ALife 2025 day 5

The final day of the conference: it's been a great event.  The day yet again started with some parallel sessions of contributed papers, and I went to the session on Cellular Automata. I was chairing the session, so couldn't take a full set of notes, but I remember the talks being interesting.

Then the final keynote, by Anna Ciaunica, a philosopher, on the importance of embodiment: why we need our toes (and the rest of our body) to think.  We associate the mind with the brain; why not with the rest of the body? Why is the brain not a part of the body? Why do we have a "Brain and Body Institute", but not a "Liver and Body Institute"?  Why do we imagine Mary in her black and white world not understanding seeing red, rather than not understanding eating a red apple?  Why do we have an adult-centric neuro-centric bias, a male embodiment bias?  Think about pregnancy, not just from the mother's point of view, but from the point of view of the embryo: pregnancy is a universal experience, we have all been inside another person.  The opposite of death isn't life, it's birth.  We should think beyond autopoiesis to co-poiesis.  Currently our philosophical tools have been mostly invented by male philosophers.  Consider a parallel world, with a Renée Descartes, who is pregnant.

Well, after that mind-altering experience, it was time for the final boxed lunch.  And then the closing ceremony, bast paper awards, and thanks to everyone!

Next year, in Waterloo! 

ALife 2025 day 4

The day again started with some parallel sessions of contributed papers: this time I went to some interesting ones on cellular automata and related topics.  Then was the day's keynote, by Blaise Agüera y Arcas, who, with colleagues at Google, is working on in silico ALife ideas.  He related some of their results to ideas of symbiogenesis from biology, with a strong dash of physical processes and lots of tasty equations.

one of Blaise's denser slides

Then another efficient lunch box distribution was followed by two further contributed paper sessions.  Here I chose one on Evolution, followed by one on Morphological computation.  Lots of interesting ideas I need to assimilate.

These were followed by a Panel session, where I was the moderator. I got to interview Emily Dolson, Dominik Chen, and Hiro Iizuka about all things ALife.  I enjoyed it; I hope the panellists and audience did too.

Then a bunch of us went to dinner.

Shop name spotted on the way back from the eatery

ALife 2025 day 3

The day started with some parallel sessions of contributed papers: I went to some interesting ones on evolution.  Then was the day's keynote, by Chris Kempes.  He covered a variety of physical constraints on living organisms: physical laws limit what life can be.  Scaling laws, metabolic rates and growth rates, the way proteins are more dilute in larger cells, the physics of the small: these are all consequences of physics, and so may be universal laws.

On to lunch, and then the excursion!  On registration, we were given a choice of several excursions to a variety of local temples.  I had little to guide my choice, except I saw the words "bamboo forest" on one, so chose that.  On the bus on the way there, the guide taught us to count to ten in Japanese (knowing what little I know of Japanese, I expect the reality is much more complicated than these ten little words!)

We walked from the bus park along a road of very touristy shops, to the temple complex.  We went in the main temple, after removing our shoes.  There was a magnificent painting of a dragon on the ceiling, with a gaze that followed you around the room, but sadly, there was also a no photography rule.  After that, shoes back on, we walked through the complex, admiring the outside of buildings, the raked gravel, and the carefully tended trees.

If you look carefully, you can see three guys up in this tree, pruning it. (Click to embiggen.)

We carried on around the garden, to a lovely lake, also carefully curated.  It's just a little too early in the year for the full spectacular autumn colours.

If you look carefully, you can see a big carp in the bottom right corner.

We carried on up around the garden, to a vista point.

Temples below, mountains in the distance, framed by trees. The tree on the right has amazing spikey leaves

spikey

On to the bamboo forest.  I wouldn't want to make my way through that; fortunately there was a nice wide path.

impenetrable

Bamboo is a kind of grass.  A special kind of grass.

a colleague providing a hand for scale

After exiting the forest, we descended back down the valley, and made our way back to the bus via an ice cream shop in the touristy road.  

The bus then took us directly to the conference dinner, held in a huuuuuge and very modern convention building.  We had a delicious buffet, and wonderful conversation.  There was a bus to take us back to the starting point, but the little group I was with decided to walk back down instead.  It made a very pleasant 30 minute stroll on a warm evening to finish off a great day.

ALife 2025 day 2

Tuesday, and the full conference gets going.  

We started with a great keynote from Hector Zenil, on using Kolmogorov Complexity measures.  Just because it is uncomputable doesn't mean it is useless.  Most of mathematics is undecidable, but that doesn't stop us using maths!  We shouldn't just give up and use simple lossless compression algorithms. We can do better, and getter better bounds, better approximations, by trying harder and harder (running the search algorithm for longer), which means we can be better at filtering out randomness from things that superficially look random, but that have algorithmic patterns.  It was a fascinating talk.  I've had Hector's book on Algorithmic Information Dynamics on my to-read pile for a while now -- I need to bump it up nearer to the top.

Then on to the contributed papers.  During one of these, on Synthetic Ecosystems, I became one of today's "Lucky 10,000", when I discovered that until relatively recently, Ascension Island was an essentially lifeless desert.  Darwin stopped there on his way home, and decided it needed "greening".  Today, after a lot of work, it has a lush, totally artificial, ecosystem!

The poster session has lots of interesting work on display, including a contribution by my student.  There were also some beautiful "Analogue Atomic Automata": large, carefully grown bismuth crystals, by Solvi Arnold:

In the evening we had a fascinating online keynote from Michael Levin, on biological embodiment, agential material, and bioengineering.  It might have been 8pm for us, but it was 7am for him!  (We live on a sphere.)

view from a Kyoto hotel window

Today was the first day of the Artificial Life conference in Kyoto.  I took the obligatory photo out of my window, then joined my colleague at breakfast.

