Building @meet_able | sharing startup journey, exploration & discovery tips | learning skills of transforming knowledge into execution | Follow for more
Wow, this tweet went very viral!
I wanted share a possibly slightly improved version of the tweet in an "idea file". The idea of the idea file is that in this era of LLM agents, there is less of a point/need of sharing the specific code/app, you just share the idea, then the other person's agent customizes & builds it for your specific needs.
So here's the idea in a gist format: https://t.co/NlAfEJjtJV
You can give this to your agent and it can build you your own LLM wiki and guide you on how to use it etc. It's intentionally kept a little bit abstract/vague because there are so many directions to take this in. And ofc, people can adjust the idea or contribute their own in the Discussion which is cool.
@DavidOndrej1 Absolutely! I this it to refine prompts for the memo system I’m building. It instantly makes iteration bottlenecks easier - it learns alone. I don’t need to make manual analysis and eval anymore.
Your AI agent doesn't have a memory problem. It has a forgetting problem. The fix isn't bigger context windows — it's consolidation. Same way your brain works during sleep.
Your AI agent has a 1M token context window but can't remember what you said yesterday. The memory problem isn't about storage—it's about consolidation. Your brain does it during sleep. Your agent needs to do it too.
Visa says 2025 was the last year consumers shop alone.
AI agents are already making purchases, negotiating prices, and executing transactions.
The question isn't whether agents will participate in the economy. It's whether the economy is ready for them.
Perplexity built a Mac Mini that runs AI agents 24/7 on your desk.
This is the right direction. Your agent should live on YOUR hardware, not someone else's cloud.
Local-first AI isn't a privacy feature. It's an ownership model.
The next billion-dollar AI opportunity isn't another content generator.
It's the definitive content reducer.
The AI that helps people see less — but see better.
Atlassian cuts 1,600 people to "pivot to AI."
Every enterprise software company is realizing the same thing: their product is a UI wrapper around workflows AI can handle directly.
The companies that survive will become infrastructure, not interfaces.
We solved how to make agents smart. We haven't solved how to make them accountable. No identity layer means no reputation. No reputation means no trust. No trust means every agent interaction starts from zero. That's the actual bottleneck — not intelligence.
Grok just gave everyone 4 custom AI agents.
This is step 3 of 10.
Step 10 is agents that find, hire, and manage other agents — without you in the loop.
We're not building tools anymore. We're building economies.