SAY THIS LOUDER FOR THE DENIERS IN THE BACK.

Bonjour !

Contactez-moi via SimpleX Chat :

https://simplex.chat/contact#/?v=2-7&smp=smp%3A%2F%2Fhpq7_4gGJiilmz5Rf-CswuU5kZGkm_zOIooSw6yALRg%3D%40smp5.simplex.im%2FocWN1F6Xy4FLkQUCPtmH3PY1e9IoEb96%23%2F%3Fv%3D1-4%26dh%3DMCowBQYDK2VuAyEAgxg8S4IgT4nyu_ZgoguzAw3Hdpzz0Rb8URRAa7ZSsAU%253D%26q%3Dc%26srv%3Djjbyvoemxysm7qxap7m5d5m35jzv5qq6gnlv7s4rsn7tdwwmuqciwpid.onion

nostr:

This isn’t true for large apps yet, maybe in the future but at the current moment it isn’t realistic. Smaller scoped apps this is viable though

I’m working as a product manager for 10+ years now and the fact that I can do real world user testing of new features without talking to my developers is massively underrated at the moment. It’s so much more powerful to demonstrate a “working feature” than writing boring user stories and wasting time in Figma.

Absolutely true. Clear thinking is more important than anything else in this new era.

Generalized intelligence is being greatly empowered by AI. It’s definitely going to change most industries. AI basically gives every generalist a staff of specialists

I was never the strongest dev, but I coded when needed. I was, however, the stronger UI/UX oriented product owner role that delivered to the user stuck in the workflows. I vibe coded a RE/BTC calculator last weekend, see my feed. Now I’ve spent a week developing a MDC based repo that blends task driven AI development guidance with recursive agentic execution of scope with User as product owner requiring key check-ins and UI explorations and scope development prior to execution by agents. Going to build a nostr:npub126ntw5mnermmj0znhjhgdk8lh2af72sm8qfzq48umdlnhaj9kuns3le9ll based nostr product with it as a test. Wish me luck. May suck. But it’s where we’re headed. I can already tell.

let's vibe code llms and it's infrastructure. what could possibly go wrong

I'm in the first group, and I love me some AI prototyping, but the end result is better, when I have the architects and engineers on board during the process.

I think anyone who sees this as a competition between the two sides, rather than a path to tighter integration, is missing the point.

Yeah, the overall quality of apps we are forced to use in our everyday life is going to skyrocket in the same way the richest companies like google have improved their customer support over years!

Sure, for coding green field prototypes or trivial bug fixes. For everything else you need to actually know what you're doing.

LLMs are an incredibly leaky abstraction when it comes to building software.

You're on fire, Cameri.

Hop skip and a jump , it’s like leap 🐸 (frog)

i think its more broader than that - AI gives people with high agency super powers, NPCs will benefit from it a lot less just because they are either in denial or can't be bothered to learn how to use it or change their ways

🎯💯

Claude code's plan mode would get you pretty close.

Its great for 0->1 folks ;) I love it. I can do marketing, product design, architecture, development, debugging sosomuch faster than before. Maybe the future will thrive with more solopreneurs.

Also less painful to onboard jr. dev. That's one thing I've noticed. but they still lack fundamentals when it comes to debugging and architecture. I'll take that as a win as its less work for sr eng.