The unusual experiments of ADIA's 150 AI experts
The Abu Dhabi Investment Authority (ADIA) has been growing its elite team of quants and data scientists; Bloomberg says it now has around 150 of them. Many of these hires are experts in machine learning, AI, and LLMs with impressive academic pedigrees, but their experiments in the space haven't always been conventional.
Click here to follow our new WhatsApp channel, and get instant news updates straight to your phone 📱
Oliver Sheridan-Methven joined ADIA in September. He's a PhD mathematician and a former quant developer for Citadel Securities' advanced scientific computing team. Amongst his academic publications, however, is a study focused on developing a model to predict the body weight of turkeys by measuring their kinetic movements on a force plate.
Mathieu Ravaut, who joined in July, is a computer science PhD from Nanyang Technological University Singapore. He previously worked as a machine learning research scientist in the healthcare industry with Layer 6 AI; his publications include using deep learning to detect underwater mines from sonar imagery and using deep learning to aid in lung cancer detection. A 2024 study of his suggests that closed-source LLMs like OpenAI's GPT-4 are becoming "contaminated" and could soon cease to be reliable.
Dominik Seitz, who joined in September, was previously a quantum machine learning engineer at Pasqal, a French startup building "programmable quantum simulators and quantum computers made of 2D and 3D atomic arrays." He also has multiple academic publications around AI modelling for poker and other games.
Karolis Jankauskas joined ADIA last month from Millennium spinoff Worldquant, where he was VP of deep research data science. He's a PhD biochemist from University College London, and so much of his academic publications are in that space. His personal site reveals a number of quirky AI side-projects, however, including multiple facial recognition projects and one in which he teaches an AI agent to play space invaders using deep learning.
He's not the only ADIA quant to experiment with facial recognition; others have used it in a financial context. Alexis Marchal, who has been a quant at ADIA since 2021, previously used the technology to analyze video from Federal Open Markets Committee conferences to predict the market volatility. Marchal's model looked at how often FOMC chair Janet Yellen would look at internal documents when answering complex questions in the Q&A. Using the weight of turkeys to predict market volatility may pose more of a challenge...
Have a confidential story, tip, or comment you’d like to share? Contact: Telegram: @AlexMcMurray, WhatsApp: (+1 269 237 3950). Click here to fill in our anonymous form, or email editortips@efinancialcareers.com.
Bear with us if you leave a comment at the bottom of this article: all our comments are moderated by human beings. Sometimes these humans might be asleep, or away from their desks, so it may take a while for your comment to appear. Eventually it will – unless it’s offensive or libelous (in which case it won’t.)
