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

Modern-Day Oracles or Bullshit Machines? How to thrive in a ChatGPT world

thebullshitmachines.com

Post reply on HN
Print view View original

101–110 of 652 posts

Re: Modern-Day Oracles or Bullshit Machines? How to thrive in a ChatGPT world

#102

Post by PartiallyTyped »

vladms wrote: post #53
einrealist wrote: post #49

This website is so important! Now ask yourself why AI companies don't want to be regulated or scrutinized. So many companies (users and providers) jump on the AI hype train because of FOMO. The end result might be just as destructive as this mythical "AGI". Edit: I am not saying to not use the technology. I am just on the side of caution and constant validation. The technology has to serve society. But I fear this hy…

My impression is that companies in most of the fields do not like to be regulated or scrutinized, so nothing new there. While observing some people using LLMs, I realized that for a lot of people it really makes a huge difference in time saved. For me the difference is not significant, but I am generally solving complex problems, not writing nicely formatted reports where words and not numbers are relevant, so YMMV.

Anecdata, N=1; I recently used aider — a tool that gives LLMs access to specific files and git integration. The tools are great, but the LLMs are underwhelming, and I realized that — once in the flow — I am significantly faster at producing large, correct, and on-point pieces of code, whereas when I had to review LLM code, it was frustrating, it needed multiple attempts, and it frequently fell into loops.

Re: Modern-Day Oracles or Bullshit Machines? How to thrive in a ChatGPT world

#103

Post by tmnvdb »

sgt101 wrote: post #68
silvestrov wrote: post #50

Earlier quoted context omitted.

LLMs are always bullshitting, even when they get things right, as they simply do not have any concept of truthfulness.

They don't have any concept of falsehood either, so this is very different from a human making things up with the knowledge that they may be wrong.

I think the first part of that statement requires more evidence or argumentation, especially since models have shown the ability to practice deception. (you are right that they don't _always_ know what they know)

Re: Modern-Day Oracles or Bullshit Machines? How to thrive in a ChatGPT world

#104

Post by sesm »

TeMPOraL wrote: post #95

"The LLMs have no ground truth" claim (around chapter 2) that's core to the "bullshit machines" argument is itself wrong. Of course LLMs have ground truth. What do the authors think here, that the text in training corpus is random ? Hint: it isn't. Real conversations are anything but random. There's a lot of information hidden in "statistical ordering of the words", because the distribution is not arbitrary . Statist…

> Statistical ground truth isn't any worse than explicitly given one

There are multiple kinds of truths.

'Statistical truth' is at best 'consensus truth', and that's only when LLM doesn't hallucinate.

Re: Modern-Day Oracles or Bullshit Machines? How to thrive in a ChatGPT world

#106

Post by TheOtherHobbes »

tmnvdb wrote: post #71

Hmm, it seems that the author takes very clear (and sometimes cynical) positions on some controversial questions. For example, "They don't have the capacity to think through problems logically." is an hotly debated claim, and I think with the advent of reasoning models this has at least become something one should not state in entry level material, which would hopefully reflect common understanding rather than the au…

Many claims don't stand up to scrutiny, and some look suspiciously like training to the test.

The Apple study was clear about this. LLMs and their related modal models lack the ability to abstract information from noisy text inputs.

This is really obvious if you play with any of the art generators. For example - the understanding of basic prepositions just isn't there. You can't say "Put this thing behind/over/in front of this other thing" and get the result you want with any consistency.

If you create a composition you like and ask for it in a different colour, you get a different image.

There is no abstracted concept of a "colour" in there. There's just a lot of imagery tagged with each colour name, and if you select a different colour you get a vector in a space pointing to different images.

Text has exactly the same problem, but it's less obvious because what the grammar is usually - not always - perfect and the output has been tuned to sound authoritative.

There is not enough information in text as a medium to handle more than a small subset of problems with any consistency.

Re: Modern-Day Oracles or Bullshit Machines? How to thrive in a ChatGPT world

#108

Post by light_hue_1 »

einrealist wrote: post #49

This website is so important! Now ask yourself why AI companies don't want to be regulated or scrutinized. So many companies (users and providers) jump on the AI hype train because of FOMO. The end result might be just as destructive as this mythical "AGI". Edit: I am not saying to not use the technology. I am just on the side of caution and constant validation. The technology has to serve society. But I fear this hy…

AI companies desperately want to be regulated. OpenAI is lobbying hard to be regulated.

Didn't call for it. The whole point is to keep it competition with regulation against mythical vague harms that don't exist.

Re: Modern-Day Oracles or Bullshit Machines? How to thrive in a ChatGPT world

#109

Post by tmnvdb »

dzogchen wrote: post #105

This will not age well. Yesterday my bullshit machine wrote a linker argument parser to hook a C++ library up in a Rust build config. Oh it also wrote tests for it. https://chatgpt.com/share/67a89e5f-b5b4-8011-9782-472d469cc2...

[flagged]
Post reply on HN