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

I want to wash my car. The car wash is 50 meters away. Should I walk or drive?

mastodon.world

Post reply on HN
Print view View original

941–950 of 998 posts

Re: I want to wash my car. The car wash is 50 meters away. Should I walk or drive?

#941

Post by NicuCalcea »

ndriscoll wrote: post #926
NicuCalcea wrote: post #582

Earlier quoted context omitted.

> What about when you use it? I use AI for coding, but not for anything involving writing text, it's just horrendous at it. It just spews verbose slop, devoid of meaning, original thought or nuanced critique. > That doesn't mean the model isn't capable; it means an incapable person won't know what they'd want to ask of it. So it's user error again then, eh? PhD experts are able to help even "incapable" students, that…

Weird, my experience is that they are full of nuance. e.g. here is a snippet of my discussion with Gemini: > Would you like to see why Q is "flexible" (Flat) while Z/2 is "rigid" (Not Flat) using this "crushing" vs. "preserving" logic? It explains why localized rings are almost always better to work with. > Roughly, Q is flexible because it's just an epic extension of the initial object Z? > That is a very "categoric…

> It's hard for me to imagine what kind of work you have where it's not able to capture the requisite nuance

I teach journalism. My students have to write papers about journalism, as well as do actual journalism. AI is very poor at the former, and outright incapable of doing the latter. I challenge you to find a single piece of original journalism written by AI that doesn't suck.

> Obviously if someone just says "write my paper" or whatever and gives that to you, that won't work well

But it would work extremely well if they told a PhD to do it.

Re: I want to wash my car. The car wash is 50 meters away. Should I walk or drive?

#942

Post by ben_w »

scott_w wrote: post #525
DharmaPolice wrote: post #453

Earlier quoted context omitted.

Rule of thumb for everyone (i.e. both). If I ask you a question, start by assuming I want the answer to the question as stated unless there is a good reason for you to think it's not meant literally. If you have a lot more context (e.g. you know I frequently ask you trick or rhetorical questions or this is a chit-chat scenario) then maybe you can do something differently. I think being curious about the motivations b…

> Rule of thumb for everyone (i.e. both). That's never been how humans work. Going back to the specific example: the question is so nonsensical on its face that the only logical conclusion is that the asker is taking the piss out of you. > Certainly when dealing with technical problem solving I often find myself asking extremely simple questions and it often wastes time when people don't answer directly Context and t…

> That's never been how humans work. Going back to the specific example: the question is so nonsensical on its face that the only logical conclusion is that the asker is taking the piss out of you.

Or a typo, or changing one's mind part way through.

If someone asked me, I may well not be paying enough attention and say "walk"; but I may also say "Wa… hang on, did you say walk or drive your car to a car wash?"

Re: I want to wash my car. The car wash is 50 meters away. Should I walk or drive?

#944

Post by tantricked »

All of the recent LLM advancements have just been training the model to self-talk to force it to see clearly.

I hate to be a “proompter.” but I used this prompt and got the right answer without thinking:

  Before answering, do the following:

  Clearly restate the user’s actual objective.
  Identify what must physically or logically change for the objective to be achieved.
  Check for hidden assumptions or trick framing.
  Ask: “Does my answer actually accomplish the stated goal?”
  If multiple interpretations exist, briefly list them and choose the most logically consistent one.
  Do not optimize for surface efficiency if it conflicts with the core objective.
  Use strict common sense before answering.

Re: I want to wash my car. The car wash is 50 meters away. Should I walk or drive?

#946

Post by entech »

kshacker wrote: post #899

We built Windows and DOS and got Viruses. We built phone lines and got Spam and "Do Not Call" registries. We built the Internet and got Ads, Scams, and Spoofing. We built Google Search and got SEO gaming. We built Facebook and it was Hijacked to influence elections. We built AirTags to track our keys, and people used them for Stalking. We built High-Frequency Trading and got Flash Crashes. We built Encryption to prot…

Well said. I feel that every technology has a benefit to harm ratio and we as society put guardrails around use of the technology to reduce and mitigate the harm outcomes.

Unfortunately a lot of the guardrails are developed in response to the bad outcomes occurring (or rather the guardrails refined).

I feel that we are much better at dealing with obvious physical harms like planes being hijacked and viruses causing damages to your data than with higher level impacts that are not clearly felt and where you cannot hold someone accountable.

