Jump to content

Adam Tauman Kalai

From Wikipedia, the free encyclopedia
Adam Tauman Kalai
NationalityAmerican
Alma materHarvard University
Carnegie Mellon University
Scientific career
FieldsComputer Science, Artificial Intelligence
InstitutionsToyota Technological Institute at Chicago
Georgia Tech
Microsoft Research
OpenAI
Doctoral advisorAvrim Blum

Adam Tauman Kalai is an American computer scientist who specializes in Machine Learning and recently moved to OpenAI[1][2] after being a Senior Principal Researcher at Microsoft Research New England[3][4].

Education and career

[edit]

Kalai graduated from Harvard University in 1996 and received a PhD from Carnegie Mellon University in 2001, where he worked under doctoral advisor Avrim Blum. He did his postdoctoral study at the Massachusetts Institute of Technology before becoming a faculty member at the Toyota Technological Institute at Chicago and then the Georgia Institute of Technology. He joined Microsoft Research in 2008[4] and subsequently moved to OpenAI in 2023.[1][2]

Contributions

[edit]

Kalai is known for his algorithm for generating random factored numbers (see Bach's algorithm), for efficiently learning learning mixtures of Gaussians, for the Blum-Kalai-Wasserman algorithm for learning parity with noise, and for the intractability of the folk theorem in game theory.

More recently, Kalai is known for identifying and reducing gender bias in word embeddings, which are a representation of words commonly used in AI systems[3][5] and for his work on Hallucinations in large language models.[1]

References

[edit]
  1. ^ a b c Levy, Steven (January 5, 2024), In Defense of AI Hallucinations, retrieved 2024-03-19
  2. ^ a b Adam Tauman Kalai, retrieved 2024-03-19
  3. ^ a b Pinkerton, Byrd (August 12, 2016), He's Brilliant, She's Lovely: Teaching Computers To Be Less Sexist, National Public Radio (NPR), retrieved 2019-01-28
  4. ^ a b Artificial Intelligence and Statistics Conference, 2016, retrieved 2019-01-28
  5. ^ Gholipour, Bahar (March 10, 2017), Algorithms Learn From Us, and We Can Be Better Teachers, NBC, retrieved 2019-09-01
[edit]