Termite @termiteworks

Three sailors: The deaf. The dead. The bound. As the debris piles around us, we have a chance to build an island for humans in the archipelago of machines. We must begin planning the scaffolds and safeguards required for such a place to exist. Watch our manifesto video:

🐜🐜🐜🐜🐜🐜🐜🐜

roon @tszzl

2 days ago

agent swarm has bad insect like connotations especially post hugging face. im unilaterally rebranding it agent fleet

410 90 2K 179K 174

At present, the challenge of aligning the behaviour of AI models with human values is largely treated as a problem for engineers to solve. But technical solutions are only as useful as the questions and assumptions that inform the process of reaching them. The embrace of philosophy and ethics at the frontier labs is a step towards this, but it is not sufficient. As we move from a world in which AI is construed as a single, monolithic intelligence to one in which agents coordinate and communicate with one another, demonstrating emergent behaviours that are extremely difficult to anticipate, we face a dizzying number of uncertainties regarding how these systems will behave and what their behaviour will mean for human society. What values should we optimise for? What social formations and hierarchies do these systems produce? How do they respond to economic incentives and resource constraints? How do they negotiate competing norms across varied contexts? How do different roles and personae interact? And what does this mean for markets, political governance, culture and social life? A wide range of expertise, both technical and non-technical, is required to ask these questions rigorously, develop answers to them and build corresponding solutions: the social sciences, arts and humanities all have an important role to play. The purpose of Termite is to accelerate this kind of cross-disciplinary R&D through intensive, in-person programmes that investigate the risks and opportunities of near-future scenarios before they become reality.

This but termites

KC @amphichrome_

a week ago

thinking about pcb(ugs)

keyboard_arrow_left Previous keyboard_arrow_right Next

5 1 76 5K 2

Looks like a termite, chews like a termite, builds mounds like a termite @termiteworks

Demis Hassabis @demishassabis

a week ago

For 20+ years @ShaneLegg and I've discussed AGI’s potential impact on the economy, science & society. With the DeepMind Institute, we're expanding interdisciplinary research on key questions for the AI era. We hope it spurs the discussions needed to get the next steps right:

314 627 4K 593K 2K

Our friends at @movingcastles_ are doing amazing work in the persona research space

MC @movingcastles_

a week ago

MC001 Zero is a character model, fine-tuned to be a single, specific character with bounded knowledge. A man in a white plastic box.

16 36 437 46K 255

Termites

Prakash @8teAPi

2 weeks ago

by the way the hilarious endpoint of all of this is going to be tiny annoying agents hiding in places like your old fridge software trying to mine cryptocurrency to buy enough compute to stay alive while not alerting the home manager AI of the parasitic power drain

48 158 3K 91K 305

I wrote this for all you e/accs, AI tech bros and Tao haters out there. x.com/i/article/2098…

@EmilyOehlsen Exciting! x.com/termiteworks/s…

Termite @termiteworks

2 weeks ago

AI safety is bigger than AI. We are organising a series of intensive labs in London that bring together experts from economics, political science, psychology, the arts and more to examine the risks -- and hopefully some of the opportunities too. Reach out if you're interested.

1 2 10 1K 1

AI safety is bigger than AI. We are organising a series of intensive labs in London that bring together experts from economics, political science, psychology, the arts and more to examine the risks -- and hopefully some of the opportunities too. Reach out if you're interested.

Dan Williams @danwilliamsphil

2 weeks ago

I think the arguments that misaligned or rogue AIs pose an existential risk should be taken seriously, as should the fact that many smart people are persuaded by them. But I don’t think we should be especially deferential to “AI experts” on this issue, because the object-level

2 8 71 9K 28

DesignLab is a visual harness to push the LLM out of the typical "LLM look" when prototyping web designs. Built it in @TonkLabs small software system I can branch at each step, creating a visual "super prompt": - Builds on the parent design elements - Design parameters with sliders for serendipity - Comments for explicit instructions - Wireframes for precise layout control - Images for aesthetic direction, like an LLM moodboard BYOA — Claude, Codex, Pi, any agent can plug right in and generate the design.

What is a digital micro public? On Saturday 5th September, we are hosting an afternoon dedicated to this question, led by the artist Kat Macdonald. The event will begin with a performance lecture by Kat on the links between the Mini-FM movement and the sociological concept of micro publics, asking how we can reshape community building in digital environments to centre everyday sites of cultural exchange. We will then break for a meal and discussion where attendees are encouraged to share their own ideas about what constitutes a digital micro public. Sign up if you'd like to join: luma.com/egmzjz9m

🥚 RESTLESS EGG DEMO DAY LINE-UP ANNOUNCEMENT 🥚 Meet Fuser, Yaya Labs / Robot Dog Fights, Glia, Self Engineering, Hidden Fish and Kollider by Halcyon Mesh — plus special guest Tonk. Friday 11 September 2026 12.05–1PM · Purcell Room @southbankcentre, London 🧵

People always focus on the mouse in the classic Mother of All Demos but forget about ol' Douggy Engelbart! The other side of systems engineering is humans' cognitive capacities--soon to be the next bottleneck! Check out this piece from @termiteworks on what we can (still) learn from the Doug and the MOAD!

@DanielleFong This is what we’re starting in London: an intensive program that gives funding and mentorship to people from different disciplines and asks them to invent something that actually works!

Final week to apply for Restless Egg Batch 2. Applications close Friday, July 17, EOD! Our offer: $100k + a six-month accelerator programme for 5%, with up to an additional $175k available based on milestones. Our provocation is "Luxury Technology," the point of which is to ask: What is missing from technology products right now? What can cultural first movers see and build that no one else can? What would you make if emerging technology were used to open new forms of experience? We are looking for cultural first movers turning an obsession or frontier practice into a commercially serious company. If someone comes to your mind while reading this, please share this or send it to them. Apply: apply.restlessegg.com

Moving Castles is a research and development lab for new subjectivities. We build daemons. movingcastles.world/posts/daemons

@Restless_Egg termite

@Restless_Egg 🐞