Interested in RL and its links to control?

Join us March 24 and 25 at the CWI, Amsterdam, for a great workshop on these topics!

Speakers include: Sean Meyn, Ann Nowe, Bert Kappen, yours truly and more...

https://www.cwi.nl/en/events/cwi-research-semester-programmes/workshop-themes-across-control-and-reinforcement-learning/

We are gearing up for the @ELLISforEurope community event & closing event for the ELISE project here in Helsinki! We'll be livestreaming 3 keynotes (from Bernhard Schölkopf, Neil Lawrence, Yoshua Bengio) and 1 panel (AI, ELLIS and the European Public) starting Thursday morning June 27 (6:00 UTC). Watch on YouTube: https://www.youtube.com/playlist?list=PL6SVf9IEsQ7p70wgcg1MfSwgbTV3fbRzk

What a response! More than 3,000 candidates from all over the world registered on our #ELLISPhD portal this year – more than in any year before! We are now reviewing every application in detail and will inform shortlisted candidates in January. Stay tuned!

Learn more about the ELLIS PhD Program here: https://ellis.eu/phd-postdoc

#machinelearning #phd #AI

The ELLIS Doctoral Symposium is the annual highlight among the many networking and training events for #ELLISPhD students and will take place in Paris next year! Apply to our program by November 15 and join our international #PhD community!

Call for applications 2023: https://ellis.eu/news/ellis-phd-program-call-for-applications-2023

#AI #artificialintelligence #ML #machinelearning #PhDProgram

Congratulations! ELLIS President Bernhard Schölkopf receives the 2022 ACM - AAAI Allen Newell Award: https://awards.acm.org/newell

The very first citation in this stupid letter, https://futureoflife.org/open-letter/pause-giant-ai-experiments/, is to our #StochasticParrots Paper,

"AI systems with human-competitive intelligence can pose profound risks to society and humanity, as shown by extensive research[1]"

EXCEPT

that one of the main points we make in the paper is that one of the biggest harms of large language models, is caused by CLAIMING that LLMs have "human-competitive intelligence."

They basically say the opposite of what we say and cite our paper?

This is a huge milestone for #ELLISforEurope! Our network has just welcomed the 1000th member! 🎉 This means: 1000 researchers are now working with us towards the same goal: securing Europe’s sovereignty in modern #AI! Thank you all! #JoinELLISforEurope https://ellis.eu/members

Our #ELLISPhD Program will be featured as one of the case studies for PhD AI Education in Europe during the AIDA AI Education online event on March 9 (10:00 CET). Register now and ask your questions about our program during the event! Other topics on the agenda include risks and opportunities of ChatGPT for education, and factors that make a good PhD program in #AI. https://us02web.zoom.us/webinar/register/WN__fku2aYbQWaxQTrwACZolw

At the City of Amsterdam, we've been using #AI to assess, analyse, and hopefully improve pedestrian #accessibility. My colleagues Iva Gornishka and Shayla Jansen explain how we've been doing so so far, and what our current interns from the Universiteit van Amsterdam and Universiteit Maastricht - Andrea Lombardo, Alisa Todorova, and Aditi Mishra - are currently working on.

https://amsterdamintelligence.com/posts/automatically-collecting-pedestrian-accessibility-information

Hee, KLM-mevrouw Rintel, als we die kerosinetax nou eens inzetten om de dienstverlening en overstaps bij internationale treinen te verbeteren (de infra ligt er namelijk al grotendeels), dan kunnen internationale treinen een flink deel van het vliegen binnen Europa overnemen hoor.
https://www.volkskrant.nl/nieuws-achtergrond/klm-baas-marjan-rintel-natuurlijk-kun-je-zeggen-minder-vliegen-maar-daar-gaan-we-spijt-van-krijgen~bbbd66e5/

Revisiting adversarial training for the worst-performing class

Thomas Pethick, Grigorios Chrysos, Volkan Cevher

https://openreview.net/forum?id=wkecshlYxI

#adversarial #classes #training

A few days to go to submit a 1-2 pages document to this #weather and #climate workshop in Kigali. Deadline extended to 6 March.

Details: https://climate-workshop23.github.io/

Excited about new work with Nikhil Vyas and Sham Kakade on formally defining and providing rigorous guarantees on potential copyright infringements by generative models. See the blog https://windowsontheory.org/2023/02/21/provable-copyright-protection-for-generative-models/

As we say in paper, there are many legal and ethical issues in generative models, and we are only focused on one of them. However I believe making a formal definition is important, also to ground future discussions.

a quick high-level intro to LLMs (the discussion was the best part but unfortunately isn't in the video)
https://twitter.com/EthicsInAI/status/1628543640711602179

At TU Delft, we have a vacancy for an Assistant/Associate Professor Causal Reinforcement Learning:

https://tinyurl.com/5am94765

This is a great opportunity to join the ELLIS Unit Delft, and the larger AI ecosystem 😃

Please forward, boost, etc. - thanks!

Attention Young Europeans!

Only 30 days left until the next round of applications.

As an 18-year-old European, you have the unique opportunity to receive a #DiscoverEU travel pass and explore the diversity of Europe by train.

Immerse yourself in local cultures, soak up the rich history, and indulge in delicious cuisine.

Mark your calendar and be prepared to embark on an unforgettable journey.

#ErasmusPlus #EuropeanUnion

I'm happy to share that our paper "Bayesian Structure Scores for Probabilistic Circuits" has been accepted at #aistats23! This is joint work with Yang Yang (MSc @mastodon.world, incoming PhD @leuvenai and Gennaro Gala.

The main contribution of the paper is to propose a new (actually old) way to learn the structure of #ProbabilisticCircuits. We take lessons from classical structure learning in Bayesian networks which optimizes some structure score, where a principled choice is a Bayesian score.

Senior Research Fellowship in #MachineLearning for #Weather or #Climate at #UCL

https://www.ucl.ac.uk/work-at-ucl/search-ucl-jobs/details?nPostingId=3344&nPostingTargetId=6338&id=Q1KFK026203F3VBQBLO8M8M07&LG=UK&mask=ext

For more information, please reach out to me directly.

@AbelTM @avandeursen @nicolaspapernot @timnitGebru

I tend to agree. I don't think that there is a clearly agreed definition of AGI and so firm statements about its merits or dangers should be subject to scrutiny.

That said, I do believe, that there is consensus about the fact that humans more general intellectual capabilities have led to many benefits as a species. So the idea that more general capabilities in our digital assistants would be helpful has a clear basis, I think.

Introductory lecture slides, first practical, and example code now online for a half-day course "Foundations of GPU Computing" (with a machine learning flavor) that I have in preparation. See more here: https://indii.org/gpu-course #GPU #CUDA #course #MachineLearning