2 x PhD studentships in King's Arts & Humanities on responsible AI, in conjunction with Responsible Ai UK: interdisciplinary research looking at technology and people. Funded at UKRI stipend rates paid for three years, with UK Home Student Fees also covered.
kcl.ac.uk/study-legacy/f…

No. Training LLMs on purely factual data STILL WON'T cure them of 'Hallucinations' #SundayHarangue

There is a persistent myth that LLM hallucinations are just a result of them being trained on un-curated and 'non-factual' data, and will go away with high quality/factual data.…

📖 👍🏼Michael Wooldridge
Enjoyed reading it

My dear and long time friend Daniel Dennett passed away today. He was a towering figure in philosophy and in particular in the philosophy of AI. Now we have only memories of him. [[And do keep up to date with your correspondence--I have a 30 day old email from him in my inbox…

My neighborhood in SF, except for my block, is always thick with humanless Waymo vehicles. In tonight's rush hour I saw my first humanless Waymo on a highway, on 101, driving north to the city around the 280 interchange. I first took a ride in autonomous mode in a Waymo (then…

Our upcoming Communications of the ACM piece (with Eric Horvitz Sheila McIlraith Peter Stone) is now available on arXiv! arxiv.org/abs/2404.04750

Current LLM guardrails are flimsy. We need much more robust & deeper fixes.

Which news org well be first to rebrand as artisanal news? Small-batch, even. Certified AI-free and human-written.

📢 The detailed schedule for #AAMAS2024 is now online: aamas2024-conference.auckland.ac.nz/program/progra…

🗓 Don't forget to register by April 11 to benefit from the early bird rates: aamas2024-conference.auckland.ac.nz/registration/

It's safe to auto-delete all email that contains the word 'esteemed', right?

AI-generated sad girl with piano performs the text of the MIT License

Event:
On Wednesday 10 April 6 pm-7.15 pm, Professor Charlotte Deane will give a free talk for the Oxford Schmidt AI in Science Annual Lecture 2024, before discussing the effects AI is having on science with experts.

For details & booking: saiis.web.ox.ac.uk/event/oxford-s…

#compscioxford

What is 'general intelligence'?

'Can do most cognitive tasks'?

Is a baby GI? A 120 year old? You, a minute before you are struck by lightning and die? You, blinded? You, in a coma?

What is 'cognitive'? Driving a racing car? Carpentry? Running a tech business? Care work?

With all the humility in the world, neither leads to a better grasp of what intelligence is. It may be practical, but it does not advance understanding. That is the central point of the criticism of LLMs.

Open asteroid impact…. Fascinating new startup 🙂

openasteroidimpact.org

I've been learning and researching these past few days about AI, history, and underlying concepts. It is absolutely revolutionary. This amazing lec by Michael Wooldridge at Royal Institution amazing. We're a lucky gen to witness this🚀
youtu.be/b76gsOSkHB4?si…
#AI #ChatGpt #100DaysOfCode

Can AI ace law exams?

Last year, OpenAI announced GPT-4 got 90th percentile on the bar.

Here we (a) refute 90th percentile claim; (b) replicate/extend recent work on GPT capabilities; (c) discuss implications for law profession.

Now open, AI&Law: link.springer.com/article/10.100…
1/

Rough draft of a paper “A Human Right to a Human Decision” which was distributed at a workshop in June 2023 held in Oxford. New and improved version coming soon.

academia.edu/116775511/A_Hu…

I liked my talk about citizens' assemblies so much that it took me three years to switch to a different topic. But the last iteration — an ECAI'23 keynote, which I consider to be the definitive version — is now online, courtesy of the PC chair Kobi Gal's AI & DS Lab.

youtube.com/watch?v=ji-uVx…

10 years ago we published a paper estimating the exposure of jobs to automation. In this paper, reassess what AI can and cannot do in the era of Gen AI. Finally out in the The Brown Journal of World Affairs Michael A Osborne

bjwa.brown.edu/30-1/generativ…