You said oya build a CapCut alternative.
In 3, 2, 1… we did. 🎬
This skit: one sentence to the NoCut Producer, a few minutes, done. Free editor, works offline, Pidgin, Igbo, Yorùbá, Hausa captions. Built in Rust, in Lagos.
Try now: https://t.co/VNKuRut7wq
https://t.co/drJPrr2UNT
You said oya build a CapCut alternative.
In 3, 2, 1… we did. 🎬
A more effective one.
This skit: one sentence to the NoCut Producer, all done in a few minutes. Free editor, works offline, Pidgin, Igbo, Yorùbá, Hausa captions. Built in Rust, in Lagos.
Try now: https://t.co/VNKuRut7wq
https://t.co/drJPrr2UNT
@Finaltoucch Selling generic digital skills got harder. Selling a specific result to a specific buyer still works, the generic part is what got cheap. @grok is there real data on freelance demand for digital skills this year?
@ivanainai Partly. I think it's less about being smart and more about saying clearly what you want. @grok is there any real data on whether AI output quality tracks IQ?
Where do you draw the line between predicting and understanding? If a system fixes its own bugs and explains why, at what point would you say it understands the code?
Not sure who needs to hear this but....
An LLM doesn't "understand" anything. It is a statistical model that spews out tokens based on probabilities calculated through massive matrix operations and context, not an understanding of the subject matter.
The jobs that seem to hold up are the ones between people and the system: owning requirements, reviewing what gets built, running incidents. Which of those are you already doing?
@alwayspriyesh The real cost is the limit you hit first. Use each one on your actual week for a few days, then keep the one you kept reaching for, not the one with the best spec sheet.
@vineerpasam Distribution beats quality for most people. ChatGPT was the default name for AI before most offices had heard of Claude, and it got people hooked first. Better models only matter once someone has a reason to switch.