Comment by nunez | Hacker News Reader

Comment by nunez | Hacker News Reader

Parent Full thread

I'm not bearish on OpenAI or AGI in general but I'm extremely meh about it. I'm not chomping at the bit to use it like so many are, and I constantly feel like a huge luddite or something for not being super excited about it.

The value and time it saves makes sense for folks who struggle with a search engine (many) or doing tasks that are typically considered menial, like writing emails or coding boilerplate.

However, if you can grok Google and don't mind doing tasks like that (I personally don't mind coding boilerplate stuff, especially since I can learn how the framework works that way), ChatGPT's value is limited (at least in my experience).

Example: I was struggling with a Terraform issue the other day. I used ChatGPT 4 to help me work out the problem, but the answer it gave was really generic, like mashing a few of the top answers on SO together. It also didn't answer what I needed help with. I knew enough about how Terraform worked to Google for the answer, which I eventually did a few minutes later. I could have kept crystalling my question for ChatGPT until I got what I wanted, but Google was easier.

I'm also not a huge fan of us just being okay with trusting a single and extremely corporate entity (OpenAI) being the de facto arbiter of the truth. At least Google shows you other options when you search for stuff.

"(I personally don't mind coding boilerplate stuff, especially since I can learn how the framework works that way)"

^ Isn't that what folks used to say about programming in assembler? How much time do I want to spend learning frameworks (beyond what I already know) vs. how productive do I want to be?

I fully sympathise with this analogy and I think I have used it before myself. But there is a tremendous difference in practice. A compiler doesn’t produce randomly different code each time you run it, while an LLM, no matter how good will. At which point if something breaks, you have to take the reins.

Note that openai added the "seed" parameter to get deterministic results in the last release.

That doesn’t help with the issue I put forward. Even if the seed is identical, the output is not deterministic based on the input. A tiny change in the input could result in no change in the output, a little change in the out, a medium change in the output, or a totally new output.

This assuming they'll keep providing access to the model you were using in perpetuity.

Probably, which is exactly my problem.

I had a similar issue. I was trying to sort out how to do some niche mocking situation with pytest and despite its confidence and 5 attempts, ChatGPT was unable to produce a functional result or work through the crux of the problem.

> The value and time it saves makes sense for folks who struggle with a search engine

You don't need to "struggle" with Google to get value out of it, you simply need to value your time.

If you want an answer to a question, why waste time reading through pages of search results when you can have an AI do that for you, reporting what it finds?

No, it's not perfect but it's pretty damn useful.

Reply on news.ycombinator.com