Ian Goodfellow | |
|---|---|
 Goodfellow, 2017 | |
| Born | 1987 (age 37–38)[1] |
| Nationality | American |
| Alma mater | Stanford University (BS, MS) Université de Montréal (PhD) |
| Known for | Generative adversarial networks,Adversarial examples |
| Scientific career | |
| Fields | Computer science |
| Institutions | Apple Inc. Google Brain OpenAI DeepMind Google DeepMind |
| Thesis | Deep Learning of Representations and its Application to Computer Vision (2014) |
| Doctoral advisor | Yoshua Bengio Aaron Courville |
| Website | www |
Ian J. Goodfellow (born 1987[1]) is an Americancomputer scientist,engineer, andexecutive, most noted for his work onartificial neural networks anddeep learning. He is a research scientist atGoogle DeepMind,[2] was previously employed as a research scientist atGoogle Brain and director of machine learning atApple as well as one of the first employees atOpenAI, and has made several important contributions to the field ofdeep learning, including the invention of thegenerative adversarial network (GAN). Goodfellow co-wrote, as the first author, the textbookDeep Learning (2016)[3] and wrote the chapter on deep learning in the authoritative textbook of the field of artificial intelligence,Artificial Intelligence: A Modern Approach[4][5] (used in more than 1,500 universities in 135 countries).[6]
Goodfellow obtained hisB.S. andM.S. in computer science fromStanford University under the supervision ofAndrew Ng (co-founder and head ofGoogle Brain),[citation needed] and his Ph.D. in machine learning from theUniversité de Montréal in February 2015, under the supervision ofYoshua Bengio and Aaron Courville.[7][8] Goodfellow's thesis is titledDeep learning of representations and its application to computer vision.[7][9]
After graduation, Goodfellow joinedGoogle as part of theGoogle Brain research team.[10] In March 2016, he left Google to join the newly foundedOpenAI research laboratory.[11] Barely 11 months later, in March 2017, Goodfellow returned to Google Research[12] but left again in 2019.[13]
In 2019, Goodfellow joinedApple as director of machine learning in the Special Projects Group.[13] He resigned from Apple in April 2022 to protest Apple's plan to require in-person work for its employees.[14] Shortly after, Goodfellow then joinedGoogle DeepMind as a research scientist.[2][15][16]
Goodfellow is best known for inventinggenerative adversarial networks (GAN), using deep learning to generate images. This approach uses two neural networks to competitively improve an image's quality. A “generator” network creates a synthetic image based on an initial set of images such as a collection of faces. A “discriminator” network tries to detect whether or not the generator's output is real or fake. Then the generate-detect cycle is repeated. For each iteration, the generator and the discriminator use the other's feedback to improve or detect the generated images, until the discriminator can no longer distinguish between the fakes generated by its opponent and the real thing. The ability to create high quality generated imagery has increased rapidly. Unfortunately, so has its malicious use, to createdeepfakes and generate video-baseddisinformation.[17][18]
At Google, Goodfellow developed a system enablingGoogle Maps to automatically transcribe addresses from photos taken byStreet View cars[19][20] and demonstrated security vulnerabilities of machine learning systems.[21][22]
In 2017, Goodfellow was cited inMIT Technology Review's 35 Innovators Under 35.[23] In 2019, he was included inForeign Policy's list of 100 Global Thinkers.[24]
