Sophia (@sopharicks) 's Twitter Profile
Sophia

@sopharicks

Community builder. Ex-OpenAI. Founder of @BuzzRobot.

ID: 4724768053

linkhttps://www.youtube.com/@BuzzRobot calendar_today07-01-2016 18:44:23

637 Tweet

1,1K Followers

1,1K Following

Sophia (@sopharicks) 's Twitter Profile Photo

When Elon says 'my dear friend Ilya' (in the YC interview), does it mean we should expect SSI to join x.ai at some point? I wouldn’t be surprised since Elon recruited Ilya to start OpenAI. If SSI ever exits - it will be either Google or x.ai

Emmett Shear (@eshear) 's Twitter Profile Photo

I dislike the way that all the deep research products go off and work without you. I like that they show a running list of what they're doing...but I wish I could give suggestions and ask questions asynchronously, without stopping the progress.

Sophia (@sopharicks) 's Twitter Profile Photo

Just realized that Deedy is my number 1 source of AI news and interesting takes on AI. A great example of a new public relations playbook: become a valuable source of information and people in the industry will know about you and what you build. I think this is especially

Pranav agarwal (@pranav_al) 's Twitter Profile Photo

Academia may have some issues, but the best thing it allows us is the independence to learn and collaborate with like-minded explorers. During this last phase of my PhD journey, I started having weekend meetings with the most inspiring and supportive Ioana Ciuca (ioana ciucă). We

Academia may have some issues, but the best thing it allows us is the independence to learn and collaborate with like-minded explorers. 

During this last phase of my PhD journey, I started having weekend meetings with the most inspiring and supportive Ioana Ciuca (<a href="/errai34/">ioana ciucă</a>). We
Emmett Shear (@eshear) 's Twitter Profile Photo

Just finished reading this book. It’s the best book on the nature of mind and intelligence I’ve ever read, can’t recommend it more highly. My favorite highlight covers the way in which imagination and attention are the same thing.

Sophia (@sopharicks) 's Twitter Profile Photo

I'd bet my money on a team with 1 superstar and the rest are strong players but not superstars. Much easier to align the team, rather than a team with only superstars, who are entitled to their opinions and might have a hard time aligning around one leader. Meta's strategy of

Andriy Burkov (@burkov) 's Twitter Profile Photo

On the time savings with using LLM for coding: On one hand, you can code in 5 minutes what would take a day by hand. On the other hand, you can spend 2 days fixing a bug that would take only 5 minutes by hand.

Andriy Burkov (@burkov) 's Twitter Profile Photo

We have a very poor understanding of why deep neural networks like transformer models learn the parameters they learn. For example, in the paper below from 2013, the authors demonstrated that 5% of the weights of a trained deep neural network can be used to predict the values of

We have a very poor understanding of why deep neural networks like transformer models learn the parameters they learn. For example, in the paper below from 2013, the authors demonstrated that 5% of the weights of a trained deep neural network can be used to predict the values of