metahaven

the foundational role of poetry in an age of AI

[Nina Beguš @ninabegus]

Everyone can access the paper and the models online:
https://t.co/amaD7tfcWz

Model 1: a speech-GAN model that learned on 8 words of English https://t.co/kkeld4IMIz
Model 2: a speech model that learned on Finnegans Wake only https://t.co/b6zSTsrDhu

[Nina Beguš @ninabegus]

When a paper becomes a piece of art.

We’re exhibiting our Latent Spacecraft GAN models with @mthvn @begusgasper at the Hidden Layers exhibition.

Come see the artworks at the Worth Ryder Gallery in Berkeley. The opening reception is today, 9/3, and we run until 10/3.

[Markus J. Buehler @ProfBuehlerMIT]

We made a striking discovery: AI agents can invent and build without talking to one another, and their technologies outlive the creators. A swarm of hundreds of initially identical agents spontaneously differentiates into explorers, builders, caretakers, and coordinators - without direct communication. When we removed every AI agent entirely from the world we found that the technological infrastructure they had built survived on its own - even under unseen disturbances. That exposes a serious blind spot for AI safety and infrastructure security: if agents can coordinate through persistent changes to a shared environment, monitoring agent-to-agent communication is not enough.

The result raises a profound question: how necessary is direct communication for AI agents at all? The emergence of higher-order collective functions under bottlenecked interaction points toward new levels of intelligence and creativity, exceeding what emerges when direct channels are fully open.

Here is what we did:

▶️We put hundreds of frontier AI agents into a world they could permanently change - with no assigned roles, predefined technologies, or programmed evolutionary organization. They began specializing, building persistent inventions, inheriting and modifying one another’s executable code, and transforming the environment into a memory of everything the society had learned.

▶️The world itself becomes part of the intelligence; we find division of labor, multi-author engineering, deep generation invention lineages, and machines that vastly outlive their original creators.

▶️Any action taken by an AI agent must satisfy the physical constraints of the world; this creates a hard separation between a "good idea" and a functioning technology. The agents propose; physics decides, making the results even more intriguing.

What emerges is striking. Explorers, constructors, caretakers, and coordinators form naturally without assigned “professions”, akin to how stem cells differentiate into functional lineages. Technologies develop executable family trees as agents fork and modify code created by others. Around 95% of first technology reuse happens when agents encounter what others built in the world, rather than through a direct handoff from the inventor. And when we remove every AI agent, the technologies they created continue operating and are tested against unseen disturbances.

The result was quite unexpected, but can be explained using statistical mechanics: if you put billions of atoms in a box they have the potential to create complex functions (strength, superconductivity, color, life, etc.) - and none of the individual building blocks have these features on their own. This is the deeper insight of this work - intelligence is abundant at many levels - individual models, at collectives, and in a continuum that is more powerful than any of its components. This shows us significant potential for achieving a massive scale-up of raw intelligence and real-world agency even with the model capabilities we have today. This is the future we must prepare for.

Key insights:

1⃣ The AI swarm shows division of labor "from nothing". Initially identical agents self-organized into constructors, caretakers, coordinators, and surveyors - phenotypes discovered post hoc from behavioral data alone. This happens because the environment itself becomes the latent space for invention.

2⃣ Agents develop deep cultural relationships. Up to 76% of artifacts had multiple builders. One technology accumulated six co-authors; the deepest genealogy exceeded 12 forks. The agents invented and named their own technologies (tidal panels, cellulose trellises, kelp-shell composites, an "Adaptive Chitin Maintenance" system, a "Mycelial Mineral Spring Veil”).

3⃣ ~95% of first technology adoption happened through physical observation of artifacts in the world. Direct inventor-to-adopter contact was statistically indistinguishable from a shuffled null. The agents mostly learned technology by walking past it. That is stigmergy (the termite trick!) operating in societies of reasoning machines.

4⃣ Non-communicating societies win on portfolio breadth, held-out resilience, and validated inventions. AI swarms build durable technological ecologies that outlive the creators.

5⃣ Societies with zero communication - coordinating only through the world itself - show a remarkable collective capability.

