Will keep saying this : super bull on OpenAI

You do think as a parent how to turn your child into an independent thinker You do think as an LLM user how to make sure the model can think on its own and doesn’t try to overly please you

Coming soon. 😉😉😉

The CEO of Nvidia has a legendary LinkedIn profile:

I wait for the sun to set and the night to fall to put my kiddo to bed , to go do my sets, put away my weights and wrap with some chin-ups and sit down to——- vibe code 👩‍💻😌😌😌

If your timeline is overwhelmed with tech and politics and memes , maybe just time to touch some Grass and dive into some Art . Isabella is my choice of the day. Isabella Ducrot | Studio | Frieze Masters 2024 youtu.be/AYsBjyJ1I4w?si… via YouTube

Sunday inspo

Sunday fun

There’s a lot of arbitrage in picking just at the quality threshold about to pass or capability and area with LLMs vs forever long tail, I feel like I’m getting there , it seems very close but might be another cycle or couple of years if not more

Everything is a scam until it isn’t

Let’s go ! Never been this bullish on someone . Drones to the moon

It’s losing it with my art which is not sane in the first place 🤣🤣🤣🤣

What’s the price tag on smth that’s extremely useful vs smth that’s more intelligent but less useful . Can the delta be covered by the input and the guidance ? Just thinking out loud. The shift to useful vs simply smart is clear and I wonder if that’s going to accelerate

My biggest fear with these models rn and the amount of time I spent with them is becoming super predictable and banal myself ; also certain brain muscles are definitely on holiday for a while now and others are overworked as hell

I’m going to say this. There’s many claims I am seeing and an overall underwhelming sentiment around GPT5 . The issue is that when you get super capable models already the next leap in intelligence is not obvious and also depends on what you define as intelligence . I see the

The gap between creatives and AI needs to be bridged , we can do so much this