iiinova — where your digital self keeps evolving

Milton @miltonllera.bsky.social · 05/07/2026

Don’t know about ultimate but it is a start 😅

Milton @miltonllera.bsky.social · 04/07/2026

We are also running this again next year in Norway: aliceworkshop.org. Plenty of people from ALife here (and adjacent fields), so do apply ;)

aliceworkshop.org

ALICE Workshop

Copenhagen, Denmark February 2nd to 6th, 2026

141

Milton @miltonllera.bsky.social · 04/07/2026

Main journal is the ALife journal direct.mit.edu/artl and ISAL is the overarching org. GECCO (gecco-2026.sigevo.org/HomePage) and CCS (ccs26.cssociety.org) are also popular with people in the field. There may be others that I am missing

direct.mit.edu

Artificial Life | MIT Press

131

Milton @miltonllera.bsky.social · 04/07/2026

One of my colleagues created this back when BS started gaining traction: bsky.app/starter-pack...

1147

Milton @miltonllera.bsky.social · 22/06/2026

Unpopular opinion (on BS at least): they solved this in the US, we are just too stubborn over here.

nbes @nbes.bsky.social · 18/06/2026

Reposted by Milton

Our paper with @eplantec.bsky.social was accepted at @alifeconf.bsky.social How does criticality move from agents to populations? We study how near-critical dynamics inside individual agents shape collective dynamics at the population scale. arxiv.org/pdf/2605.01818

0166

Learning Developmental Scaffoldings to Guide Self-Organisation

7/n Thus we show that the system can offer a valuable insight related to the functional utility of pre-patterns. Furthermore we can use it to replicate how embryogenesis combines growth and self-organisation. Check out the preprint here: arxiv.org/abs/2605.14998

arxiv.org

From subcellular structures to entire organisms, many natural systems generate complex organisation through self-organisation: local interactions that collectively give rise to global structure withou...

050

6/n Most surprisingly, we find that the pre-patterns are not used to directly approximate the target. Indeed, even though more complex SIREN functions support better target generation, the initial patterns are not closer to the target when we increase complexity

130

5/n Using a goal-directed NCA formulation, we show that for similar parameter counts, using pre-patterns leads to systems that can encode more patterns:

4/n To test the first result we trained the model using different levels of reliability in the cell update and compared to a standard NCA which can only self-organise. The results show that as we decrease cell reliability, pre-patterns help with robustness:

130

3/n Our results show 3 things: 1. Pre-patterns improve robustness, 2. they increase model capacity, 3. and, perhaps surprisingly, are not used to directly approximate the target (or some approximation of it)

2/n To study this question we paired an NCA with a learned spatial pre-pattern generator in the form of an implicit representation model, specifically a SIREN. The pre-pattern acts as a scaffolding which sets initial conditions of the NCA but cannot reproduce the target directly

1/n Self-organisation is studied as a mechanism that works in isolation, unfolding development from random patterns. But biology rarely works this way; embryos start with spatial information guiding development. In this work we model this computationally arxiv.org/abs/2605.14998

Milton @miltonllera.bsky.social · 31/03/2026

New research out — A developmental model with morphogenetic scaffoldings to guide self-organisation: a single model, many grown patterns. The memory-compute trade-off but in self-organizing systems. A collab featuring @najarro.science, Jakob Schauser and @risi.bsky.social and yours truly

1162

Maybe the question here is what it say about peer-review, even if in an admittedly low-standards venue. Sort of an unintentional Sokal Affair for AI research...

Milton @miltonllera.bsky.social · 25/02/2026

We are thrilled to be returning to GECCO for a second edition of the Evolving Self-Organisation Workshop. We are now accepting paper submissions, so come join us! Check out our website for more info and upcoming announcements evolving-self-organisation-workshop.github.io/gecco-2026/

0114

Reposted by Milton

ALICE Workshop @aliceworkshop.org · 10/02/2026

What a week—projects, tutorials, discussion groups, art demo & even a complexity-alife quiz! Thank you to all the attendees who brought a fantastic energy to this year's ALICE workshop. We can't wait to see you again next year, this time in Norway 🇳🇴 Thank you to @aicentre.dk and @itu.dk !

0168

and 4) which phenomena we should model together given that we don’t want to do model then in isolation but neither can we model them all at the same time.

000

We agree on that. For me the disagreement is more about: 1) the relative importance of both types of tests, 2) how many conclusions you can draw from naturalistic stimuli given increased model complexity, 3) how particular phenomena manifest if at all in those stimuli 1/2

100

I would also agree with this. The question is are we that much closer to achieving this? The main issue seems to be that to scale up we need models that are increasingly hard to understand. Thus we end up with not one but two different systems we have to explain, no?

000

The main concern on my side is that naturalistic stimuli may introduce unknown unknowns into your experimental design which throw off your conclusion. Thus you need to be extra careful. But I agree that testing a lot of these phenomena on more realistic settings is important

I agree with this, yes. The artificial stimuli are diagnostic tools. Like edge cases for algorithms in computer science; we should use them to find out what is missing. I guess the disagreement (in so far as there is one) is on the relative importance between artificial and naturalistic stimuli.

Milton @miltonllera.bsky.social · 09/02/2026

Sure, but it may give you some indication of how things should work. For example classical experiments that illustrate Gestalt phenomena tell you something important about how visual stimuli are grouped. I don't think (and don't believe you think) these don't manifest for naturalistic stimuli.