What have the eight ex-Google Brain authors of the Transformers paper been doing since December 2017? 🧐 Let’s find out together! 🧵 Paper: arxiv.org/pdf/1706.03762.pdf Sources: Google, Pitchbook, LinkedIn // @ashVaswani @NoamShazeer @aidangomezzz @nikiparmar09 @YesThisIsLion @kyosu @lukaszkaiser @ilblackdragon

Apr 15, 2023 · 7:57 PM UTC

Ashish Vaswani (@ashVaswani) left Google in late 2021 to start @AdeptAILabs which has raised $415 million so far and is valued by VCs at > $1 billion…looks like the company is growing fast and just crossed 40 people. It also looks like Ashish just left Adept a few months ago. 🧐
After TWO DECADES at Google, Noam Shazeer (@NoamShazeer) left G in late 2021 and founded @character_ai with his friend @dan_defr… they’ve raised nearly $200 million so far and are ALMOST a unicorn... and pretty damn capital efficient so far with only 20 ~ employees. 🦄 😅
Niki Parmar (@nikiparmar09) is the only woman author on the Transformers paper. She left G in late 2021 and co-founded @AdeptAILabs. Niki left Adept a few months ago as well. 😎
Jakob Uszkoreit (@kyosu) — cited by many as the main brain behind discovering / inventing the Transformer architecture — left G in mid 2021 and co-founded @Inceptive_labs.. designing mRNAs with neural nets. 🧬 🤯 They’ve raised $20M so far and just crossed 20 people. 🔥
Llion Jones (@YesThisIsLion) is still at Google. 🤓 He claims to have made the most meaningful contribution to the paper: writing the title. 😜
Aidan Gomez (@aidangomezzz) left Google Brain in September 2019 to found @CohereAI — they’ve raised (looks like) nearly $400 million. Cohere also recently crossed 180+ employees and is clearly scaling after 3 years going strong. 🚀
Lukasz Kaiser (@lukaszkaiser) is one of the co-creators of @TensorFlow and left G in mid 2021 to join @OpenAI.
Illia Polosukhin (@ilblackdragon) left G in Feb 2017 🧐 … he founded @NEARProtocol (competitor to Ethereum / Solana) in June 2017. NEAR is currently valued at $2 billion ~. Looks like they have raised $375 million ~ and done lots of secondary rounds. 😮
They'll be burning through $ pretty quickly since not only is attention not all that we need, it's not even a real process, value or function in terms of cognition. It's a phantom category. Silicon Valley is essentially the Land of Zombie Intuition. pubmed.ncbi.nlm.nih.gov/3148…
I’ll read this paper, the abstract is interesting. Thanks.