It's easy to get carried away with AI — either by avoiding it completely or by replacing your entire team overnight. But balance is everything. ⚖️
Real signals — like a DM, comment, or social post — can impact your business in ways AI can't always detect. 👀 That's why oversight and human judgment still matter. Trust AI, but always verify. ✅
🎧 Listen to the full episode here 👉 https://youtu.be/sGtPfP7cBao
#AIinBusiness #MarketingStrategy #TrustButVerify #AIBalance...
But, on the other hand, just now ChatGPT made me waste about 30 minutes because it lied to me about the push() method on Rust vectores. It says it gives back an usize option with the index, which is what I wanted in order to get something done. Spoilers, it does not.
So, the bot I'm writing for the fedi is intended to be used for simple games. I'm writing it in rust and because mastodon-async is, well, async, I'm having to learn corners of the language I hadn't touched before.
Using ChatGPT as a tutor is very handy. In principle I could reread the rust book until I get it, or I could find a real human to ask questions from. The first option I already tried, and inevitably I forget details or fail to understand edge cases or the like until I gain some practice. The latter would be a waste of a human's time.
I have a Game trait which implements the interface I want to be able to add different games. Then I added the one element I have to a vector I iterate on. I didn't realise that the vector is taking the concrete type, not the trait.
ChatGPT took me from understanding this situation, to using a boxed dyn trait object, rewriting my trait so that it is Send, setting Sized as a trait bound for associated methods, and getting tests to pass.
Could I have done it without it? Sure. Especially if I had someone to guide me. This is a privilege not anyone has (just using the word without political valence here).
So it's definitely made my life easier.
I'm going to start posting on my experiences with #AI. My impression is a lot of AI posting lacks balance, either in the "it's all a Markov bot and useless" direction, in the "omfg it is perfect" direction, or in the yet weirder "the world is ending" direction.
I'm going to use the #AIBalance hashtag for it and I invite people to use it and engage with it if they so choose.
If I get enough ideas written down without getting shouted out too much, I will try to write some longer article.
servenomaster