Nello Cristianini
[introductory] Machina Sapiens — Towards More General Forms of AI
Summary
Intelligence existed long before human beings, and does not require language nor logic. All intelligent systems in use today have a form of intelligence entirely different from ours. The current paradigm of Artificial Intelligence is the result of a series of choices that shape what we can expect from it. From the adoption of a data-driven approach in the late 1990s to the use of “data from the wild” in the early 2010s to the introduction of ideas such as implicit feedback, commonplace systems like recommender systems have been determined by those choices. Issues such as bias, polarisation and explainability follow from those choices.
The later introduction of Large Language Models builds on those ideas, adding new concepts, such as self-supervised learning from text, as well as emergent properties. This new generation of AI systems is now close to passing Turing’s test. How was our first encounter with the first non-human entity capable of sustaining a credible conversation? Finally, is the next goal that of a human-level intelligence? How would we pursue it, and how would we evaluate it? Performing the same tasks that humans perform is not a requirement for intelligence, but has great economic and cultural value. The journey continues.
Syllabus
- The shortcut — how machines became intelligent without thinking in a human way.
- Machina sapiens — the story of Turing’s test and Large Language Models
- The next stage: AI competing on traditionally human tasks?
References
Book 1 – The Shortcut, CRC Press, 2023.
Book 2 – Machina Sapiens, Il Mulino, 2024 – forthcoming also with CRC Press, 2025.
Book 3 – to be decided.
Pre-requisites
No prerequisites.
Short bio
Nello Cristianini is a Professor of Artificial Intelligence at the University of Bath, specialising in machine learning. Before that, he was a professor at the University of Bristol and the University of California, Davis. He is the author of the books: The Shortcut – why intelligent machines do not think like us, and Machina Sapiens. https://www.linkedin.com/in/nello-cristianini.
