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

I'd rather read the prompt

claytonwramsey.com

Post reply on HN
Print view View original

301–310 of 880 posts

Re: I'd rather read the prompt

#301

Post by lurking_swe »

soerxpso wrote: post #253
sn9 wrote: post #221

> I should hope that the purpose of a class writing exercise is not to create an artifact of text but force the student to think; a language model produces the former, not the latter. It's been incredibly blackpilling seeing how many intelligent professionals and academics don't understand this, especially in education and academia. They see work as the mere production of output, without ever thinking about how that…

> I should hope that the purpose of a class writing exercise is not to create an artifact of text but force the student to think I'm there for the degree. If I wanted to learn and engage with material, I could save $60,000 and do that online for free, probably more efficiently. The purpose of a class writing exercise is to get the university to give me the degree, which I cannot do by actually learning the material (…

That’s a disingenuous argument. You don’t know what you don’t know. Literally. A completely self guided high school graduate following random online materials will not learn nearly as much on their own. Or they will go down rabbit holes and waste countless hours, and not having an expert unblock you or guide you down the right path would waste a lot of time.

Further, some high school graduates (like myself at the time) literally don’t know HOW to learn on their own. I thought I did but college humbled me, made me realize that suddenly i’m in the drivers seat and my teachers won’t be spoon feeding me knowledge step by step. it’s a really big shift.

If you were the perfect high school graduate, then congrats, you’re like the 0.01%! And you should be proud (no sarcasm). This doesn’t describe society at large though.

For the very few that are extremely motivated and know exactly what job they want, i do think we need something in between self guided and college? No BS - strictly focusing on job training. Like a boot camp, but one that’s not a scam haha.

The other aspect of college you ignore is, it is a way to build a network prior to entering the workforce. It’s also one of the best times to date, but that’s another story.

Completely agree that the cost of college in the US is ridiculous though.

Re: I'd rather read the prompt

#302

Post by _AzMoo »

derefr wrote: post #187
musicale wrote: post #141

Earlier quoted context omitted.

> LLMs are like anti-compression algorithms when used like that, a simple idea gets expanded into a bloated mess by an LLM, I think that's the answer: LLMs are primarily useful for data and text translation and reduction, not for expansion. An exception is repetitive or boilerplate text or code where a verbose format is required to express a small amount of information.

There is one other very useful form of "expansion" that LLMs do. If you aren't aware: (high-parameter-count) LLMs can be used pretty reliably to teach yourself things. LLM base models "know things" to about the same degree that the Internet itself "knows" those things. For well-understood topics — i.e. subjects where the Internet contains all sorts of open-source textbooks and treatments of the subject — LLMs really…

I use LLMs a huge amount in my work as a senior software engineer to flesh out the background information required to make my actual contributions understandable to those without the same background as me. eg, if I want to write a proposal on using SLO's and error budgets to make data driven decisions about which errors need addressing and which don't, inside a hybrid kubernetes and serverless environment, I could do a few things:

* Not provide background information and let people figure it out for themselves. This will not help me achieve my goals.

* Link them to Google's SRE book and hope they read it. Still not achieving my goals, because they won't.

* Spend 3 hours writing the relevant background information out for them to read as part of my proposal. This will achieve my goals, but take an extra 3 hours.

* Tell the LLM what I'm looking for and why, then let it write it for me in 2 minutes, instead of 3 hours. I can check it over, make sure it's got everything, refine it a little, and I've still saved 2.5 hours.

So for me, I think the author has missed a primary reason people use LLMs. It saves a bunch of time.

Re: I'd rather read the prompt

#303

Post by bumby »

ebiester wrote: post #276
efavdb wrote: post #269

Earlier quoted context omitted.

If this attitude prevails I would think the value of degrees will quickly diminish.

This seems to be the general feeling of students right now. Academia put itself as a gateway and barrier to the middle class. Why would we be surprised when people with no interest in anything but the goal are not enthralled by the process?

Academia clearly lost their monopoly on information. Since their last moat is a monopoly on credentials, I expect them to defend it intensely.

We could make it less meaningful if employers weren’t so keen on using credentials as their own gateway. That may have more of a chance of happening if the OPs perspective becomes more prevalent and the credential becomes an increasingly worse signal for meaningful skills.

Re: I'd rather read the prompt

#304

Post by casey2 »

Teachers say they would rather read the prompt but the truth is plain that they wouldn't

It's the old joke of the teacher who wants students to tried their best and that failure doesn't matter. But when the student follows the process to the best of their ability and fails they are punished while the student who mostly follows the process and then fudges their answer to the correct one is rewarded.

