I did both the back end in N8n and the front end in HTML using Claude/MCP and was able to bang this site out in about three hours.
I’ll shut up because y’all aren’t interested in it, but I figured I’d give you one more to look at before I shut up.
Really, three and a half hours from nothing to this.
awesome!
I dont think its no one is interested
More like This doesnt apply to what I AM working on so its an interesting curiosity for me at best
i, personally, love to hear what other people are doing, how, and why.
I dont disagree
Just what Jays doing & using AI for I have no need for at this time
I’m not building the sort of things he is
In fact I’m learning another toolset so even if I could use AI for it I probably wouldnt since I want to understand things very well
And that means me sitting down & learning it
I cant judge if the AI writes me good code or crap (like it does with Xojo) if I cant even read it myself
So while interesting, its not something I’m likely to use at this time
I haven’t said much because I’m tempted to offer (pen)testing and I don’t have the time. 
I’ve been recruited by my company to do some ‘vibe’ coding. Don’t know the details yet but it should be…interesting.
Bob, I’ve learned a lot of tricks. You know how to get a hold of me.
Oh, we’re getting ‘training’ on it in a couple of weeks. Our company loves its meetings.
Basically what I think Jay is saying is consistent with what this guy is claiming
IF you happen to be working in a toolchain where this works great for you
I tried vibe coding a Discourse plugin, and for all the condescending attitude the AI had, eventually we got something working. For languages that are well documented, it can work well enough.
how do you ensure security?
A different perspective on using AI extensively
Norman, it clearly sucks as a software architects, but for grunt programming, it can’t be beat.
Depending on what languages you’re using
Some are better understood than others
Any time I see Xojo code from prompts its littered with VB, VB.Net and lots of things that even seem completely hallucinated
PHP, C# Java and others not so much
YMMV
I have a long running thread with an AI that is still trying to convince me it can write PHP to sanitize HTML using regex.
And every time I ask it a question, it’ll be like “wait, I’ve got it this time”
And it’s an increasingly convoluted way of breaking everything every time 
hallucinating AI’s is a thing
the more they get trained on their own slop the worse they get
I’ve seen academic research papers showing this happens
Not to say they CANT be useful or helpful but they need to be trained with KNOWN good inputs
Garbage in garbage out still holds true IMHO
same here. The AI was really nice, friendly, trying to help, excused their mistakes. I ended up with a version that compiled with no errors - but with zero usability. Anyway - the AI was much more ‘human’ than the vendors developer site…