Miscellaneous

Administrator

Administrator
Staff member

Luc Steels: Can Robots invent their own language?

Published on Jan 19, 2015

For more than a decade we have been doing robotic experiments to understand how language could originate in a population of embodied agents. This has resulted in various fundamental mechanisms for the self-organisation of vocabularies, the co-evolution of words and meanings, and the emergence of grammar. It has also lead to a number of technological advances in language processing technologies, in particular a new grammar formalism called Fluid Construction Grammar, that attempts to formalise and capture insights from construction grammar, and a new scheme for doing grounded semantics on robots.

This talk gives a (very brief) overview of our approach and discusses some details of the technical spin-offs that have come out of this work. The talk is illustrated with live software demos and videos of robots playing language games. The talk ends with a number of open problems and issues that we need to tackle before having adaptive open-ended language communication between humans and robots.

Luc Steels - ICREA/IBE, Barcelona - Sony CSL Paris

Luc Steels studied linguistics at the University of Antwerp (Belgium) and computer science at MIT. After working in the domain of expert systems for geophysical data interpretation at Schlumberger, he founded in 1983 the Artificial Intelligence Laboratory of the Free University of Brussels (VUB). There he worked with his students on many topics of AI, ranging from knowledge engineering and machine learning to computational linguistics and robotics. In 1996 Steels founded the Sony Computer Science Laboratory in Paris which he directed until april 2014. This laboratory has made many contributions to developmental robotics, musical creativity, and computational linguistics. At the moment he is ICREA research professor at the Institut de Biologia Evolutiva in Barcelona (UPF-CSIC) working on the evolution of language.

Steels is a member of the European Academy of Science, the New York Academy and the Royal Academy of Science of Belgium. His publications include a dozen books and hundreds of papers in top level journals, such as (Nature Physics, BBS, AI Journal, Trends in Cognitive Science, etc.)

Soon a re-edition of his 'Talking Heads Experiment' book will appear with the Language Science Press (Berlin).

The A-Talks are gathering world-reknowned experts in fields related to robotics & AI, presenting ground breaking ideas in various domain from robotics and computer science, to psychology, language or social sciences. We want to bring to the audience fresh new ideas, just beyond the level of popular science into the real core of today's top most advanced research in the field. Our ambition is that if you listen to an A-Talk, there is a good chance you'll hear ideas that you've never heard before!

Languages AI and robots studies by Luc Steels
 

Andrej Karpathy: Software Is Changing (Again)

Jun 19, 2025
Andrej Karpathy's keynote on June 17, 2025 at AI Startup School in San Francisco.
Chapters

00:00 - Intro
01:25 - Software evolution: From 1.0 to 3.0
04:40 - Programming in English: Rise of Software 3.0
06:10 - LLMs as utilities, fabs, and operating systems
11:04 - The new LLM OS and historical computing analogies
14:39 - Psychology of LLMs: People spirits and cognitive quirks
18:22 - Designing LLM apps with partial autonomy
23:40 - The importance of human-AI collaboration loops
26:00 - Lessons from Tesla Autopilot & autonomy sliders
27:52 - The Iron Man analogy: Augmentation vs. agents
29:06 - Vibe Coding: Everyone is now a programmer
33:39 - Building for agents: Future-ready digital infrastructure
38:14 - Summary: We’re in the 1960s of LLMs — time to build
Drawing on his work at Stanford, OpenAI, and Tesla, Andrej sees a shift underway. Software is changing, again. We’ve entered the era of “Software 3.0,” where natural language becomes the new programming interface and models do the rest.

He explores what this shift means for developers, users, and the design of software itself— that we're not just using new tools, but building a new kind of computer.

Thoughts (From Andrej Karpathy!)
0:49 - Imo fair to say that software is changing quite fundamentally again. LLMs are a new kind of computer, and you program them *in English*. Hence I think they are well deserving of a major version upgrade in terms of software.
6:06 - LLMs have properties of utilities, of fabs, and of operating systems → New LLM OS, fabbed by labs, and distributed like utilities (for now). Many historical analogies apply - imo we are computing circa ~1960s.
14:39 - LLM psychology: LLMs = "people spirits", stochastic simulations of people, where the simulator is an autoregressive Transformer. Since they are trained on human data, they have a kind of emergent psychology, and are simultaneously superhuman in some ways, but also fallible in many others. Given this, how do we productively work with them hand in hand?Switching gears to opportunities...
18:16 - LLMs are "people spirits" → can build partially autonomous products.
29:05 - LLMs are programmed in English → make software highly accessible! (yes, vibe coding)
33:36 - LLMs are new primary consumer/manipulator of digital information (adding to GUIs/humans and APIs/programs) → Build for agents!
 
Back
Top