Hacker News Reader

Comment by innodendo | Hacker News Reader

My experience programming is so much different than theirs makes me wonder what I missed out on. I have always programmed on my own, and I can't even think about a single time I have talked to a person in depth about programming (both online and in person). It sounds fun and exciting, but unfortunately I have simply never had the opportunities in life to do so.

For me, AI is the first time I have ever been able to get something resembling an opinion on specific problems/situations that I encounter. I can ask it a very specific question about what the best approach is for what I am working on and it can give me an answer that I read over and consider before deciding on what approach to take. I still frequently get answers that are nonsense, but even then it helps me think deeper on how I should approach the problem because I can ask myself "Are the statements made by the AI true?".

View on HN

AI is your Google and your yellow rubber ducky (rubber duck debugging) all rolled into one.

Absolutely. It can also be your brain if you let it, so…don’t let it.

It’s an absolutely fantastic educational tool if you choose to use it that way.

So are you saying we should ...

... Put down the duckie[0]?

[0]: https://www.youtube.com/watch?v=acBixR_JRuM

Curse you for this ear worm of a song, but also my brain keeps wanting to interpret this as "duck... You're bad at your job and will never amount to anything. Your mother is ashamed of you and you're the reason your father is an alcoholic." hahaha

For the longest time the Android community was super close-knit where people talked about this stuff online and in person constantly, and OP was a pretty active contributor.

Unfortunately Twitter pre-acquisition was probably the focal point and since then, I don't think the community has been the same.

Same, so we started organising workshops, super super cool and you learn so much every time. That said, if you want to meet people around any sorta topic, do it yourself: host a workshop!

You dont need to be an expert to do this. Just share what you have and encourage others to do so as well.

Second order effects should enumerated and reasoned about. This companion perspective is perhaps only one, but very strong indeed

The problem with (propietary AI) is that they (anthropic/google/openai/etc) gain more from the usage of AI than you. Other tools like postgres, gcc, git, HTTP, emacs, etc. don't "gain" anything if you use them (well, they gain popularity and perhaps more contributions, but that's it). The more you use Claude, the richer anthropic gets and the easier for them to position themselves in a place of power, power to dominate the programming of the world. That's sad. So even we all like so much propiertary AIs, we should think twice what we are giving in exchange (and no, it's not just the $200/month what we are giving)

You know it's random nonsense, and yet you still take it on faith that it "gives you answers". That's what makes me despair about all the AI nonsense: it's the death of intellect. Not the death of deep thinking, but of common sense.

I can't wait for the day until all these people collectively snap out of it, and go "what the hell were we thinking with these chatbots".

If AI nonsens is doing the work I ask it to and I confirm it works and is useful .. then what exactly should I snap out of it?

It is just a tool and a useful one. Your loss for not making use of it, but .. you seem to live on a different timeline. I am reminded of old Japan, where the Samurais banned gun powder, so they could continue to fight with swords and not change their ways of living.

When I say it gives me an answer I mean an answer purely in the sense of it is a response to a question, whether or not the answer is correct is for me to decide and I put no faith in whether or not the answer will be correct. I treat the output like an opinion that can sometimes give me an idea for how I can better tackle the problem I am trying to solve.

AI can give good answers to design/architecture problems.

Have it do research (HN, Stack Overflow, technical blogs) and summarize.

Why do you think this wouldn't work?

You'll still need to be capable of understanding and judging the output, and it might miss some options, but it is certainly no worse than pre-AI.

Reply on news.ycombinator.com