Events

Date: 11 May 2026   Time: 11:00 - 15:00

Location: Maths Lecture Theatre, Mile End E1 4NS Map 

These events will be of interest to members of the centre: The school of mathematical sciences will be hosting Professor Enrique Alba (University of Málaga) for two talks on the 11th of May.

Modern AI: From Engineering to Scientific Approaches – 11am in the Maths Lecture Theatre and then

Reshape Your Field: An Experienced Guide's Playbook for the European Research Council – 2pm in the Maths Lecture Theatre.

Details

Modern AI: From Engineering to Scientific Approaches – 11am in the Maths Lecture Theatre.Abstract:

Modern AI is rewriting the contract between research and society at extraordinary speed, and the researchers who have spent decades studying intelligent systems are now being asked to adapt — to read the new architectures with classic instruments, and to forge new ones where the old ones fail. This talk presents an ongoing research programme that approaches the field from two complementary points of view: an engineering approach, where AI is applied to solve real problems, and a scientific approach, where AI itself becomes the object of study, explanation, and redesign.

The engineering side is anchored in three decades of work on AI for software engineering: testing, code repair, agentic orchestration, automated cognitive-complexity reduction, and more. Besides this, application domains were faced, where AI meets the physical world: smart cities and mobility, circular economy, bioinformatics, and healthcare. Innovation transfer to companies has been a constant: industrial pilots, contracted research, spin-off-grade software.

The scientific side is newer and bolder. The talk will present a dive into quantum computing (QAOA, hybrid metaheuristic-quantum solvers) and into other new reserarch lines that turn classical optimisation and search tools into instruments for studying modern AI: white-box transformer refactoring for 30–40% energy savings on production-class LLMs, heterogeneous mixed-precision quantisation reducing ≈42% of the needed memory, per-token energy profiling of tiny LLMs on real hardware, recovering training data from weights in MLPs, evolutionary algorithms applied to statistical optimal transport (Wasserstein), and contract-based inference as a bridge toward the new world models.

The talk closes with the open lines that connect this programme to academy, industry, and society: the three communities modern AI must answer to.

Reshape Your Field: An Experienced Guide's Playbook for the European Research Council – 2pm in the Maths Lecture Theatre.

Abstract:

There is one place in the world where a single scientist can walk in with the most ambitious, most uncertain, most personally meaningful idea of their career: no consortium, no industrial partner, no policy alignment, no deliverables to a funder's roadmap... and still be evaluated on the one right thing only: scientific excellence at the highest world level. That place is the European Research Council (ERC).

The ERC was built for the researcher who wants to reshape a field, not iterate inside it. Five or six years of freedom to pursue the question that has been quietly haunting you. A team of your choosing. A host institution that signs to protect your independence. And a panel of peers whose only job is to ask: is this ground-breaking, and is this the mind to do it?

This talk is for the scientist who suspects that idea is in them, and for those already drafting and wondering why the door has not yet opened.

Drawing on personal thoughts, the session translates a 200-page Work Programme into a pragmatic and oriented playbook: how to recognise an ERC-grade idea, how to choose your scheme and panel without overthinking, how to write Part I/II so generalists and experts see your correct vision, how to build a CV that tells a story. And how to navigate the 2026 changes and what is coming in 2027.

These talks are open to researchers at any career stage, anywhere in the world. Bring the idea you are afraid to write down. Leave with the plan to defend it.

Bio:

Dr. Alba is Professor of Computer Science at the University of Málaga (1993-). He earned his Ph.D. on parallel neuroevolution in 1999, and now works at the intersection of optimisation, learning, and the engineering of modern AI systems, including large language models and agentic AI, with applications in software engineering, quantum computing, bioinformatics, vehicular networks, logistics, and smart cities. He is the fifth most influential Computer Science researcher in Spain (JCR), the first of his university across all disciplines, and is listed in Stanford's top 2% of world researchers, with an H-index of 71 and 24,700+ citations on DBLP. He has opened pioneering lines on AI for software engineering, AI and energy, quantum AI, and circular economy with swarm intelligence, and is an established expert in parallel and high-performance algorithms, multi-objective and dynamic optimisation, federated learning, IoT, and heterogeneous self-adaptive techniques.

Prof. Alba has authored 12 monographs, 140+ ISI-indexed journal papers, and 350+ conference papers, directed many national and international projects and PhD theses, and collaborates with 20+ international labs. His industrial impact includes several patents and software packages in exploitation.

He is deeply involved in ACM SIGEVO, an ACM Distinguished Speaker (since 2020), a SICSA Distinguished Visitor (Scotland, 2021), three times designated by Japan as an excellent foreign expert in Informatics, and has served on AIDA, CLAIRE, and AI&Space. Over the last five years he has worked as Seconded National Expert at the European Research Council (ERC) in Brussels, managing computer-science funding programmes across Europe, a chapter now closed as he returns to research, this time on the modern AI agenda.