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

Ask HN: Is prompt engineering just snake oil?

news.ycombinator.com

Post reply on HN
Print view View original

191–200 of 206 posts

Re: Ask HN: Is prompt engineering just snake oil?

#191

Post by mdp2021 »

throwaway675309 wrote: post #157
ZephyrBlu wrote: post #61

Earlier quoted context omitted.

Adding a working control system doesn't take away the fact it requires skill though. It seems like people think that one day they will be able to give a vague prompt and generate exactly what they were looking for, which I don't believe will ever happen. It feels like there will always be some skill in navigating the model.

If by "some skill" you mean the basic ability to type vaguely coherent English, than I agree. If anything the recent generative art advancements coming out (Firefly, MJ v5) would seem to refute your point. There was a time when diffusion models require some level of knowledge and skill to use. To get high quality output, it was in your best interest to learn about the different types of samplers, upscalers, hyper net…

> pre-processing LLM models that transform more accessible amateur prompts

Dangerous! It looks like instead of driving directly, you are tapping on the shoulders of the one at the steering wheel.

Which is useful for children, and frustrating for drivers.

Re: Ask HN: Is prompt engineering just snake oil?

#192

Post by Veen »

throw22423 wrote: post #139
Veen wrote: post #39

Earlier quoted context omitted.

You should rethink that whenever you drive under a bridge or get on a boat. The reason they collapse or sink so infrequently is largely because engineers are accredited and held responsible.

Whoever built the old bridges in my city is lost to history. There was probably no regulation then, and yet, they haven't collapsed.

That's survivorship bias. The ones that didn't collapse are still there. You don't know about the ones that did collapse because they aren't there to see.

Re: Ask HN: Is prompt engineering just snake oil?

#193

Post by hackernoteng »

Its hacking not engineering. But what helps is understanding a bit how the attention mech works and how your prompt effects the output. This "prompt engineering" is a vastly different programming paradigm. The notion of time is different, as output is being formed (one token at a time), the attention mechanism can focus more or less on different parts of the original prompt (along with the previous output tokens). This means that your "code", (your original prompt), is essentially getting "revisited" at different spots and different times while the output is getting generated and you are essentially "steering" the output in different directions based on the instructions in the prompt. This is different than how traditional code works which is always sequential and top-down while prompt writing, its better to think and speak like a human with human thoughts than to try to narrow things down to logic and code-like syntax. The LLM isnt going to process your prompt in a top-down one-time manner - its going to keep revisiting it (reweighing different parts of it, emphasising different parts of it) based on the output tokens chosen, and those weights will help determine the next token chosen, etc. Its like a whole new world of hacking, and it can be super frustrating but also amazingly effective.

Re: Ask HN: Is prompt engineering just snake oil?

#194

Post by toss1 »

mdp2021 wrote: post #60
ZephyrBlu wrote: post #55

Earlier quoted context omitted.

100%. People who have not tried to interact with these models don't understand there is actually quite a lot of skill in navigating the latent space.

It really should be just an intermediate phase. Proper, solid control system must be created. First you have Stable Diffusion, then you create ControlNet - and this as just a beginning. This is also one of the reasons why the problem of transparency ("but why and how does it work?") is so important in Machine Learning: allowing control.

This sounds like a joke I heard decades ago when I first worked with LISP programmers - when I commented about some unexpected result, he said "ah, you're looking for the DWIM command..." "What?"... "The Do What I Mean command, they're still working on it..."

Decades later, and it looks like everyone is still working on the DWIM command...

Re: Ask HN: Is prompt engineering just snake oil?

#195

Post by ImPostingOnHN »

wg0 wrote: post #31
atahanacar wrote: post #18

Is social engineering just snake oil?

Good question. Is political engineering also engineering? My question is - for social or political engineering, you can gain definite knowledge (spying, observing, researching, information extraction) and then devise something. How to do the prompt engineering exactly based on what knowledge? How to formulize and document that knowledge so that it isn't just about intuition and gut feeling but rather a learnable and…

> Is political engineering also engineering?

Good question, even if I've never heard of "political engineering"

Is social engineering snake oil, too?

Re: Ask HN: Is prompt engineering just snake oil?

#196

Post by lmeyerov »

than3 wrote: post #50
lmeyerov wrote: post #37

Earlier quoted context omitted.

