talkachman.bsky.social

talkachman.bsky.social

@talkachman.bsky.social

Ass Prof at Radboud, member of Doners Institute PI of the CATALYST Lab: www.cerebrnita.com Founder, Exiter, Angel investor Ex : MIT, Technion, IBM, AQR, Optiver All opinions are personal talkachman.com Tamer of transformers, annoyance of bureaucrats

108 followers163 following291 posts

talkachman.bsky.social
@talkachman.bsky.social

OPT is not like h1b. It's one of the strongest mechanisms we have to keep good people. It's one of the strongest mechanisms we have for people to do their internships it's an economic engine........

Melanie Mitchell@melaniemitchell.bsky.social

RIP OPT? What an unmitigated disaster for everyone.

talkachman.bsky.social
@talkachman.bsky.social

A bit of sanity finally ...

Melanie Mitchell@melaniemitchell.bsky.social

Hey @ericlipton.nytimes.com 👋 AI is not “evolving” on its own. *People* are choosing how to build it. *People* are choosing how to train it. *People* are choosing how to run it. Maybe the problem is that they are the *wrong people* to make these choices.

talkachman.bsky.social
@talkachman.bsky.social

I am always baffled and feel really uncomfortable with these kind of statements. It's basically completely demolishing the modern perspective of science. Long thread and rant incoming 🤌🧵👇

Ethan Mollick@emollick.bsky.social

It is less than a decade since the development of the transformer. Less than four years since the release of GPT-3.5 (ChatGPT). Less than two years since the release of o1-preview (the first Reasoner).

View full thread
talkachman.bsky.social
@talkachman.bsky.social

More and more, especially in CS I see that we have completely lost track of what led us to our current state which is completely shutting down out innovation. Closing the loop, it's only been a decade since transformer is discarding the 20+ years of everything that led up to it.

talkachman.bsky.social
@talkachman.bsky.social

I am always baffled and feel really uncomfortable with these kind of statements. It's basically completely demolishing the modern perspective of science. Long thread and rant incoming 🤌🧵👇

Ethan Mollick@emollick.bsky.social

It is less than a decade since the development of the transformer. Less than four years since the release of GPT-3.5 (ChatGPT). Less than two years since the release of o1-preview (the first Reasoner).

View full thread
talkachman.bsky.social
@talkachman.bsky.social

This is also why I always tell my grad students that negative results are as important if not more than, positive ones cause they are all setting up the stepping stones for us to be closer to the next breakthrough, the next discovery, the next frontier. Which sadly in today's landscape of ....

talkachman.bsky.social
@talkachman.bsky.social

Grad student descent is being lost, that is horrendous from an educational perspective for our next generation of scientists. Perspective BTW, being the important point. When one approaches science in the above manner the perspective of where we are going where we went and what led us there is key.

talkachman.bsky.social
@talkachman.bsky.social

I am always baffled and feel really uncomfortable with these kind of statements. It's basically completely demolishing the modern perspective of science. Long thread and rant incoming 🤌🧵👇

Ethan Mollick@emollick.bsky.social

It is less than a decade since the development of the transformer. Less than four years since the release of GPT-3.5 (ChatGPT). Less than two years since the release of o1-preview (the first Reasoner).

talkachman.bsky.social
@talkachman.bsky.social

Very early in my career, especially as a PI, I was asked by one of my first students, Why are you doing this, and what's science to you? It took me a minute to formulate my reply, which was a combination of both. In my mind, doing science is laying the stepping stones for the next discovery

talkachman.bsky.social
@talkachman.bsky.social

I love playing online chess; my go-to has been, and recently I am really enjoying. The problem is that at certain exact times of the day I am starting to be paried with players that actually go from a start game that is ok but then mid game that is 96% correlated with an engine

talkachman.bsky.social
@talkachman.bsky.social

Wow another huge company waiting to buy #datasets! Where have I seen this before www.bbc.com/news/busines...

bbc.comAmazon to buy Whole Foods for $13.7bnThe deal marks the biggest push yet by Amazon into traditional bricks and mortar retailing.
Jon Keegan@jonkeegan.com

Wow www.theinformation.com/articles/nvi...

theinformation.comNvidia Agrees to Buy Open Source Model Repository Hugging Face For $12.9 BillionNvidia has agreed to buy Hugging Face, a company known for its GitHub-like repository of open-source AI models, for $12.9 billion, according to a person with knowledge of the agreement. The move will ...
talkachman.bsky.social
@talkachman.bsky.social

Can confirm max is an awesome person to work with. Bonus points, I am right down the corridor as well!

Max Hinne@maxhinne.bsky.social

Come work as a postdoc at the Donders Institute at the Radboud University, with prof. Sabine Hunnius @sabinehunnius.bsky.social and myself, on a model-based approach for understanding cognitive development in early childhood! More info here: www.ru.nl/en/working-a...

ru.nlPostdoc Position: A Model-Based Approach to Understanding Cognitive Development in Early Childhood | Radboud UniversityDo you want to work as a Postdoc at the Donders Centre for Cognition? Check our vacancy!