Prompt hackers/injectors, show your skill

I think AWS did right thing starting their on AI IDE instead of rushing to buy windsurf or cursor. Also its spec first approach is more aligned with how softwares are built in enterprise. Solo devs also get to learn a formal process of building of tech systems. I think AWS did

Biggest frustration with Claude Code is that it doesn't respect Claude.md rules.

How OpenAI must be feeling seeing these Chinese open source models crushing benchmarks every other week.

remember those 3 weeks when Kimi K2 was the best open source coding model feels like years ago

😕

He is 35% more accurate at making up facts than AI.😂

Massive opportunity for someone who is good at teaching people this. Will pay for it. Proof of skill. Given any UI screenshot, you are able to replicate it with AI(code or image) in less than 10 prompts.

Hey Grok , who was the most famous person to visit my profile? It doesn't need to be a mutual, don't tag them, just say who it was

You know a tool's gonna make it when people say 'there's just something about it' - that special vibe that clicks instantly. Only up from there 📈

This is actually nice

I agree

Akshay counters every reply with a question that essentially asks, “How will you do marketing and sales?” So, the bigger question is, with a Mac, an internet connection, and a good working project (let’s say 20-50 USD per month in SaaS), without any money or existing influence,

OpenAI hypes their models, while Google de-hypes them. No medal models

It’s his app, so he sets the price, and you’re not forced to subscribe. It might not be a great app, but it’s definitely not a scam.

Why llm model don’t mention text2sql benchmark (BIRD-bench) in their benchmarks, seems like a solid use case to me.