AIhub monthly digest: September 2026 – tracking animal populations, recommender systems, and an interview with Ken Goldberg

Welcome to our monthly digest, where you can catch up with any AIhub stories you may have missed, peruse the latest news, recap recent events, and more. This month, we meet AI pioneer Ken Goldberg, use AI techniques to track animal populations, investigate how to improve recommender systems, and find out how external knowledge is used in AI systems.

Combining cultures, from code to canvas: an interview with Ken Goldberg

In the latest instalment in our AI pioneers series, we spoke to award-winning roboticist, filmmaker, and artist, Ken Goldberg. We discussed the culture clash within robotics, what art and science have to learn from each other, and how AI will shape the future of art.

Understanding animal behaviour and tracking populations

Deep learning is a powerful tool for understanding animal behaviour and tracking populations. We caught up with Isla Duporge to find out about her work developing remote-sensing and AI methods to investigate animal movement in the wild.

Optimising sensor placement for estimating wildlife populations

Hannah Murray is also researching wildlife populations. In this interview, she told us more about work developing optimization methods that help ecologists determine where to place sensors, such as camera traps, to get precise estimates of species population counts from the data they collect.

Improving the process for large-scale recommender systems

In large-scale recommender systems, platforms often need to handle massive volumes of items at web latency. To meet this requirement, systems often employ a two-stage decision process. We heard from Haruka Kiyohara about her research trying to improve the first part of this decision process – an early-stage ranker, which filters up to billions of items cheaply into a smaller candidate set.

How is external knowledge used in AI systems?

We kicked off our series meeting the IJCAI-ECAI doctoral consortium participants meeting Yash Saxena. Many AI systems now look through external documents before answering a question. In his research, Yash follows what happens to that evidence as it moves through the system.

How much can fair budget-division rules resist manipulation?

Haris Aziz, Patrick Lederer and Jeremy Vollen won a distinguished paper award at IJCAI-ECAI for their work “Approximate Strategyproofness in Approval-based Budget Division”. In this blog post, they summarise their paper and ask what the best achievable compromise is between fairness, efficiency and complete resistance to manipulation.

When compression techniques don’t just add up

What happens when you shrink a language model twice — first by removing some of its weights, then by storing what’s left with fewer bits? As Iheb Bouriel explains, the two techniques don’t just add up, and the effects can be quite different to what you’d expect.

Misleading metaphors and real risks

In her piece Misleading metaphors and real risks, Melanie Mitchell writes about the media clamour over reports that OpenAI “AI agents” under test attempted to hack into Hugging Face’s servers. If the press was to be believed these agents “went rogue” and “escaped”, with OpenAI “losing control”. However, as Melanie notes, these were just misleading anthropomorphic metaphors, with the incident actually down to poor cybersecurity measures and reward hacking.

GPT-6 Astra released

Earlier this month, the latest model in the GPT series was released with GPT-6 Astra being made available to paid users. In this blog post, Sebastian Raschka details his first impressions and improvements on previous models.

A severe misalignment of AI in mathematics

A number of Fields Medal recipients have signed a declaration warning of a “misalignment between the outcome of the use of AI and its initial purpose”. They write that we need to be weary of using generative AI for mathematics. The push by AI companies to solve mathematical problems as a benchmark misses key fundamental aspects of mathematics, those of conceptual understanding and insight.

Our resources page
Our events page
Seminars in 2026
AI pioneers series
AAAI/ACM SIGAI Doctoral Consortium interview series
AAAI Fellows interview series
AfriClimate AI series

Lucy Smith is Senior Managing Editor for AIhub.

Lucy Smith is Senior Managing Editor for AIhub.