BEGIN:VCALENDAR
VERSION:2.0
PRODID:-//Talks.cam//talks.cam.ac.uk//
X-WR-CALNAME:Talks.cam
BEGIN:VEVENT
SUMMARY:Large language models and human cognition - Charles Elkan (Compute
 r Science and Engineering\, UC San Diego)
DTSTART:20241129T100000Z
DTEND:20241129T113000Z
UID:TALK224626@talks.cam.ac.uk
CONTACT:Hanni Sondermann
DESCRIPTION:Large language models (LLMs) have remarkable capabilities\, bu
 t they make mistakes that seem surprisingly elementary. For example\, one 
 of today's best LLMs says "A word like delicious but with one letter diffe
 rent is precious" and other state-of-the-art LLMs make similar mistakes. I
  will argue that LLM failures are understandable given the fundamental tec
 hniques on which LLMs are built\, namely tokenization\, transformers\, pre
 training\, and alignment. I will then discuss how the failure modes of LLM
 s are similar to known failure modes of human thought\, and how each of th
 e four central techniques has an analog in human cognition and learning. M
 oving on to multimodal models and the latest agentic and chain-of-thought 
 models\, I will discuss how similar analysis applies to them also. The dis
 cussion will lead to three conclusions: human cognition is based on proces
 sing similar to that done by LLMs\; human-level artificial intelligence is
  hence in reach\; but superintelligence\, for common meanings of that word
 \, is not on the horizon.\n\n*Charles Elkan* is currently an affiliate pro
 fessor of computer science at the University of California\, San Diego\, w
 here for many years previously he was a tenured full professor. In recent 
 years he has worked in New York as the co-founder of Ficc.ai (https://www.
 ficc.ai)\, as a venture partner at Fellows Fund (https://www.fellowsfundvc
 .com)\, and as a consultant. Until 2020 he was a managing director and the
  global head of machine learning at Goldman Sachs\, while from 2014 to 201
 8 he was the first Amazon Fellow\, leading scientists and engineers in Sea
 ttle\, Palo Alto\, and New York doing research and development in machine 
 learning for both e-commerce and cloud computing. He earned his Ph.D. at C
 ornell University and his undergraduate degree in mathematics at Cambridge
 . His students have gone on to professorships at universities that include
  Columbia and Carnegie Mellon\, and to central roles in industry\, includi
 ng as team lead for multiple generations of ChatGPT and for AI in Google s
 earch .\n\nLink to join virtually: https://cam-ac-uk.zoom.us/j/82340807096
 \n\nPasscode: 012190
LOCATION:Computer Lab\, FW11
END:VEVENT
END:VCALENDAR