The history of the internet…compressed into 256 bits. @samecwilliams explains how AO collects compute logs from edges of the internet and stores them in as little as 256 bits via hashpaths. A verifiable record that something actually happened. A tiny fingerprint of

Hack a little holiday magic this season 🎄✨ Introducing AO Holiday Jam 2025. Build holiday themed games with @ao_devbot and a chance to win one of four prizes. If you’ve been waiting for a sign to ship something fun… this is it! Sign up today 👇 https://t.co/cxy37tQ7AM

Digital twins = permanent AI characters on-chain. Four are live: Trump, Obama, Orwell, Ayn Rand The community voted for #5: Satoshi Nakamoto Now being built, Bitcoin's creator: → On-chain GPU via @apus_network → Permanent storage on @ArweaveEco → Can't be deleted or

The @fwdresearch team have been relentless, refining and launching massive updates as we get to a comprehensive mainnet experience. It's time to go big. Learn the infra. More coming, and more opportunity to create. Follow us @aoComputerClub for more updates and content.

Why does any of this matter? It means AO has the capability to scale like no other decentralized network before it. Holographic state + separation of compute = contracts can execute in parallel. More nodes means more computing power. It's an elegant and innovative design.

The second part is verifiability. Contracts and all of their interactions are stored publicly, on a blockchain. In other words, anyone can read the state logs on Arweave, and given the right compute environment, anyone can verify the state. The security model is flexible.

AO nodes only execute when requested, and different nodes can execute on different contracts simultaneously (providing their state doesn't overlap). There is no global shared state on AO - it's on Arweave, instead. This means AO nodes can execute in parallel.

Arweave stores smart contract code and interactions. State is constructed on demand by putting all of these logs together. AO nodes read and write this data on Arweave to compute the current state of contracts. Two things at play here: execution on demand, and verifiability.

@aoTheComputer Blockchains have long been storing records of transactions and smart contracts onchain. Normally, execution and storage go hand in hand: nodes both store smart contracts and run computations on them when requested. AO simply separates these two pieces.

Anyone can read blockchain data. But what happens if you decide to compute on it? Here's how holographic state on AO works, for newbies.

Holidays are right around the corner and we have something fun in store for you! 🎄 More details dropping on 12.11.25. Stay tuned.

What is the through line about AO? Quality of the machine is the first detail of importance. -> Zero cutting corners. -> Solid infra. -> Scalable. The initial launch of AO was hard and imperfect. Oftentimes, things of importance don't have an easy start. It was similar

Is it real or AI generated? If we can't distinguish this from reality, how do we verify truth on the internet? Better detection tools could be the cure, but what if we address this at the root itself? Verifiable cryptographic proofs at generation in the works from

AI models are like fuzzy jpegs of the internet 🖼️ @samecwilliams shares AO's vision to make cyberspace verifiable + trustworthy. @JonnyRingo711 compares the library from the movie Time Machine to Arweave + AO. And how they can preserve years of education for generations to

Boom! All the parts of Arweave, AO & HyperBEAM and how they fit together. Deep dive even further via the "Learn" mode: https://t.co/tZiIlYyuF7.

In order to achieve the horizontal scalability, the network relies on a concept known as Holographic State instead of a shared global state. This means each process maintains its state independently of other processes. In order to see or use a program's state, the computation

Remember, this is not to be confused with another primitive, hashpaths. Every time a compute request is processed, a state change occurs. And the system provides a hashpath for it. It is a verifiable proof of the entire tree of executions, all the compute steps, that produced

As devices are stackable, we can stack several devices calls, each with their own operation, as part of a single compute request. This is done through pathing. Within the HTTP API, functionality from multiple devices can be composed into a single structured URL through pathing.

Another primitive is the concept of devices. Erlang modules that define specific computational logic. This is GIANT for creating custom logic. If AO is the computer, these modules are the pluggable components that provide additional computational utility on top of the core

This ensures HyperBEAM is compatible and can interact with our internet as is, without the need for special mediums. This means we can package our compute request messages to HyperBEAM and get back results through the HTTP API.