Deal ex Machina

Web3 beyond hype & greed, taking the long view, with in-depth knowledge, multiple fields. Focused on firms, markets, institutions.

[Deal ex Machina @dealexmachina]

« In democracies, the art of association is the mother science. » Tocqueville, 1835

[Deal ex Machina @dealexmachina]

Solving agents coordination. https://t.co/xIhlAauQhM

[Deal ex Machina @dealexmachina]

Merchants can integrate cool multi garments virtual try-ons and trigger promo for the combo.

Had fun setting up a demo with @bfl_ai Flux Virutal Try On. Public repo for fun. Proxy deployed on @gokoyeb.

https://t.co/BpfjUyEbab

[Deal ex Machina @dealexmachina]

Say Hi to Wagmi https://t.co/xdchGyPwBJ via @LinkedIn

[Deal ex Machina @dealexmachina]

Governed swarms are constitutional systems, not chat topologies. A deterministic causal layer records what happened. A semantic lattice decides what can count as admissible knowledge, with non-scalar finality that prevents compensation attacks. A propagation layer diffuses candidate evidence.

[Deal ex Machina @dealexmachina]

https://t.co/akqe71Kai7

[Deal ex Machina @dealexmachina]

https://t.co/Uzj6IBFXln

[Deal ex Machina @dealexmachina]

Can we replace orchestration-as-topology with governance-as-policy for agent coordination -- with formal convergence guarantees, bitemporal auditability, and a perpetual compliance lifecycle?

[Deal ex Machina @dealexmachina]

Hint : yes we can. And it's exposing useful properties for regulated context.

[Deal ex Machina @dealexmachina]

With Riemann, the notion of a manifold emerged — a space that may be globally complex but locally smooth.
A system no longer “chooses” a path in ordinary space.
It evolves on a manifold — an abstract space of possible configurations.

[Deal ex Machina @dealexmachina]

In large regulated orgs, we need governed state graphs for always-on agent swarms coordination, don't we ?

- Wu Wei style
- Least Action
- Margulis and lie groups, high dimensionality.

Like or follow if interested.

[Deal ex Machina @dealexmachina]

Most firms can’t release AI agents “into the wild.” The stakes are too high. See: Agents of Chaos (MIT / Harvard et al., 2026)
https://t.co/b1Wcemg7pc

So they wrap them in orchestration layers, approval logic, monitoring stacks, exception handlers. The whole enterprise grade enchilada. At scale (10 agents), it becomes a Rube Goldberg machine. Coordination cost grows quadratically.

We build agents for autonomy and deploy them in containment.

We’re reaching an architectural inflection point.

Time for a different geometry.

[Jean-Baptiste Dézard @jbdezard]

« Contact sales », c’est l’abandon de panier assuré. Le tue l’amour.

On a Stripe aujourd’hui, X402 demain.
Vous avez combien de « CTA » qui tuent votre revenu passif?

Cc @Maldoju

[Deal ex Machina @dealexmachina]

testing https://t.co/Yd1MDwkyM2

[Deal ex Machina @dealexmachina]

testing https://t.co/Yd1MDwkyM2

[Deal ex Machina @dealexmachina]

testing https://t.co/Yd1MDwkyM2

[Deal ex Machina @dealexmachina]

I like to cook with my friends @DragonLLM .

[OVHaiLLM @OVHaiLLM]

https://t.co/dBM2pTJ38s

[Koyeb @gokoyeb]

❤️‍🔥 Perks of running AI on high-performance serverless infrastructure 👇