” really undersells how hard it is to solve people problems.
- dylan604 3y agoPeople on social media are absolutely 100% posting things deliberately to fuck with people. They are actively seeking to confuse people, cause chaos, divisiveness, and other ill intended purposes. Unless you're saying that the LLM developers are actively doing the same thing, I don't think comparing what people find on the socials vs getting back as a response from a chatBot is a logical comparison at all
- jen20 3y ago"Computer says no" is not a meme for no reason.
- altruios 3y agowhy should all computing be deterministic? let me show you this "genius"/"wrong-thinking" person as to say about AL(artificial life) and deterministic computing. https://www.cs.unm.edu/~ackley/ https://www.cs.unm.edu/~ackley/ https://www.youtube.com/user/DaveAckley https://www.youtube.com/user/DaveAckley To sum up a bunch of their content: You can make intractable problems solvable/crunchable if you allow just a little error into the result (which is reduced the longer the calculation calculates). And this is acceptable for a number of use cases where initial accuracy is less important that instant feedback. It is radically different from a Von Neumann model of a computer - where there is a deterministic 'totalitarian finger pointer' pointing to some registry (and only one registry at a time) is an inherently limited factor. In this model - each computational resource (a unit of ram, and a processing unit) fights for and coordinates reality with it's neighbors without any central coordination. Really interesting stuff. still in its infancy...