Post

I don't think "smart" actually has all that much to do with effective use of LLMs. Experience, sure. But I bet that good conversationalists have the most important skills: clarity, patience, curiosity, empathy. These skills translate pretty directly from human conversation to LLM chat.

The issue is that there's significant nuance between "using modern AI tools" and "using modern AI tools *effectively*"

Absolutely that I can see why people assume it must be easy given the interface, but wow there's so much hidden depth there

(I get so sad when I hear people dismiss their own ability to learn things, I expect that the real problem is that they were given an inaccurate mental model of the level of difficulty involved and how much work they would need to put in - I see this for learning to program all the time)

I feel like there is a split in the community, I've tried AI enough in a coding context to know it will never be safe to use so I wouldnt want to work with or consider someone who uses it a programmer personally

How about someone like me who uses it on a daily basis but meticulously reviews and revises the code it writes (almost the exact same process as code reviewing PRs from a new contributor) and lands them only after ensuring they have the same or higher quality as code they would write themselves?

Oh man, I remember Reddit writhing in scorn when the term “prompt engineering” took off. The assumption was that it’s so trivially easy there was no need to overcomplicate it. Fast forward to now, when Anthropic Console will burn 60 seconds optimizing a CoT prompt for you.

To be fair, the complaint was you won’t end up with a ‘prompt engineering’ role. Which is true. Engineers are expected to learn how to engineer prompts, you don’t go hiring for someone to do this exclusively.

I think we are in the early days yet. We don't know how specialized LLMs may become. A sufficiently specialized LLM in the future may well trade features like constraints on prompting for more deterministic or correct output. The analogy today is database query engineers.

I've seen the complaint you reference, but I'm talking about stuff like this (mostly from 2+ yrs ago): "I think I get a little more angry every time I see someone say “prompt engineer”. [...] Like holy fuck. You are typing words to an LLM. It’s not complicated and you’re not engineering anything"

I don’t know what you do, but the people developing the tools we see on commercials tidying up emails and whatnot need to be thrown into a volcano.

I haven't seen many of the ads, but if they're promoting "rewriting your emails" as the key value you can get out if this tech we have even more education work to do!