Researcher

Larry Yaeger

Larry Yaeger is a former Apple Distinguished Scientist who built Polyworld, an artificial-life ecology in which simulated organisms evolve neural network brains, and who led the neural network handwriting recognizer for Apple's Newton. Trained in aerospace engineering at Purdue, he directed software development at the effects house Digital Productions from 1982 to 1987, then joined Alan Kay's Vivarium Program at Apple in 1987. Polyworld debuted at the Artificial Life III conference in 1992, and his Print Recognizer, described in a 1996 NIPS paper, shipped in second-generation Newton PDAs and later in Mac OS X as Inkwell. In 2004 he became a professor of informatics at Indiana University, where he studied the evolution of neural complexity, and by 2015 he was doing applied machine learning at Google. He is now retired and lives in Bloomington, Indiana.

Now at
Google
Born
1950

Years link to their sources.

Timeline

  1. 1950–1972

    Aerospace engineering at Grumman

    Born in 1950, he studies aerospace engineering at Purdue University (B.S.) and the Polytechnic Institute of New York (M.S.). At Grumman Aerospace, he co-develops the Quick three-dimensional geometry modeling system in 1972 and runs some of the first fully three-dimensional hypersonic flow studies over the space shuttle.

  2. 1982–1987

    Effects software at Digital Productions

    From 1982 to 1987, he is Director of Software Development at the effects house Digital Productions, where his work on the DP3D rendering software leads to the photo-realistic effects of The Last Starfighter. He also creates the planet Jupiter for 2010 and the flying-owl title sequence of Labyrinth, which wins the NCGA Best Computer Animation award of 1986.

  3. 1987–1992

    Polyworld in Apple's Vivarium

    In 1987, he joins Alan Kay's Vivarium Program at Apple Computer, where he extends Terry Sejnowski's NETtalk speech network and designs a computer voice for Koko the gorilla. There he creates PolyWorld, an artificial-life ecology whose organisms evolve neural architectures through mutation, recombination and sexual reproduction, and presents it at the Artificial Life III conference in Santa Fe in 1992. He frames it as an attempt to approach artificial intelligence the way natural intelligence emerged, through the evolution of neural systems in a complex ecology.

  4. 1996–1998

    The Newton Print Recognizer

    In Apple's Advanced Technology Group, he is principal author and technical lead of a neural network handwriting recognizer that ships as the Print Recognizer in second-generation Newton PDAs. With Richard Lyon and Brandyn Webb, he describes it in a 1996 NIPS paper and in AI Magazine's Spring 1998 issue. He later ports the recognizer to Mac OS X 10.2 Jaguar, where it becomes Inkwell.

  5. 2004

    Professor at Indiana University

    In 2004, he joins Indiana University's School of Informatics as a professor, arriving from Apple, where he is a Distinguished Scientist. He keeps a one-day-a-week role at Apple on pen, ink and handwriting recognition and becomes a full professor of informatics.

  6. 2004–2011

    Evolving neural complexity in Polyworld

    He releases a revived Polyworld as open source on SourceForge in December 2004. In a 2008 Artificial Life XI paper with Virgil Griffith and Olaf Sporns, Polyworld experiments show that, in some conditions, evolution selects for increases in neural complexity in a driven fashion, and he presents the work at London's Gatsby Computational Neuroscience Unit in July 2008. In spring 2011, he teaches his course Artificial Life as an Approach to Artificial Intelligence.

  7. By 2015

    Machine learning at Google

    By 2015, he has returned to industry to develop and deploy applied machine learning at Google in Mountain View, California, where he writes software and creates data optimization algorithms. In May 2015, he gives a lecture series titled Artificial Life as an Approach to Artificial Intelligence, The Director's Cut.

  8. As of October 2026

    Retirement in Bloomington

    He is retired and lives back in Bloomington, Indiana. Polyworld, the program he wrote to evolve artificial intelligence through natural selection, remains open source on GitHub, with version 2.6 released in February 2015.

Career

  1. From 2015Google
  2. From 2004Indiana UniversityProfessor
  3. 1987–2004ApplePrincipal Author and Technical Lead, Distinguished Scientist
  4. 1982–1987Digital ProductionsDirector of Software Development
  5. Until 1972Grumman Aerospace Corporation

Education

  1. Purdue UniversityB.S. in Aerospace Engineering
  2. Polytechnic Institute of New YorkM.S.

Sources and credits

Photo: We have no published photo of Larry Yaeger yet. Send us a free-licensed photo.

  1. Larry Yaeger Extended Bio shinyverse.org. Checked 20 October 2026.
  2. Larry Yaeger - Wikipedia en.wikipedia.org. Checked 20 October 2026.
  3. Polyworld: Real Life in an Artificial Context? - Ars Electronica Archive webarchive.ars.electronica.art. Checked 20 October 2026.
  4. Apple-Newton Handwriting Recognition shinyverse.org. Checked 20 October 2026.
  5. Apple Newton - Wikipedia en.wikipedia.org. Checked 20 October 2026.
  6. inside the mind of a Techno-maniac - Indiana Daily Student idsnews.com. Checked 20 October 2026.
  7. Polyworld shinyverse.org. Checked 20 October 2026.
  8. [1112.4906] Passive and Driven Trends in the Evolution of Complexity arxiv.org. Checked 20 October 2026.
  9. Gatsby Unit | Research: Larry Yaeger, Trends in the Evolution of Neural Complexity gatsby.ucl.ac.uk. Checked 20 October 2026.
  10. I486/I586/H400 Artificial Life as an approach to Artificial Intelligence shinyverse.org. Checked 20 October 2026.
  11. Artificial Life as an approach to Artificial Intelligence, The Director's Cut - Donghua University News news.dhu.edu.cn. Checked 20 October 2026.
  12. A very brief, very general bio of Larry Yaeger shinyverse.org. Checked 20 October 2026.
  13. Polyworld - Wikipedia en.wikipedia.org. Checked 20 October 2026.