I've caught the AI bug and it's fun

I'm still learning my place with AI. I used to use ChatGPT - let it do it's thing, and let it create powershell scripts that would inadvertently crash themself.

Over the last 6 months, I've been paying for Anthropic's Claude, only at the £20 price range and very rarely used it. When I did, it was fantastic and did it's job. But I've now found Claude Code, which I knew was a thing but didn't quite understand it. I have been using it almost everyday for personal and work projects over the last 2 weeks and, I am astonished over how far we've come from the last 3 years.

I used AI early on, 3-6 months before IT people started to hear about it and about a year before everyone else knew about it, so I saw the huge pitfalls and it's hallucinations, it's very severe and unavoidable hallucinations. What I have been using it for lately has blown my mind. I have a very very basic understanding of HTML/CSS, and even less understanding of back-end development, but with Claude a lot of that becomes unimportant. I've found that with the right Claude Skills and knowing how to ask it the questions, I've been able to clear up a lot of the 'slop' that AI brings.

Will I be up the creek if there were to be a huge bug? Yes. But until then the bots can keep helping me develop what I'm asking for. I don't think I'm asking for anything amazing, nor am I asking for anything that is too public facing so those pesky security holes are not wide out there.

Again, one of the skills I use is to have it do security scans of the code, and it does come back with some interesting returns. Usually internal escalation paths.

Either way, AI is here to stay and I'm all for it.