Prompt testing, especially when for q/a pairs where there are multiple right answers, has been bugging me a lot The article is reasonable, but also shows a big gap in tooling, as the techniques there feel closer to linting & typing then testing once you do more interesting prompts. They don't check the interesting parts..

> The article seems reasonable but ... closer to linting then testing... they don't check the interesting parts can you elaborate a bit more on what those interesting parts are? It could just be a limitation of computation.

We are helping our users with qa tasks involving code generation, where the answers may be either JSON, executable code, or markdown discussions involving the same. We are tuning for a bunch of tools following that pattern so our users don't have to.

It's easy to make a labeled training set for grading our homework (catching regressions, ...) in the case of classifiers, and that's basically what the blog post showed.

What about for the above qa tasks? We can ask GPT4 whether a generated A was a good answer for a Q, but that's asking it to grade itself. Likewise, in the code case, we can write unit tests for the answers. (Trick: we use the former to more quickly do the latter.) But I feel like there has to be better ways

Another: OpenAI always updates models based on use, so we have to be sure our tests are real holdout sets that never get back to them...

Re: Ask HN: Is prompt engineering just snake oil?

#197

Post by marcellus23 »

vimax wrote: post #166
marcellus23 wrote: post #163

Earlier quoted context omitted.

Maybe you could explain how it's related? I feel like I did a good job of explaining how ethics makes no sense in this discussion in my above comment.

My original comment was that this list of engineering standards is not followed by most software engineering. You seem to be saying there are some that it can be applied ti, and I’m say those are in the minority.

Right — standards, not ethics. You still haven't explained how ethics matters here. Like I said, plenty of engineers that follow standards rigorously do it in the service of unethical goals.

Re: Ask HN: Is prompt engineering just snake oil?

#198

Post by wg0 »

ImPostingOnHN wrote: post #195
wg0 wrote: post #31

Earlier quoted context omitted.

Good question. Is political engineering also engineering? My question is - for social or political engineering, you can gain definite knowledge (spying, observing, researching, information extraction) and then devise something. How to do the prompt engineering exactly based on what knowledge? How to formulize and document that knowledge so that it isn't just about intuition and gut feeling but rather a learnable and…

> Is political engineering also engineering? Good question, even if I've never heard of "political engineering" Is social engineering snake oil, too?

Political engineering is pretty common in societies where deep state or establishment has a firm control and wants to engineer results their way.

Most Western democracies are not that.

Re: Ask HN: Is prompt engineering just snake oil?

#199

Post by dragonwriter »

mdp2021 wrote: post #79
ZephyrBlu wrote: post #61

Earlier quoted context omitted.

Adding a working control system doesn't take away the fact it requires skill though. It seems like people think that one day they will be able to give a vague prompt and generate exactly what they were looking for, which I don't believe will ever happen. It feels like there will always be some skill in navigating the model.

> there will always be some skill There is a world of difference between the needs in governance in just having to write "Those decorations where I marked the area, make them a bit more Hans Holbein as opposed to Klimt", compared to knowing patching tricks so that if you asked the engine to draw a 'T' you were not being served a 'B'. This latter example - ask 'B', get 'T' - I have seen from Midjourney only a few hour…

> That this is not intrinsic, but just a current stage, is only part of the story: the user will want to have a tool that allows the production of what the artist had exactly in mind

Communicating to another entity (even a human) “exactly what you had in mind” (or, more to the point, so as to get them to behave so as to produce “exactly what you had in mind”) is a skill which involves understanding of the audience, understanding of the mode of communication, and understanding of the field about which you are communicating. “Prompt engineering” is just a fancy name for the particular application of that set of skills in the domain of getting desired results out of an LLM. Whether or not the terminology ends up being durable, the skill isn’t going away, as long as we are using language to communicate with models.

(Once we get the ability to just plug our brain into an AI, will their be a skill of focussing properly so that the mind-reading tech reads the right thoughts? Probably, and that will be the new “prompt engineering”.)

Re: Ask HN: Is prompt engineering just snake oil?

#200

Post by ImPostingOnHN »

wg0 wrote: post #198
ImPostingOnHN wrote: post #195

Earlier quoted context omitted.

> Is political engineering also engineering? Good question, even if I've never heard of "political engineering" Is social engineering snake oil, too?

Political engineering is pretty common in societies where deep state or establishment has a firm control and wants to engineer results their way. Most Western democracies are not that.

would you mind answering this question, please?

it's been ignored twice despite being kind of the key point:

> is social engineering snake oil, too?

Post reply on HN