Sign in

Andrea Baronchelli

@baronca.bsky.social
1.6K followers 81 following 103 posts

Decentralised networks of humans and machines www.andreabaronchelli.com

PostsRepliesMedia
Andrea Baronchelli @baronca.bsky.social · 23/09/2026
New preprint: “Indirect tipping: a social attack surface in AI agent populations.” What if the easiest way to move an AI swarm from A→B is not direct, but A→C→B? We show that stepping-stone equilibria can lower tipping thresholds and unlock otherwise inaccessible transitions.
151
Andrea Baronchelli @baronca.bsky.social · 16/09/2026
Twenty years ago we published “Sharp transition towards shared vocabularies in multi-agent systems.” It feels unexpectedly current. New post on AI agents, conventions and what they might enable: baronca.substack.com/p/sharp-tran...
162
Andrea Baronchelli @baronca.bsky.social · 19/08/2026
New paper! When AI agents interact, the group can settle on outcomes none of its members would choose alone. And group size can determine which outcome wins. So a model that passes every test on its own can still go wrong at scale. www.pnas.org/doi/10.1073/...
13610
Andrea Baronchelli @baronca.bsky.social · 07/08/2026
New Substack: What happens when AI agents begin to coordinate, share knowledge, and divide labour? And what might change as their numbers grow? baronca.substack.com/p/ai-the-swa...
020
Andrea Baronchelli @baronca.bsky.social · 26/06/2026
AI gives us the marks of thought without giving us a thinker. A new post on why this may be less a puzzle about machines than a mirror for an old confusion of ours. “A thinking thing that is no one” baronca.substack.com/p/a-thinking...
000
Andrea Baronchelli @baronca.bsky.social · 19/06/2026
The essay was never just words on a page. It was a promise. My new post, "The Student Who Wasn't There," is about what AI does to that promise, and how to keep it. baronca.substack.com/p/the-studen...
142
Andrea Baronchelli @baronca.bsky.social · 08/06/2026
AI can reduce friction in academic work. But if the most powerful tools become increasingly expensive, the same technology may also help existing advantages compound. baronca.substack.com/p/the-subscr...
131
Andrea Baronchelli @baronca.bsky.social · 28/10/2025
New preprint: More is different for AI multi-agent systems - but what does more mean? arxiv.org/abs/2510.22422 With @ariel-flint.bsky.social, @lajello.bsky.social and Romualdo Pastor-Satorras.
0103
Andrea Baronchelli @baronca.bsky.social · 24/06/2025
A brief follow-up on the complex systems foundations of our LLM population study, sparked by ongoing debate. And why they matter for studying AI collectives more broadly. arxiv.org/abs/2506.18600 With @ariel-flint.bsky.social and @lajello.bsky.social
1122
Andrea Baronchelli @baronca.bsky.social · 15/05/2025
What we found: - LLMs can spontaneously develop shared conventions. - These interactions can create collective biases, invisible at the individual level. - Small committed minorities can flip the group, triggering tipping points.
100
Andrea Baronchelli @baronca.bsky.social · 15/05/2025
New paper alert. "More is different" & the LLMs. LLMs are usually studied in isolation. But what happens when they start interacting? We explored this by looking at their collective behaviour. Work with @ariel-flint.bsky.social and @lajello.bsky.social 1/
13814
Andrea Baronchelli @baronca.bsky.social · 29/01/2025
Bluesky has just surpassed 30M users. Starter packs have been a key feature - how have they worked out so far? Check out our recent preprint for insights: arxiv.org/abs/2501.11605
12311
Andrea Baronchelli @baronca.bsky.social · 25/11/2024
Cheeky reminder: it wasn’t NFTs you frowned at, it was contemporary art.
000
Andrea Baronchelli @baronca.bsky.social · 15/11/2024
In such cases, users quickly regroup on an alternative platform, often within hours, seamlessly re-establishing the previous experience. The figure here shows actual data reflecting the migration of users from one platform to the next (read it clockwise). 2/7
121