Tired of managing 50 Word docs for job applications? I just joined the waitlist for @ResumeFolio — one workspace to tailor, export, and track every resume. https://t.co/SgbgCBthSF
Happy to launch my space on topmate – a hub for my community.
It’s where I’ll share opportunities to engage, learn, and collaborate as I grow and create.
Come be part of the journey.
https://t.co/MdbCq6noXx
And if you say that a text/button based touch interface will go away then how do we access - say a bank app? Or send a message to family/friends? Say it out loud? Maybe these experts don’t take public transport.
No more knowledge workers but thought workers. Knowledge is almost free but thinking is still very expensive as it requires years of being a practitioner in the domain.
https://t.co/l6RDBa3TLu
Note to myself: Follow the excitement of creation, not the hunger for validation.
Creation is internal.
Validation is external.
One builds momentum.
The other pulls your attention away.
Finally approved as an Author on Medium. Still trying to feel what that means. Honestly? Feels like “no big deal.” But friends keep reminding me — writing a 190-page book is a big deal. Maybe I should pat myself on the back.
Most “new” data online is synthetic, repetitive or derivative. The next breakthrough in AI won’t come from scaling models up — it’ll come from rediscovering something deeply human: fresh ideas, authentic voices and real thinking. PS: Written by a human.
We’re running out of useful data to train AI. The sum total of digitised human knowledge is finite — and much of it is junk. Everyone’s still talking about bigger models & more compute, but the real limit isn’t hardware anymore… it’s data worth learning from.
When the cleaning goes too far, bias gets concentrated and facts get distorted (see “Google Nano Banana” & “Founding Fathers” fiascos). Sam Altman hinted it’s odd to think the best way to train a model is to generate a quadrillion tokens and feed them back.