In 2022, artificial intelligence will continue to develop in fields such as AIGC and AI for science, and some work has also explored knowledge-based models. In addition to these research results, we will also find that the work situation of artificial intelligence practitioners has also changed significantly this year.
For example, large technology companies such as Meta have conducted large-scale layoffs, causing many AI practitioners to consider new employment directions. On the other hand, many AI experts have chosen to leave large technology companies and join AI entrepreneurship.
According to a survey by Business Insider, in the past year, 26 AI experts have resigned from Alphabet alone, including the two major technology giants Google and DeepMind. Several experts left the company to start new companies and collectively raised hundreds of millions of dollars to build the next generation of AI tools.
Whether it is an AIGC model such as DALL-E 2 or a large-scale language model such as ChatGPT, they all require strong support in terms of technology and resources. In terms of resources, companies such as Google, Meta, Hugging Face, and Stability AI will provide powerful training infrastructure; in terms of technical support, many of the current major advances in the AI field can be traced back to a paper published by Google researchers in 2017. A seminal paper - "Attention is all you need".
#The Transformer architecture proposed in this paper became the basis for many subsequent AI models. Currently, this 5-year-old paper has been cited more than 60,000 times. Now the institutions where these authors work have changed a lot:
# So it shouldn’t be a surprise that Google is a source of talent for machine learning professionals. strange. In particular, six of the Transformer paper authors have launched their own startups, one has joined OpenAI, and only one is still working at Google.
Ashish Vaswani and Niki Parmar founded Adept AI in April 2022 and received $65 million in financing from Addition and Greylock. The startup is focused on creating a robot based on the same core technology as text generation tools. The CEO, CTO and chief scientist are all from Google, and it has also gathered a group of technical talents who have worked at Google:
David Luan once told his motivations for leaving Google to start a business:
At Google, we have trained more and more talented people. Transformer, dreams of one day building a universal model to support all ML use cases. However, there is an obvious limitation: models trained on text can write great prose, but they cannot take action in the digital world. You can’t ask GPT-3 to book a flight for you, write a check to a supplier, or conduct a scientific experiment.
Also there is Inflection AI, a privately held company in the general modeling space that raised $225 million in May 2022. Mustafa Suleyman is the CEO and co-founder of Inflection AI.
Mustafa Suleyman
Mustafa Suleyman is a founding member of the DeepMind Artificial Intelligence Lab and became Google’s Vice President of Artificial Intelligence Products and Policy in 2020, six years after Google acquired DeepMind. At least four Google AI employees have joined Inflexion AI this year:
Character.ai is another startup founded by the author of a seminal article on artificial intelligence. The company's product provides users with a way to talk to bots. This company was officially established in October 2021. The outside world does not know much about the company’s organizational structure and business model.
Noam Shazeer, the founder and CEO of Character.AI, is a former chief software engineer at Google and one of the authors of the Transformer paper. He joined Google in late 2000 and was one of Google's most important early employees until his final departure in 2021.
Four more Google AI technical talents have joined Character.AI:
In addition, several experts from DeepMind, another subsidiary of Google’s parent company Alphabet, have left to join startups. Foreign media The Next Web called DeepMind "a startup factory with brain drain."
EquiLibre Technologies
#In January this year, Martin Schmid, Matej Moravcik and Rudolf Kadlec left DeepMind and founded Focus EquiLibre Technologies, an AI startup in the stock market field. All three are developers of DeepMind Poker AI DeepStack.
Martin Schmid is a former senior researcher at DeepMind, one of the authors of DeepStack and Player of Games, and is now CEO and co-founder of EquiLibre Technologies.
Rudolf Kadlec is a former senior researcher at DeepMind and co-author of Player of Games. He is currently the CTO and co-founder of EquiLibre Technologies.
Jack Rae
Former DeepMind senior researcher, now an engineer at OpenAI.
Eric Jang
##Former senior researcher at Google, now at the robotics company Halodi Robotics Vice President of AI.
Tatiana Shpeisman
Former senior engineering manager at Google, worked on CPU and GPU compilers, now Director of Compiler Engineering at Modular.
Maithra Raghu
Former senior researcher at Google, now co-founder and CEO of Samaya AI.
Alex Hanna
Former senior researcher in Google’s AI ethics department, he is now the research director of the Distributed AI Research Institute (DAIR), a startup founded by Timnit Gebru, a female researcher who was previously fired from Google.
Barret Zoph
Former Google Brain researcher, now an engineer at OpenAI.
Sara Hooker
Former Google Brain researcher now leads the nonprofit machine learning research lab Cohere For AI . It is worth noting that Aidan Gomez, one of the authors of the "Attention is all you need" paper, is the co-founder and CEO of Cohere For AI, and he also interned at Google Brain.
From a personal point of view, it is very promising for these outstanding talents to leave their jobs and start businesses; from the perspective of the entire industry, the rapid growth of many start-up companies will also promote the innovation and development of the industry and enrich the industry. Domain diversity. Perhaps this is a brain drain for technology giants such as Google and DeepMind, but they have served as a cradle of talent for the development of the entire field.
The above is the detailed content of In 2022, 26 Google AI experts will leave: most of them will start their own businesses overseas, with a maximum financing of 1.5 billion. For more information, please follow other related articles on the PHP Chinese website!