Top Qs
Timeline
Chat
Perspective
Jeff Dean
American computer scientist and software engineer From Wikipedia, the free encyclopedia
Remove ads
Jeffrey Adgate Dean (born July 23, 1968) is an American computer scientist and software engineer. Since 2018, he has been the lead of Google AI.[1] He was appointed Google's chief scientist in 2023 after the merger of DeepMind and Google Brain into Google DeepMind.[2]
Remove ads
Early life and education
Jeff Dean was born to a family of a tropical disease researcher and a medical anthropologist, who moved the family frequently. However, from the fifth to tenth grade, Dean attended schools in the Twin Cities.[3]
Dean received a B.S., summa cum laude, from the University of Minnesota in computer science and economics in 1990.[4] His undergraduate thesis was on neural networks in C programming, advised by Vipin Kumar.[5][6] During freshman year, Dean met his future wife, Heidi Hopper, who graduated from the same university in 1990.[3]
He received a Ph.D. in computer science from the University of Washington in 1996, working under Craig Chambers on compilers[7] and whole-program optimization techniques for object-oriented programming languages.[8]
Remove ads
Career
Summarize
Perspective
Early career
Before graduate school, Dean worked at the World Health Organization's Global Programme on AIDS, developing software for statistical modeling and forecasting of the HIV/AIDS pandemic.[9]
After graduate school, Dean worked at DEC/Compaq's Western Research Laboratory,[10] on profiling tools, microprocessor architecture and information retrieval.[9] Much of his work was completed in close collaboration with Sanjay Ghemawat.[10][7] In early 1999 he briefly worked at comparison-shopping startup mySimon as a senior member of technical staff, designing and implementing a distributed system for web crawling, caching, and full-text indexing of product information to improve the scalability of its shopping search service, before leaving later that year to join Google.
Career at Google
Dean joined Google in mid-1999.[4] He was Google's 30th employee.[11] As of 2018, Dean and Sanjay Ghemawat are the only two employees at Google to hold the title of Senior Fellow, the highest technical level at the company.[7]
Systems design
In collaboration with Sanjay Ghemawat, Dean designed and implemented portions of Google's advertising, crawling, indexing and query serving systems, along with components of the distributed computing infrastructure that underlies most of Google's products.[7] He has also worked on improving search quality, statistical machine translation (Google translate) and internal software development tools and has had significant involvement in the engineering hiring process.
Dean and Ghemawat led the original design of Protocol Buffers, an open-source data interchange format that provides language-neutral and platform-neutral mechanisms for serializing structured data.[12] Protocol Buffers are used extensively across Google for remote procedure call (RPC) protocols and for storing structured information in persistent storage systems, offering advantages in efficiency and extensibility compared to alternatives like XML.[13]
The pair also developed MapReduce, a programming model and implementation for processing and generating large datasets that became foundational for applications.[7] MapReduce abstracts away the complexities of distributed computing (including data partitioning, inter-machine communication, and fault handling) allowing programmers without specialized distributed systems knowledge to leverage clusters of thousands of commodity machines for data processing tasks. The system's success at Google subsequently inspired open-source implementations including Apache Hadoop, which democratized large-scale data processing.
Bigtable, co-designed by Dean and Ghemawat alongside other Google engineers including Fay Chang and Mike Burrows, represents a large-scale semi-structured storage system designed to handle the petabyte-scale data requirements of Google's products.[7] Bigtable is engineered to store sparsely populated tables that scale to billions of rows and thousands of columns while maintaining low-latency, high-throughput read and write operations, and it integrates seamlessly with MapReduce for batch processing.[14] The system's architecture and design influenced the NoSQL database movement, with multiple open-source projects adopting similar approaches.
Spanner represents Dean and Ghemawat's contribution to globally distributed database systems, delivering a horizontally scalable, multi-version database that can replicate data synchronously across geographically distributed datacenters while maintaining strong consistency guarantees.[15] Spanner's architecture enables applications to specify fine-grained replication constraints, control data locality to minimize latency, and transparently migrate data between datacenters for load balancing, supporting SQL-based queries and ACID transactions across potentially millions of machines and trillions of rows.
Dean and Ghemawat developed LevelDB, an open-source on-disk key-value store released in 2011 that drew inspiration from Bigtable's design principles while minimizing external dependencies to enable broader adoption.[16] LevelDB stores sorted key-value pairs, supports atomic batch writes, forward and backward iteration, and compression via Google's Snappy library, subsequently becoming the backend storage engine for Google Chrome's IndexedDB as well as other applications including Bitcoin Core and Minecraft Bedrock Edition.[17]
Artificial intelligence
Dean joined Google X in 2011 to investigate deep neural networks, which had just resurged in popularity. This ended with "the cat neuron paper", a deep belief network trained by unsupervised learning on YouTube videos.[18] This project morphed into Google Brain, a team that studies large-scale artificial neural networks, also formed in 2011.[7] Jeff Dean became its leader in 2012.
