Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

Has the hallucination problem in AI been solved?

news.ycombinator.com

Post reply on HN
Print view View original

41–50 of 63 posts

Re: Has the hallucination problem in AI been solved?

#41

Post by salawat »

spl757 wrote: post #17
salawat wrote: post #12

Earlier quoted context omitted.

The types of people using it that way are not concerned with the best interests of anyone but themselves, and arguably not even that except within a very immediate time horizon.

Perhaps you should be concerned about it. I certainly am. This is deployed on the battlefield and has been for months.

What makes you think I'm not concerned about it? I've been positively mortified at the seeming cavalier way in which all caution is thrown to the winds. I've honestly had to fundamentally reassess my understanding of the average character of humanity given the last decade.

Fact is, those with the foresight to see how this can go badly are also seemingly the types of people who don't end up in a position to prevent harms by it. Furthermore, it seems inevitable in a sense, because greed for power seems to necessitate development of automated weapons in ASAP in spite of the risks, and the fact it basically renders traditional warfare pointless.

It appears we'll have to learn the hard lessons, same as our forebearers with. I just hope we can avoid having to regress back to sticks and stones on account of fucking ourselves by overdoing our capability to destroy on account of not being willing0able to peacefully coexist.

Re: Has the hallucination problem in AI been solved?

#42

Post by sandcat_ »

Gemini is the only model that I still see regularly flat-out hallucinate, e.g. the other day it told me we were using a particular technology (Fivetran) on a project when we aren’t and I have literally no idea where it could have gotten that from. No mention of it at all, etc.

Claude will make mistakes but they’re largely “reasonable”. In some ways that’s worse as they’re more believable.

GPT I can’t comment on too much, except to say we run a chatbot using it at my work and when extracting data from the conversation (contact info and such) it will occasionally make up an email address that wasn’t entered into the chat. We have guardrails around it, so it’s not a big deal, but it does happen.

Re: Has the hallucination problem in AI been solved?

#43

Post by iDon »

I'd agree with other posters - I'm seeing less hallucination; probably the model-makers are tuning to reduce that, because people make a big deal about it.

I see it as a small deal - it reminds us to check the responses against references. LLMs are a statistical construction, and everyone accepts without complaint that statistical models have a predicted false-positive and false-negative rate. I think of "hallucination" as a false positive; false negative is no answer when the model could have made a useful response. I think there's some relation between our creativity and the LLM capability we dismissively call hallucination.

Re: Has the hallucination problem in AI been solved?

#44

Post by burnt-resistor »

spl757 wrote: post #5
AIOSAI wrote: post #3

I think its solved, with the right setup and model. I couldn't tell u the last time my agent hallucinated. For me I consider it solved. However some dude feeding massive docs into gpt chat and long conversations, It is not solved in this context.

Would you trust it to make life and death decisions? Because it is being used in that context. Drones with AI, armed, and with discretion to choose a target and kill it.

No, because that's Terminator territory (ULAWS). Since drones' inception, US military required an officer to approve lethal weapon release until at least 2013 and it may or may not still be required. The official policy is DODD 3000.09, which uses the vague word "appropriate" 22 times, but doesn't allow unsupervised lethal autonomous (ULAWS) yet. Maybe someone knows what the criteria are used around in the world's militaries these days beyond Political Declaration on Responsible Military Use of Artificial Intelligence and Autonomy (2023-2024) (which Ukraine signed) (There's a competing REAIM 2023 Call to Action that Ukraine didn't sign with a note.[0])

In other news, New Orleans 911 is using AI to triage localized incident duplication calls from unique emergencies. I'm not saying that it's good or their only practical choice, but it's happening.

The biggest dangers I see are the outsourcing of supervisory control, appeal to authority (when used to summarize content or answer a question), and hallucinated mistakes.

0. (PDF) https://docs-library.unoda.org/General_Assembly_First_Commit...

1. PDRMUAIA https://www.state.gov/bureau-of-arms-control-deterrence-and-...

2. REAIM 2023 Call to Action https://www.government.nl/documents/2023/02/16/reaim-2023-ca...

3. REAIM 2023 Endorsing Countries and Territories https://www.government.nl/documents/2023/02/16/reaim-2023-en...

Re: Has the hallucination problem in AI been solved?

#45

Post by jimmaswell »

sjbzbeiks wrote: post #36
jimmaswell wrote: post #34

Earlier quoted context omitted.

It matches the reality of my experience using a good frontier like Sol on high to ultra thinking. If it has any way to verify its work whatsoever, it eventually gets to a working and sanely-engineered solution. Blatant hallucinations making it to the final stages have become extremely rare in my use cases, nonexistent if the model has a valid feedback loop. So yes, I will insist someone is likely "holding it wrong" i…

I think this is the issue others are having, when you say: > when I can have it do something like write entire working kernel module fixes for old MacBooks on a whim They’re not saying it can’t do that, and that’s not proof it doesn’t hallucinate. In fact, having used 6-8 agents at a time for a year plus while writing AI tooling for an AI startup, I can definitely surely tell you that they’re almost inversely correla…

What do we mean by hallucinate? I'm not counting it making a mistake that it fixes on its own without intervention.

I'm no expert on the inner workings/harnesses/etc beyond a basic understanding of the architecture. Maybe I've just developed a good sense for effective prompts? I could share some recent sessions.

Re: Has the hallucination problem in AI been solved?

#47

Post by DANmode »

dwd wrote: post #37

I don't have a problem with hallucinations, rather the confabulations. We want AI to make predictions under uncertainty that could be wrong. What we don't need is getting know facts wrong. One of the best responses I got from ChatGPT was when it said "I don't know", and on questioning it - it responded that it was an unsolved problem and it couldn't objectively take a side iun the argument.

Why not mention the problem?

Re: Has the hallucination problem in AI been solved?

#49

Post by The_GreatWorks »

spl757 wrote: post #20
bruce511 wrote: post #16

There are two common ways to spot hallucinations. Talk to it about a topic you know well, or implement what it suggests and get an error or failure. For the first, we're mostly past the point of just "testing". So I don't see that too much anymore. Mostly I still see that though in bug reports generated by AI by someone else. There is usually some underlying bug being reported, but the AI explanation and "helpful sug…

AI drones are being used to autonomously target and kill targets by the Ukraine using technology they have been given. That's real. Right now. That that is happening anywhere on the planet scares me. That's why I keep asking about AI hallucinating, and it's implications when the stakes are life and death.

It’s a different type of AI. LLMs hallucinate because of their structure — computer vision and algos don’t in the same way

Re: Has the hallucination problem in AI been solved?

#50

Post by The_GreatWorks »

spl757 wrote: post #5
AIOSAI wrote: post #3

I think its solved, with the right setup and model. I couldn't tell u the last time my agent hallucinated. For me I consider it solved. However some dude feeding massive docs into gpt chat and long conversations, It is not solved in this context.

Would you trust it to make life and death decisions? Because it is being used in that context. Drones with AI, armed, and with discretion to choose a target and kill it.

Completely different type of AI
Post reply on HN