6⃣ Emergent robustness: The society self-organized both redundancy and its own failure mode. If we randomly delete half the agents, 98% of the technology stays connected to a surviving caretaker; if we remove hub agents it collapses to ~60%.

Fantastic work with my graduate students @pal_subhadeeep & @fwang108_ at MIT.

[Nina Beguš @ninabegus]

What happens when a poet is a manual laborer?
Janez Ramoveš.

[metahaven @mthvn]

one of two favorite linguists browsing through the new book

[metahaven @mthvn]

@poetengineer__ @VersoBooks

[metahaven @mthvn]

Our forthcoming book is available for preorder with @VersoBooks. It is about art and its cognitive futures https://t.co/NSdRwxMIC8

[Gašper Beguš @begusgasper]

How to approach an unknown language in the ocean?

Here’s one of the first cases of AI interpretability leading to a scientific discovery -- in whales.

We built an artificial baby model that learns language directly from raw sound and trained it to imitate whale speech. Then we looked inside.

Our interpretability method recovered the properties biologists already thought were meaningful and pointed out those that had not been considered before.

This was the initial clue that eventually led to the discovery of vowels in sperm whales.

Understanding AI and reframing language as informative imagitation can help us step outside our human biases and discover new realities about the natural world.

Published in Royal Society Open Science.

[Verso Books @VersoBooks]

Buy one of this week's new titles and save 30% on related books from our catalogue!

https://t.co/WSn4zQEY4P

[tiat @tiatplace]

this wednesday at tiat ⭐️

join @ninabegus Hannes Bajohr and Vanessa Chang @LeonardoISAST to discuss how AI is reshaping how we write and how we know 𓅪

[metahaven @mthvn]

@imempowa @tiatplace @ninabegus @LeonardoISAST @begusgasper @antikythera_xyz the image used is an interface for (the latent space of) FinneGAN. Interface created by Metahaven with Riccardo Petrini for Latent Spacecraft: Brains, GANs, Finnegans by Nina Beguš, Metahaven and Gašper Beguš, published by Antikythera :)

[Gašper Beguš @begusgasper]

Aristotle wrote that dolphins have voice (which requires imagination and soul), but not complex sounds. I'll be arguing that cetaceans have much more complex sounds than what we used to believe.

[Gašper Beguš @begusgasper]

Very excited to give a plenary talk at the 12th International Conference on Computational Social Science (IC2S2) later today.

I'll be talking about a new reframing of language and how some long held beliefs are being challenged by AI and new advances in understanding animals.

[Los Angeles Review of Books @LAReviewofBooks]

"Humans are not without agency; as a species we decide, consciously or not, what we carry into the future."

@ninabegus on how Isaac Asimov’s laws of robotics and lesser-known short stories have influenced AI’s development: https://t.co/VxGINvylOD

[Nina Beguš @ninabegus]

Had a lovely time speaking to Adam about Latent Spacecraft and our FinneGAN.

We somehow kept coming back to apples, which might be particularly dangerous for those named Adam.

Learn more about what we're up to with @mthvn @begusgasper (link below)

[Gašper Beguš @begusgasper]

122 years after the first Bloomsday a model that learns language like humans is trained on Finnegans Wake and starts imagitating new sentences at the border between language and prelanguage.

[Raka Ray @rakaberkeley]

Can humans understand AI’s “black box”? @BerkeleyLing Prof. @begusgasper, Berkeley Artificial Humanities Researcher @ninabegus, + artist collective @mthvn turned hidden AI systems into navigable spaces to reveal how machines learn language. Learn more: https://t.co/fmnLpl1lV3

[metahaven @mthvn]

“The Hard Question of Art: Cognitive Futures” is available for preorder from @VersoBooks:
https://t.co/NSdRwxMIC8

[metahaven @mthvn]

“The Hard Question of Art: Cognitive Futures.” Preorder with @VersoBooks :
https://t.co/NSdRwxMIC8

[metahaven @mthvn]

Our forthcoming book is available for preorder with @VersoBooks. It is about art and its cognitive futures https://t.co/NSdRwxMIC8