AIhub monthly digest: December 2021 – #NeurIPS2021, sustainable cities and the Reith lectures

by Lucy Smith

20 December 2021

NeurIPS 2021

One of the main events in the AI world this month was the 35th conference on Neural Information Processing Systems (NeurIPS2021). We were lucky enough to attend and only scratched the surface of the vast array of talks, panels, workshops and posters on offer.

You can read the first of our round-ups of the invited talks, #NeurIPS2021 invited talks round-up: part one – Duolingo, the banality of scale and estimating the mean. Our articles on the rest of the talks will follow over the next few weeks.

You can find out what attendees, organisers and presenters were chatting about during the conference in our two Twitter summaries:
#NeurIPS2021 in tweets – highlights from the first two days
#NeurIPS2021 in tweets – highlights from the first week

Before the event began, the organisers revealed the winners of the prestigious outstanding paper awards, and the test of time award. This year, there were also two best paper awards for the new datasets and benchmarks track. Find out who won what here.

Sustainable cities and communities

This month saw the launch of the latest topic in our focus on the UN sustainable development goals. During December and January you’ll be able to find articles about sustainable cities and communities.

In this blog post Paul Patras writes about Harnessing the power of AI to elucidate mobile traffic consumption at city scale. Paul and his team use AI methods and aggregate data to produce high-resolution traffic maps.

In a video lecture in this post, Beril Sirmacek talks about research using AI methods to help to make cities more environmentally friendly.

CoRL awards

Last month we brought you an interview with Tao Chen, Jie Xu and Pulkit Agrawal, winners of the CoRL 2021 best paper award.

This month, we heard from the winners of the best system paper award, Huy Ha and Shuran Song. They told us about their work on a self-supervised framework for cloth unfolding which uses a pick, stretch, and fling primitive for a dual-arm setup from visual observations.

In this blog post, Robot reinforcement learning: safety in real-world applications, Puze Liu, Davide Tateo, Haitham Bou-Ammar and Jan Peters write about their research into making robots learn in the real world while ensuring safety. Their work was shortlisted for the CoRL 2021 best paper award.

Timnit Gebru on truly ethical AI

In an article in The Guardian newspaper – For truly ethical AI, its research must be independent from big tech – Timnit Gebru writes that two keys things we can do to safeguard us from unsafe uses of AI are 1) to curb the power of the companies who develop it, and 2) increase the power of those who speak up against the harms of AI and these companies’ practices.

The BBC Reith Lectures

This year, the prestigious BBC Reith Lectures focus on AI. In a series of four talks, Stuart Russell asks how artificial intelligence could transform the world, whether fears of AI are well founded, and if we can make it work for us.
Lecture 1: The biggest event in human history.
Lecture 2: AI in warfare.
Lecture 3: AI in the economy.
Lecture 4: Coming on 22 December, “AI: A future for humans”.

A new journal for the community

In this blog post “Announcing the Transactions on Machine Learning Research“, Hugo Larochelle unveiled plans to launch a new journal for the machine learning community. The review process of Transactions on Machine Learning Research will be double blind, with the referee comments and responses open for all to view.

Our resources page
Forthcoming and past seminars