Re: I'd rather read the prompt

#305

Post by devnullbrain »

jsheard wrote: post #74

I'm far from the first to make this observation but LLMs are like anti-compression algorithms when used like that, a simple idea gets expanded into a bloated mess by an LLM, then sent to someone else who runs it through another LLM to summarize it back to something approximating the original prompt. Nobody benefits aside from Sam Altman and co, who get to pocket a cool $0.000000001 for enabling this pointless exercis…

Yep. They're very closely linked.

http://prize.hutter1.net/

Note the preamble, FAQs, and that all of the winning entries are now neural networks.

Re: I'd rather read the prompt

#306

Post by liveoneggs »

Aeolun wrote: post #260
sn9 wrote: post #221

> I should hope that the purpose of a class writing exercise is not to create an artifact of text but force the student to think; a language model produces the former, not the latter. It's been incredibly blackpilling seeing how many intelligent professionals and academics don't understand this, especially in education and academia. They see work as the mere production of output, without ever thinking about how that…

> It's been incredibly blackpilling seeing how many intelligent professionals and academics don't understand this I figured this out in high school. It can’t be all that uncommon of a thought that if you are already in school and paying and given time to learn, you might as well do so?

https://www.youtube.com/watch?v=tGfiSM4H-7w

Re: I'd rather read the prompt

#307

Post by alganet »

The suggestion that an artificial intelligence follows a specific kind of writing style is a trap.

Relying on that to automatically detect their use makes no sense.

From a teaching perspective, if there is any expectation that artificial intelligence is going to stick, we need better teachers. Ones that can come up with exercises that an artificial intelligence can't solve, but are easy for humans.

But I don't expect that to happen. I expect instead text to become more irrelevant. It already has lost a lot of its relevancy.

Can handwriting save us? Partially. It won't prevent anyone from copying artificial intelligence output, but it will make anyone that does so think about what is being written. Maybe think "do I need to be so verborragic?".

Re: I'd rather read the prompt

#308

Post by casey2 »

nikanj wrote: post #183
jfengel wrote: post #45

Earlier quoted context omitted.

A LLM can't do that? I'm a little surprised. (I know jack all about robotics but that sounds like a pretty common assignment, the kind an LLM would regurgitate someone else's homework.)

The LLM is very happy to give you an answer with high confidence. The answer might be bogus, but the AI will sound confident all the way through. No wonder sales and upper management love AI

Most physics teachers who do this are happy with the BS answer, maybe 1 in a 10,000 have actually tested their problems in reality.

Re: I'd rather read the prompt

#309

Post by skydhash »

necovek wrote: post #70
mezyt wrote: post #39

Earlier quoted context omitted.

> I fear that LLMs have already fostered the first batch of developers who cannot function without it. Playing the contrarian here, but I'm from a batch of developers that can't function without a compiler, and I'm at 10% of what I can do without an IDE and static analysis.

That's really curious: I've never felt that much empowered by an IDE or static analysis. Sure, there's a huge jump from a line editor like `ed` to a screen editor like `vi` or `emacs`, but from there on, it was diminishing returns really (a good debugger was usually the biggest benefit next) — I've also had the "pleasure" of having to use `echo`, `cat` and `sed` to edit complex code in a restricted, embedded environm…

I think there’s a factor of speed there, not a factor of insight or knowledge. If all you have is ‘ed’ and a printer, then I think most of the time you will spend is with the printout. ‘vi’ eliminates the printout and the tediousness of going back and forth.

Same with more advanced editors and IDEs. They help with tediousness, which can hinders insight, but does not help it if you do not have the foundation.

Re: I'd rather read the prompt

#310

Post by pixl97 »

lionkor wrote: post #68
oncallthrow wrote: post #37

LLM cheating detection is an interesting case of the toupee fallacy. The most obvious ChatGPT cheating, like that mentioned in this article, is pretty easy to detect. However, a decent cheater will quickly discover ways to conduce their LLM into producing text that is very difficult to detect. I think if I was in the teaching profession I'd just leave, to be honest. The joy of reviewing student work will inevitably b…

You assume that the teachers job is to catch when someone is cheating; its not. The teachers job is to teach, and if the kids don't learn because their parents allow them to cheat, don't check them at all, and let them behave like shitheads, then the kids will fail in life.

This isn't the way reality works.

Or, bad money chases out good. Idiots that cheat will get the recommendations for jobs where by maxing the grade. The person that actually works gets set back. Even worse society at large loses and actually educated person. And lastly a school is going to attempt to protect their name by preventing cheating.

Post reply on HN