Ch E At Gpt Computerphile

Summary

Mike explains a paper from the University of Maryland, proposing a neat trick to 'watermark' the output of large language models ... Why didn't OpenAI release their "Unicorn" GPT2 large transformer? Rob Miles suggests why it might not just be a a PR stunt. Big data research needs high performance computing and fast networks but so do thousands of students watching Netflix. Jisc run ... What was The Imitation Game? It inspired the name for the recent Alan Turing's movie but just what was it? Professor Brailsford ... Share part of a secret without knowing which part? Dr Tim Muller explains how Oblivious Transfer works. A checksum that's robust & trivial to compute with Matt Godbolt -- Brilliant's courses and start for free at ... It's an older paper, but it checks out. Rob Miles discusses the problem of 'Sleeper Agents' - where LLMs could have hidden traits ... This video features Sydney Von Arx --- Brilliant's courses and start for free at brilliant.org/ The danger of assuming general artificial intelligence will be the same as human intelligence. Rob Miles explains with a simple ... How to we check to see if a black box system is giving us the right result for the right reason? Even a broken clock is correct twice ... Researchers stumbled upon a simple but worrying bug. Cropped images from Pixel phones contained a great deal of the original ... Why it's a bad idea to build a Virtual Private Network using TCP. Dr Steve Bagley on TCP over TCP... How do we measure harm to improve the performance of Ai in the real world? Dr Hana Chockler is a Reader in Computer Science ... A massive topic deserves a massive video. Rob Miles discusses ChatGPT and how it may not be dangerous, yet. More from Rob ... Language Models' Achilles heel: Rob Miles talks about "glitch" tokens, those mysterious words which, which result in gibberish ...