The fact that there are not even basic regulations around AI is truly mind boggling. Like how we just acknowledge a phenomenon of AI psychosis and suicidal ideation. We ban drugs and install barriers on bridges, but somehow if AI causes it - it’s seemingly ok. Let’s just hope Sam and Dario care enough to fix it.

Re: I want to wash my car. The car wash is 50 meters away. Should I walk or drive?

#947

Post by DharmaPolice »

scott_w wrote: post #525
DharmaPolice wrote: post #453

Earlier quoted context omitted.

Rule of thumb for everyone (i.e. both). If I ask you a question, start by assuming I want the answer to the question as stated unless there is a good reason for you to think it's not meant literally. If you have a lot more context (e.g. you know I frequently ask you trick or rhetorical questions or this is a chit-chat scenario) then maybe you can do something differently. I think being curious about the motivations b…

> Rule of thumb for everyone (i.e. both). That's never been how humans work. Going back to the specific example: the question is so nonsensical on its face that the only logical conclusion is that the asker is taking the piss out of you. > Certainly when dealing with technical problem solving I often find myself asking extremely simple questions and it often wastes time when people don't answer directly Context and t…

>Going back to the specific example: the question is so nonsensical on its face that the only logical conclusion is that the asker is taking the piss out of you.

I would dispute that that matters in 99.9% of scenarios.

>The person you're asking doesn't understand the link between your question and the help you're trying to offer.

Sure I, get that and I do always explain why I need to know something but it does add delays to the process (either before or after I ask). When I'm on the receiving end of a support call I answer the questions I'm asked (and provide supplementary information if I think they might need it).

Re: I want to wash my car. The car wash is 50 meters away. Should I walk or drive?

#948

Post by DharmaPolice »

vintermann wrote: post #464
DharmaPolice wrote: post #453

Earlier quoted context omitted.

Rule of thumb for everyone (i.e. both). If I ask you a question, start by assuming I want the answer to the question as stated unless there is a good reason for you to think it's not meant literally. If you have a lot more context (e.g. you know I frequently ask you trick or rhetorical questions or this is a chit-chat scenario) then maybe you can do something differently. I think being curious about the motivations b…

Sure, in a context in which you're solving a technical problem for me, it's fair that I shouldn't worry too much about why you're asking - unless, for instance, I'm trying to learn to solve the question myself next time. Which sounds like a very common, very understandable reason to think about motivations. So even in that situation, it isn't simple. This probably sucks for people who aren't good at theory of mind re…

I don't mind people asking why I asked something, I'd just prefer they answer the question as well. In the original scenario, the chatbot could answer the question as written AND enquire if that's what they really meant. It's the StackOverflow syndrome where people answer a different question to the one posed. If someone asks "How can I do this on Windows?" - telling me that Windows sucks and here's how to do it on Linux is only slightly useful. Answer the question and feel free to mention how much easier it is in Linux by all means.

I personally love explaining to people who might want to solve the issue next time so I'm happy to bore them to tears if they want. But don't let us delay solving the problem this time.

Re: I want to wash my car. The car wash is 50 meters away. Should I walk or drive?

#949

Post by bean469 »

vbezhenar wrote: post #752

It makes no sense to walk. So the whole question makes no sense as there's no real choice. It seems that LLM assumes "good faith" from the user side and tries to model the situation where that question actually makes sense, producing answer from that situation. I think that's a valid problem with LLMs. They should recognize nonsense questions and answer "wut?".

> So the whole question makes no sense as there's no real choice.

Of course there is a choice, it's even provided to the LLM directly

Re: I want to wash my car. The car wash is 50 meters away. Should I walk or drive?

#950

Post by Closi »

neya wrote: post #925
cyberrock wrote: post #175

Earlier quoted context omitted.

Does this mean we're back in favor of using weird riddles to decide programming skills now? Do we owe Google an apology for the inverse binary tree incident?

I don't know what inverse binary incident you're referring to, but the fundamental premise here is LLMs can't really think logically like humans do and they are far away from replacing humans in software, let alone senior software engineers

Because humans always say 'bread' if you ask them what you put in a toaster.

And humans will always deduce that you should switch doors if you are in a hypothetical gameshow and they show you a horse behind one of the doors.

(All I mean is - an example of an LLM answering illogically is not proof that LLMs can't really think logically, as you can equally find examples of humans answering illogically and also find examples of novel questions that LLMs get right)

Post reply on HN