Ask HN: Why does no one seem to care that AI gives wrong answers? • Hacker News Board

Re: Ask HN: Why does no one seem to care that AI gives wrong answers?

#21

Post by hpen » Wed, Jul 10, 2024, 9:15 PM UTC

proc0 wrote: post #3

There are use cases where it doesn't matter, i.e. creative writing. Additionally, I don't think AI engineers have even figured out the path for LLMs to be hallucination free and extremely accurate. It's better to ship something that is not perfect (or even not great) now, and that way the industry gains experience and the tools slowly but surely get better.

This is a false line of thinking. Karpathy say's hallucinating is what makes LLMs special. LLMs are way more like compression with a mix of hallucination rather than anything else.

Re: Ask HN: Why does no one seem to care that AI gives wrong answers?

#22

Post by threeseed » Wed, Jul 10, 2024, 9:16 PM UTC

taylodl wrote: post #18

I haven't found a human that answers every single question correctly, either. You know whom to ask a question based off that person's domain of expertise. Well, AI's domain of expertise is everything (supposedly). What gets difficult is evaluating the response, but let's not pretend that's any easier to do when interacting with a human. Experts give wrong answers all the time. It's generally other experts who point o…

a) If you ask me about surgery I will say "I don't know". LLMs won't do that.

b) Experts may give wrong answers but it will happen once. LLMs will do it over and over again.

Re: Ask HN: Why does no one seem to care that AI gives wrong answers?

#23

Post by exabrial » Wed, Jul 10, 2024, 9:17 PM UTC

Another question, why does nobody care about the enormous and absurd energy cost to train and run models?

Re: Ask HN: Why does no one seem to care that AI gives wrong answers?

#24

Post by bitshiftfaced » Wed, Jul 10, 2024, 9:18 PM UTC

You'd consider it a bug because you can fix a bug. Doesn't seem like the tech's there yet, but it's still good enough to be useful.

Re: Ask HN: Why does no one seem to care that AI gives wrong answers?

#25

Post by nashashmi » Wed, Jul 10, 2024, 9:18 PM UTC

Think of it as a system returning complex results after a search. Think of it as a synthetic search result. Think of it as a result where you still have to evaluate the source for reliability. Think of it as a junior engineer making mistakes. Think of it as a reason why you will have job security for complex tasks because the easy tasks can be done by dumber project managers. Think of it like you are now a senior engineer with a junior engineer doing the mundane stuff.

Do the mundane stuff in school/college/boot camp. Do the cool stuff at work.

Re: Ask HN: Why does no one seem to care that AI gives wrong answers?

#26

Post by latentsea » Wed, Jul 10, 2024, 9:19 PM UTC

threeseed wrote: post #22

taylodl wrote: post #18

I haven't found a human that answers every single question correctly, either. You know whom to ask a question based off that person's domain of expertise. Well, AI's domain of expertise is everything (supposedly). What gets difficult is evaluating the response, but let's not pretend that's any easier to do when interacting with a human. Experts give wrong answers all the time. It's generally other experts who point o…

a) If you ask me about surgery I will say "I don't know". LLMs won't do that. b) Experts may give wrong answers but it will happen once. LLMs will do it over and over again.

>b) Experts may give wrong answers but it will happen once. LLMs will do it over and over again.

Well... Sometimes "experts" will give the wrong answer repeatedly.

Re: Ask HN: Why does no one seem to care that AI gives wrong answers?

#27

Post by john2x » Wed, Jul 10, 2024, 9:20 PM UTC

exabrial wrote: post #23

Another question, why does nobody care about the enormous and absurd energy cost to train and run models?

Line go up. The only three words that matter nowadays.

Re: Ask HN: Why does no one seem to care that AI gives wrong answers?

#28

Post by runjake » Wed, Jul 10, 2024, 9:21 PM UTC

Of course most everyone cares, but the value proposition is high enough that people aren't going to hold off on using it until it is perfect.

Nonetheless, as with autopilot, you don't want to substitute paying attention with it. "Trust, but verify" as Reagan said.

Re: Ask HN: Why does no one seem to care that AI gives wrong answers?

#29

Post by rfjimen » Wed, Jul 10, 2024, 9:21 PM UTC

Your point is valid if you believe LLM/Generative AI is deterministic; it is not. It is inference-based, and thus it provides different answers even given the same input at times.

The question then becomes, "How wrong can it be and still be useful?" This depends on the use case. It is much harder for applications that require high deterministic output but less important for those that do not. So yes, it does provide wrong outputs, but it depends on what the output is and the tolerance for variation. In the context of Question and Answer, where there is only one right answer, it may seem wrong, but it could also provide the right answer in three different ways. Therefore, understanding your tolerance for variation is most important, in my humble opinion.

Re: Ask HN: Why does no one seem to care that AI gives wrong answers?

#30

Post by ThunderSizzle » Wed, Jul 10, 2024, 9:21 PM UTC

latentsea wrote: post #26

threeseed wrote: post #22

Earlier quoted context omitted.

a) If you ask me about surgery I will say "I don't know". LLMs won't do that. b) Experts may give wrong answers but it will happen once. LLMs will do it over and over again.

>b) Experts may give wrong answers but it will happen once. LLMs will do it over and over again. Well... Sometimes "experts" will give the wrong answer repeatedly.

Assuming they aren't convinced they are wrong - and multiple opinions can be valid while differing. The world isnt always black and white. An AI can never be convinced it's wrong permanently, and sometimes it can't be convinced ed temporarily, depending on the model