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

871–880 of 880 posts

Re: I'd rather read the prompt

#871

Post by walleeee »

godelski wrote: post #827
walleeee wrote: post #822

Earlier quoted context omitted.

Cheers. I saw that one and liked it. I do wish the title was "It's not just the incentives". Incentives are people and artifices. The latter deceptive or not. But the point being in feedback.

I think the intention of the title was to fire people up a little. Let them cook and not get away with passing the buck. Ultimately the incentive structures are a product of our actions, and the author does make that clear. So in effect, it all comes down to us.

Yeah, indeed it does. Can't know what each ear needs to hear to come alive to that, I guess.

Re: I'd rather read the prompt

#872

Post by musicale »

_AzMoo wrote: post #302
derefr wrote: post #187

Earlier quoted context omitted.

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…

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

If they won't read a relevant section of Google's book, why would they read an LLM-written version?

Re: I'd rather read the prompt

#873

Post by robertlagrant »

helloplanets wrote: post #167
robertlagrant wrote: post #107

Earlier quoted context omitted.

I wonder if you could invent a teaching language so the LLM wouldn't know about it. A little drastic, but still.

Pretty sure you could just give an LLM the given docs / course material and it would be able to write the language to a reasonable standard. Especially if it had sensible error messages.

Hah, that's true.

Re: I'd rather read the prompt

#874

Post by bumby »

andai wrote: post #486
hnfong wrote: post #449

Earlier quoted context omitted.

[flagged]

Our CS department head was overheard saying, "Our job is to create researchers." I found this quite striking, since something like 10% of undergrads go into research. Most people really are there to help them get a job. So the program is designed not to meet the needs of 90% of its "customers".

There’s an argument that treating students as “customers” has led to all kinds of bad outcomes. One example is it creates an incentive to invest in all kinds of fancy infrastructure (fancy halls/dorms and even lazy rivers) because that is how they attract more “customers” but this ultimately becomes a huge driver of educational costs. The same can be said for watered down courses etc.

Re: I'd rather read the prompt

#875

Post by MengerSponge »

sgustard wrote: post #410
MengerSponge wrote: post #387

Earlier quoted context omitted.

We (America) made university educations super expensive. Causality is complicated, but it probably started as an effort to destroy the anti-war movement in the 70's.

"Most financial experts attribute he sudden increases that started in the 1970s with an influx of federal funding designed to make college more affordable." https://educationdata.org/average-cost-of-college-by-year

Expensive things can be affordable. A house is expensive, but with stable employment and a bank loan you can afford it.

When we're talking about college costs, public schools are really the only institutions that matter. The Ivies (or even Ivy+) are a rounding error compared to the big midwestern land-grant universities and the UC and Cal State systems.

States have substantially reduced their per-student support for universities: https://www.ppic.org/publication/higher-education-funding-in...

This coincides with federal funding programs, but student loans are famously not dischargeable, which makes them more instruments of social control than conventional financial vehicles.

Re: I'd rather read the prompt

#876

Post by thomasvn »

Generally agree, using LLMs to expand on your work is usually just boilerplate or garbage.

What about using LLMs to refine or sharpen your existing work? Similar to a Rubber ducky? If you're intentional about maintaining and understanding the theory behind the work, I've found it a useful tool.

Re: I'd rather read the prompt

#877

Post by icelancer »

godelski wrote: post #505
icelancer wrote: post #475

Earlier quoted context omitted.

I am generally education blind. Mostly I would say graduate school experience has been a net negative on my teams overall, if anything.

The second half seems to contradict the first. But I do condone the actions, as that's the only way things can change.

The second point led to the first one over time.

Re: I'd rather read the prompt

#878

Post by fennecfoxy »

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…

Blame capitalism & endless growth. The blame for the shortening of the journey from input to output lies solely with those that we love to see become billionaires.

Re: I'd rather read the prompt

#879

Post by tcmart14 »

bumby wrote: post #874
andai wrote: post #486

Earlier quoted context omitted.

Our CS department head was overheard saying, "Our job is to create researchers." I found this quite striking, since something like 10% of undergrads go into research. Most people really are there to help them get a job. So the program is designed not to meet the needs of 90% of its "customers".

There’s an argument that treating students as “customers” has led to all kinds of bad outcomes. One example is it creates an incentive to invest in all kinds of fancy infrastructure (fancy halls/dorms and even lazy rivers) because that is how they attract more “customers” but this ultimately becomes a huge driver of educational costs. The same can be said for watered down courses etc.

I would also add, look at the astonishing success (sarcasm), of for-profit colleges. They would have to be considered the extreme of treating students as customers and I don't really know any serious employer who accepts a degree from for-profit colleges. Unless the college is so new the employees hadn't heard of it before and don't know much about it.

Re: I'd rather read the prompt

#880

Post by _AzMoo »

musicale wrote: post #872
_AzMoo wrote: post #302

Earlier quoted context omitted.

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…

> Link them to Google's SRE book and hope they read it. Still not achieving my goals, because they won't If they won't read a relevant section of Google's book, why would they read an LLM-written version?

Because it's formatted as an introductory part of a document they're already going to read.
Post reply on HN