Dwarkesh Patel(@dwarkesh_sp) 's Twitter Profileg
Dwarkesh Patel

@dwarkesh_sp

Being pretrained

Host of Dwarkesh Podcast

https://t.co/3SXlu7fy6N
https://t.co/rEhnfYywXY
https://t.co/hQfIWdM1Un

ID:1209960539390201864

linkhttps://www.dwarkeshpatel.com/ calendar_today25-12-2019 22:14:46

4,1K Tweets

55,6K Followers

704 Following

Follow People
Dwarkesh Patel(@dwarkesh_sp) 's Twitter Profile Photo

AI timelines should make us take Peter Thiel's famous question much more seriously:

'What is your 10 year plan, and why can't you do it in 6 months?'

account_circle
Jason Crawford(@jasoncrawford) 's Twitter Profile Photo

Announcing the 2024 The Roots of Progress Blog-Building Intensive, the second cohort of our 8-week program for aspiring progress writers to start or grow a blog.

Learn about progress studies, get into a regular writing habit, improve your writing, and build your audience

Announcing the 2024 @rootsofprogress Blog-Building Intensive, the second cohort of our 8-week program for aspiring progress writers to start or grow a blog. Learn about progress studies, get into a regular writing habit, improve your writing, and build your audience
account_circle
sophie(@netcapgirl) 's Twitter Profile Photo

this is what people were saying when Dwarkesh Patel was starting out & now he has one of the best podcasts in the entire game because of an unwavering commitment to quality. i think a relevant takeaway is that you should focus on something people want vs overall market saturation

account_circle
Dwarkesh Patel(@dwarkesh_sp) 's Twitter Profile Photo

Meta is going to have 350,000 H100s by the end of the year.

Given lead times, they probably had to start ordering them in 2022.

How did Zuck know he'd need all these GPUs?

account_circle
Dwarkesh Patel(@dwarkesh_sp) 's Twitter Profile Photo

'More of what we call training for these big models is actually going to be inference generating synthetic data to then go feed into the model

We trained Llama 3 on around 15 trillion tokens.

Our prediction was that it was going to asymptote more, but even by the end, it was…

account_circle
Dwarkesh Patel(@dwarkesh_sp) 's Twitter Profile Photo

Last 28 days 🤯

While the Zuck & Trenton/Sholto episodes are doing extremely well on YouTube, what I'm proudest of is that most of these views are actually from Sarah Paine content!

She is one of the greatest living historians, but her work wasn't really publicly well known…

Last 28 days 🤯 While the Zuck & Trenton/Sholto episodes are doing extremely well on YouTube, what I'm proudest of is that most of these views are actually from Sarah Paine content! She is one of the greatest living historians, but her work wasn't really publicly well known…
account_circle
John Coogan(@johncoogan) 's Twitter Profile Photo

Pretty remarkable that the only press Zuck shared from the Llama 3 launch was Dwarkesh Patel and Roberto Nickson

Meta did give embargoed comments to mainstream publishers, but only the independent interviewers got posted to his story.

GOING DIRECT! Lulu Cheng Meservey

Pretty remarkable that the only press Zuck shared from the Llama 3 launch was @dwarkesh_sp and @rpnickson Meta did give embargoed comments to mainstream publishers, but only the independent interviewers got posted to his story. GOING DIRECT! @lulumeservey
account_circle
Dwarkesh Patel(@dwarkesh_sp) 's Twitter Profile Photo

Every time I fail a spaced repetition review for a card which I remember thinking was almost too trivial to write down, I become more convinced that everything I read without making cards for is a waste of time.

account_circle