Dean worked on DistBelief, a proprietary machine-learning system for distributed training of deep neural networks, whose name reflects the possibility to train deep belief networks. DistBelief was eventually refactored into TensorFlow, an open-source machine-learning software library, for which he was among the designers and implementers of the initial release.[13][7] The DistBelief system was used to train the network in "the cat neuron paper".[18][19] TensorFlow initially dominated the machine-learning research landscape, but by 2023 accounted for only 8% of exclusive models on Hugging Face compared to 92% for PyTorch, which was also used in 80% of academic papers utilizing deep learning frameworks.[20]
Dean also worked on Pathways, an asynchronous distributed dataflow system for neural networks.[21][13] It was used in PaLM.[22]
In April 2018, Dean was appointed the head of Google's artificial intelligence division, after John Giannandrea left to lead Apple's AI projects.[23][24]
In December 2020, Google’s Ethical AI team co-lead Timnit Gebru was terminated after a draft paper she co-authored on the risks and ethical implications of large language models didn't receive publication approval. The incident received significant attention in the media, and in the aftermath, Dean addressed an email to staff acknowledging that the episode had “surfaced large, important issues” around research culture, bias, and inclusion within Google’s AI organization.[25][26][27][28] In an open letter, critics demanded that members of Google's senior leadership, including Dean, "explain the process by which the paper was unilaterally rejected by leadership."[29] Later, Margaret Mitchell, then co-lead of the Ethical AI team and an outspoken supporter of Gebru, was fired on February 19, 2021 for "reportedly using automated scripts to find examples of mistreatment of Dr. Timnit Gebru."[30] On the same day Mitchell was fired, Dean issued an internal memo acknowledging Google "could have and should have handled this situation with more sensitivity."[31]
In 2023, DeepMind was merged with Google Brain to form a unified AI research unit, Google DeepMind, headed by Demis Hassabis. As part of this reorganization, Dean became Google's chief scientist.[2][13] Dean proposed the name Gemini for the chatbot developed by Google DeepMind, "because it's like twins coming together".[11]
Remove ads
Philanthropy
Dean and his wife, Heidi Hopper, started the Hopper-Dean Foundation and began making philanthropic grants in 2011. In 2016, the foundation gave $2 million each to UC Berkeley, Massachusetts Institute of Technology, University of Washington, Stanford University and Carnegie Mellon University to support programs that promote diversity in science, technology, engineering and mathematics (STEM).[32] The Menlo Park-based foundation gave $22.1 million to a variety of universities and non-profits in 2023 and ended the year with $54.4 million in assets, according to its Form 990.[33]
In 2025, Dean joined the board of Laude Institute, steering the organization with David Patterson, Joelle Pineau, and Andy Konwinski.[34]
Awards and honors
- ACM SIGOPS Mark Weiser Award (2007)[35]
- Elected to the National Academy of Engineering (2009)[36]
- Fellow of the Association for Computing Machinery (2009)[37]
- ACM Prize in Computing (2012)[37]
- Mark Weiser Award[38]
- Fellow of the American Academy of Arts and Sciences (2016)[39]
- Recipient of the IEEE John von Neumann Medal (2021)[40]
Publications
Selected papers
- Jeffrey Dean and Sanjay Ghemawat. 2004. MapReduce: Simplified Data Processing on Large Clusters. OSDI'04: Sixth Symposium on Operating System Design and Implementation (December 2004)
- Fay Chang, Jeff Dean, Sanjay Ghemawat, Wilson C. Hsieh, Deborah A. Wallach, Mike Burrows, Tushar Chandra, Andrew Fikes, and Robert E. Gruber. 2006. Bigtable: A Distributed Storage System for Structured Data. OSDI'06: 7th Symposium on Operating System Design and Implementation (October 2006)
Remove ads
Personal life
Dean is married and has two daughters.[7]
He is the subject of an Internet meme for "Jeff Dean facts". Similar to Chuck Norris facts, the Jeff Dean facts exaggerate his programming powers.[41] For example:[42]
Once, in early 2002, when the index servers went down, Jeff Dean answered user queries manually for two hours. Evals showed a quality improvement of 5 points.
See also
References
External links
Wikiwand - on
Seamless Wikipedia browsing. On steroids.
Remove ads