Learn to fine tune a model this year
Learn to fine tune a model this year
Learn to fine tune a model this year
Learn to fine tune a model this year
Learn to fine tune a model this year
Start small
Start slow
But make sure you start
@mattpocockuk What is the ideal number of user stories the spec skill should create
For reference, just creating scaffolding, the skill created 65 user stories! 🤯
@markpinc loving your book! While reading I realised, it should be an app! It will be of great value to founders building with AI in consumer space. Should I go for it?
Day 3/n - Managing Agents
Came across this very similar project - VINEYARD. Which does the same. But the long game is different...
Time to learn from it...
I have multiple active projects
Keeping track became harder than managing the actual work. So I built AI Kanban Board (v0)
- Every card is a Claude instance
- Each task maintains its own context and state
Auto PRs
Unopinionated workflow
Instead of managing chats, manage work
Been exactly a year since we launched water purifiers
We now have TDS data from tens of thousands of Atomberg water purifiers installed across the country
And we combined that with publicly available data from different government sources like central pollution control board, jal jeevan mission and central ground water board
And built a free tool where you can enter your pincode and check your water quality- TDS levels, TDS variance, chemicals present etc
Claude Fable has been disabled worldwide.
This is a sharp wake-up call for India. We must urgently invest in and build sovereign AI solutions right here at home.
We’ve already outsourced our communication infrastructure to Meta. Let’s not repeat the same mistake with AI.
Once Indians get a real taste of what AI can do, there’s no going back.