Live data from Hacker News

Workers with less experience gain the most from generative AI

mitsloan.mit.edu

21–30 of 120 posts

Re: Workers with less experience gain the most from generative AI

#21
it should be pretty obvious, it's true for devs as well. ChatGPT helped me greatly with getting started with Qt/Qml because I'd never used it before, it's fantastic for exploration. I've worked about a decade in Python and there it does little for me. In fact often code can look very unidiomatic. It makes sense, these models are large and generic and so whether you're ahead or below an average level of knowledge usually determines how much you learn from the output.

When people claim that technology disadvantages workers it usually just reflects their prejudices about working people. So called knowledge workers like to think only they can leverage tech while the call center workers are fixed automatons. But in reality people at a lower skill ceiling have a much easier time leveling up.

ChatGPT when it works well is basically a Young Lady's Illustrated Primer (https://en.wikipedia.org/wiki/The_Diamond_Age). Giving underskilled people free access to information will always benefit them the most.

Re: Workers with less experience gain the most from generative AI

#22
post #6

My experience with Bard/ChatGPT has been that you'll easily shoot your foot off if you don't know what you're doing. Dangerous for workers with less experience. I see claims that it makes you 7-8x more effective. That hasn't been my experience. Maybe 5-10% improvement at best. Instead of Googling myself, LLMs can sometimes give me the answer more quickly than I could find it. Once you ask it anything not easily scrap…

The funny thing about the 10x claims (I've even seen folks say 100x) is that results like these would be readily apparent. Even a broad 1.5x productivity increase would be world changing. And yet, we clearly are not seeing this currently. I'm not saying it will never happen, just that it will be very obvious if it does. 5-10% sounds about right to me.

I dunno computers should be world changing. And they are. Yet the lives of ordinary people have hardly improved and are arguably worse since the 1980s

Re: Workers with less experience gain the most from generative AI

#23
post #6

My experience with Bard/ChatGPT has been that you'll easily shoot your foot off if you don't know what you're doing. Dangerous for workers with less experience. I see claims that it makes you 7-8x more effective. That hasn't been my experience. Maybe 5-10% improvement at best. Instead of Googling myself, LLMs can sometimes give me the answer more quickly than I could find it. Once you ask it anything not easily scrap…

I use it extensively both local ML models and LLMs, LLMs are extremely flexible but can cause headaches if you trust the answers too much, it does help in some situations especially quickly spitting out Python data massaging scripts(after 5-10 back and forths). But it does it, I could write the script myself but I rather save energy on problems LLMs can’t solve.

For simple things LLMs can excel but as you add complexity, you really need to double check

Re: Workers with less experience gain the most from generative AI

#24
post #6

My experience with Bard/ChatGPT has been that you'll easily shoot your foot off if you don't know what you're doing. Dangerous for workers with less experience. I see claims that it makes you 7-8x more effective. That hasn't been my experience. Maybe 5-10% improvement at best. Instead of Googling myself, LLMs can sometimes give me the answer more quickly than I could find it. Once you ask it anything not easily scrap…

This matches my experience. for me it’s really just a better google (but still not good enough to replace it). This includes all the caveats where if you’re truly inexperienced it’s difficult to distinguish the good answers from blogspam quality or confidently incorrect answers.

Re: Workers with less experience gain the most from generative AI

#25

This is exactly the opposite of my experience with it. Getting it to code, it can save me, a fairly experienced coder, quite a lot of time because it can do some boring things - boilerplate stuff, or things I can’t be bothered to look up. And I can tell if it’s done it right or not. I’ve seen people with less coding experience use it and just take its suggestions at face value with painful results and definite detrim…

I think my best experience with chatgpt is when I was digging around a typescript library without knowing the language. it was very convenient to enter type signatures and learn what they mean. I rarely am doing work like this where it’s worth opening chatgpt though.

Re: Workers with less experience gain the most from generative AI

#26
post #17

Earlier quoted context omitted.

The funny thing about the 10x claims (I've even seen folks say 100x) is that results like these would be readily apparent. Even a broad 1.5x productivity increase would be world changing. And yet, we clearly are not seeing this currently. I'm not saying it will never happen, just that it will be very obvious if it does. 5-10% sounds about right to me.

The problem with these claims is that you can't really quantify a "10x" change. I have found a lot of emergent benefits to using LLMs. For example, because my cognitive load of wrangling APIs and understanding and refactoring legacy code and all the other nonsense of my day to day can be so heavily delegated, I actually feel refreshed at the end of the day, and can bang out a decently chunked feature on my opensource…

I'd be really curious if you're willing to expand more on how it has helped with those workflows. Do you copy/paste chunks in and ask it to explain them? Have it try to refactor them and then clean up?

Re: Workers with less experience gain the most from generative AI

#27
Using ML tooling can be incredibly useful for new and experienced folks, - but in different ways. New folks are more likely to ask questions they’d otherwise feel uncomfortable asking but are far more at risk of being lead down the wrong path and skipping over gaining a fundamental understanding of how things work . It’s a balance of risk vs reward.

I’ve been in tech 18~ years and get a lot out of copilot and chatGPT every day.

Re: Workers with less experience gain the most from generative AI

#28

This is exactly the opposite of my experience with it. Getting it to code, it can save me, a fairly experienced coder, quite a lot of time because it can do some boring things - boilerplate stuff, or things I can’t be bothered to look up. And I can tell if it’s done it right or not. I’ve seen people with less coding experience use it and just take its suggestions at face value with painful results and definite detrim…

The study in the article was on call center workers so I wouldn't necessarily expect it to scale to more complex tasks.

Re: Workers with less experience gain the most from generative AI

#29
post #6

My experience with Bard/ChatGPT has been that you'll easily shoot your foot off if you don't know what you're doing. Dangerous for workers with less experience. I see claims that it makes you 7-8x more effective. That hasn't been my experience. Maybe 5-10% improvement at best. Instead of Googling myself, LLMs can sometimes give me the answer more quickly than I could find it. Once you ask it anything not easily scrap…

The funny thing about the 10x claims (I've even seen folks say 100x) is that results like these would be readily apparent. Even a broad 1.5x productivity increase would be world changing. And yet, we clearly are not seeing this currently. I'm not saying it will never happen, just that it will be very obvious if it does. 5-10% sounds about right to me.

I don't think anyone is claiming that chatgpt has made everyone 100x more efficient. Either they are retarded or you misunderstood them.

Re: Workers with less experience gain the most from generative AI

#30
ah their test environment was a call center

feels like there is a viewpoint floating around that chatgpt is valuable in part because of the RLHF work

but does this makes it generally capable? or just prove that you can hire essentially the people you would hire to staff a call center (language skills + good generalists) to create a language model that behaves essentially like a call center (refuses to take any actions that are not allowed)

Post reply on HN