Then off to find the conference venue, and join the registration queue (at essentially midnight UK time, but I felt fine, having slept reasonably well).  Queuing also involved saying hi to many friends and colleagues I hadn't seen since last year.  This first day was dedicated to workshops. I went to an excellent Neuroevolution tutorial in the morning.  The lunch queue was long, but moved quickly, as pre-packed food boxes were being briskly handed out.  

In the afternoon I went to a workshop on Generative ALife with a wide range of talks.  I was quite fascinated by the various talks using LLMs to implement the very simple agents in Minsky's Society of Mind.  Now, I'm an LLM mega-skeptic: these stochastic parrots are most definitely not "intelligent".  However, the whole point of SoM is that intelligence emerges from lots of unintelligent agents communicating. These researchers are using LLMs as SoM's internal agents, communicating via natural language. There are agents controlling the "mouth", which talks to/at people through a robot avatar; (these conversations still have typical "hallucination" issues), other agents are attached to the "ears", "eyes", etc, and more are just internal agents, making up the whole internal society.  In addition to the transcript of the robot/people conversations, the researchers can also get transcripts of the internal conversations between the internal agents, giving insight into how the agents are interacting to give the resulting output.  I found this an interesting novel use of these devices.

Then a bunch of us went out for dinner.

A great first day!

travelling to Kyoto

I caught a Japanese Airlines flight from Heathrow early Saturday morning, to arrive in Tokyo early this Sunday morning.  The view as we were arriving had an interesting cloud.

We landed at Haneda airport after a 14 hour flight just a little late (due to a slightly delayed takeoff), to profuse apologies from the flight crew.  The next two hours shuffling slowly back and forth in the immigration queue with about 1000 other people were not fun.  But at least my luggage was ready for me after that.  

I next figured out how to buy a ticket for the monorail (I needed to pay cash for this, although every other type of ticket could be bought with a card).  I travelled to Hamamatsucho station, changed line, then on to Tokyo station.  I found a locker to hold my luggage, and had a coffee while I waited for a colleague who I had arranged to meet up with in Tokyo and travel with to Kyoto.  We explored a bit around the station, had lunch, then went and caught the Shinkansen bullet train to Kyoto.

The train was roomy, clean, busy, quiet, and on time.  It was also fast, but that was difficult to tell, as it was all so smooth.  

It is supposed to be possible to see Mount Fuji from the train, but it was too cloudy.

Not Mount Fuji, in the clouds.  Look at that blur in the foreground!

We arrived in Kyoto early evening, checked in to our hotel, then went out for a meal.  Sunday evening, many places were full.  But my colleague, who had been to Kyoto before, and knew some Japanese, found a lovely little place in a backstreet, where we ordered a set menu.  They kept bringing us lots of little plates of wonderful food, and I discovered I like wasabi, in small quantities. (I seem to have missed this staple on previous trips.)  The meal ended with a lovely desert of a variety of not-too-sweet ices.

Yes, I photographed my desert.

Back to the hotel, to recover from the journey, and prepare for the conference starting tomorrow.  Only an eight hour time difference, the "wrong" way; I'm sure that won't be a problem!

no view from a lack of hotel window

I travelled down to Heathrow last night, to the hotel at T3, ready for my early flight to Japan today for the ALife conference.  I went to take a photo from the window, as usual.  No window!  I looked at the emergency exit floor map. No outside wall!  Okay.  Perfectly comfortable room, nonetheless.

That wall covering over the bed is not a blind: there is no window.

Towards Origins of Virtual Artificial Life

I have a new paper, "Towards Origins of Virtual Artificial Life: an overview".  This is in a special issue of PhilTransRoySocB, on Origins of Life.  I am also one of the three editors of that SI (but this paper was still properly peer reviewed, handled by one of the other editors, I hasten to add!)  

abstract:

The field of artificial life (ALife) studies ‘life as it could be’, in contrast to biology’s study of ‘life as we know it to be’. This includes a wide range of potential physical substrates, from synthetic biology (new genes), through xenobiology (new amino acids and DNA bases), inorganic chemistry (different structural elements), soft and hard robotics (new kinds of bodies) and also virtual life (existing inside a computer). Since any such life forms are artificial, the originating mechanisms can be similarly artificial, or can attempt to emulate natural mechanisms. Given the wide range of possible substrates and origins, it is crucial to have good definitions, and well-defined ways to detect and measure life, if and when it originates. This overview examines the current state of the art in ALife in defining, detecting and originating its subject matter, with its main focus on virtual life. After discussing common properties of several definitions of life, the overview synthesizes an engineering-focussed definition, in terms of abstract requirements, generic designs and specific implementation mechanisms, and then reviews the current state of the art through this lens. Although virtual ALife that satisfies all these requirements is yet to be exhibited, significant progress has been made on engineering individual mechanisms and, arguably, partially alive systems.

I had fun writing it, thinking about ALife in the context of origins of life.  Given it is artificial life, that implies an artificer, so it has to be an engineering origin rather than a natural origin.  So I get to exercise my Requirements Engineering knowledge.

It's open access, and can be found at doi:10.1098/rstb.2024.0298

A foolish consistency is the hobgoblin of little minds

I am off to the Artificial Life conference in Kyoto soon; I am filling in some online forms to save time in person later.  As is often the case, such forms want date of birth.  I am bemused by the day-of-month dropdown menu.