Adam Gleave co-founded FAR.AI and leads its research on making advanced AI safe and reliable. His work spans reinforcement learning, reward models, and adversarial robustness, including research at DeepMind and doctoral work at UC Berkeley.
Adam Gleave co-founded FAR.AI and leads its research on making advanced AI safe and reliable. His work spans reinforcement learning, reward models, and adversarial robustness, including research at DeepMind and doctoral work at UC Berkeley.
Career
Co-founder & CEO at FAR.AI from 2022-01 to PresentLeads the AI safety institute’s research agenda and technical team.SABoard Member at Safe AI Forum until Present
LIBoard Member at London Initiative for Safe AI until Present
MBoard Member at METR until Present
LFFund Manager at Long-Term Future Fund from 2020-02 to 2022-01Worked with fellow fund managers on grants to charities addressing humanity’s long-term future.
Research Intern at DeepMind from 2021-01 to 2021-05Researched uncertainty in language reward models with Geoffrey Irving; previously interned with Jan Leike in 2019.GCJunior Researcher at GSA Capital from 2016-10 to 2017-08Developed a futures trading strategy.
JSTrading Intern at Jane Street Capital from 2015-06 to 2015-09Researched commodity trading and fixed-income models; previously a developer intern in 2014.
RPSummer Intern at Raspberry Pi from 2013-06 to 2013-08Wrote C and Python software for a low-cost weather station.
IMathematics Intern at i2OWater from 2012-08 to 2012-08Modeled pressure loss in water networks.