https://youtu.be/AlkkAKzkXQM?si=gmpJx192uHDyJ6qP

I’m afraid to watch this because every conversation between a left brained person and a right brain person doesn’t seem to go well.

Investing @variant, writing at https://backoftheenvelope.substack.com/

School really taught us that you need to attain some level of accuracy before you’re creative.

The truth is the creativity was already there before you even entered school. 

Accuracy is correlated with creativity, but too much accuracy and it becomes automation.

https://x.com/fchollet/status/1870169764762710376?s=46

Oh oh oh 🎅🎅

Today OpenAI announced o3, its next-gen reasoning model. We've worked with OpenAI to test it on ARC-AGI, and we believe it represents a significant breakthrough in getting AI to adapt to novel tasks.

It scores 75.7% on the semi-private eval in low-compute mode (for $20 per task… https://t.co/ESQ9CNVCEA

François Chollet

Gemini-2.0-Flash-Thinking #1 across all categories! https://t.co/mRctNA31B9

This model is performing pretty well on the lmsys arena.

https://t.co/UdznaNGaL5

Jeff Dean

https://x.com/jeffdean/status/1869794852516573497?s=46

https://www.notboring.co/p/the-return-of-magic?r=ezk3i&utm_medium=ios&utm_campaign=audio-player

Interesting and weird read. I wonder if most people have weird experiences but don’t say it.

Upgraded my phone. Made me realize half of the ux is the phone itself. Everything looks better.

frames-v2

Building @warpcast and @farcaster, new dad, like making food

https://x.com/fchollet/status/1869578315952197797

o1 performs poorly on ARC-AGI eval. When we say it's a reasoning model what do we mean in technical terms?

Verified o1 performance on ARC-AGI's Semi-Private Eval (100 tasks)

o1, Low: 25% ($1.5/task)
o1, Medium: 31% ($2.5/task)
o1, High: 32% ($3.8/task)

o1 performance on the ARC-AGI pub leaderboard at different compute levels: 25-32%, solid progress compared to o1-preview

learnonline

/higher

sowing the seeds of Qilin’s Garden /qilins-garden 



The cofounder started a company called Synthesis. They had complex games for children to play, and they are currently trying to build an ai tutor.

Co-founder @bountycaster (bountycaster.xyz). Co-producer @ethereumfilm. Previously co-founder of Scalar Capital and product at Coinbase

This is a smart blog

https://blog.usv.com/four-futures

12 hour lay over on my flight back home 🫠

I really like deconstructed pho (pho to go). The veggies don’t get cooked so it tastes fresh. The fat of the meat doesn’t go into the broth so it’s still clear and not foggy.

Sapiens qui prospicit - Wannabe Polymath. Interested in art, music, history, philosophy and religion.

More posts like this! It’s been a while since I read this book.

Notice that the future is built on the backs of definite optimists. Everyone else is just a bystander.

Indefinite optimism says things have been trending up and will continue to do so, but it presents no solutions. 

Definite optimism is the confluence of deregulation, crypto, ai, nuclear, drones, solar, and other technologies to thread the needle.

There are four possible futures

Indefinite pessimism says America is going to fall similar to Rome falling.

Definite pessimism says things will at best return to the mean and continue on trend. The debt will grow. Doge will fail. Violence will grow.

Founder of MidJourney https://t.co/D0yt1zjLIt

Sar Haribhakti

https://x.com/sarthakgh/status/1868881401283764698?s=46

John Hoang

@jhoang