Seven AI plays beyond TAO: io.net, NEAR, Nillion, Ritual, Fhenix, RISC Zero
Post607 words
Watchlist of seven DeAI plays beyond TAO, each with a short thesis: GPU aggregation, user-owned AI, a subnet network, private compute, inference, FHE and ZK compute.
Key claims
Bittensor has 45 subnets
Nillion, Ritual, Fhenix and RISC Zero were not yet live
Crypto x AI will be the leading narrative of the cycle. But while there are dozens of potentially life-changing opportunities, everyone is just focused on $TAO.
Don't get me wrong, $TAO is cool. However, there is much more to the DeAI space, so let me share 7 high-potential AI plays you should have on your watchlist.
Alpha ↓
$IO - @ionet is a DePIN network that aggregates idle GPU resources globally and allows anyone to spin up geo-distributed GPU clusters. It serves AI/ML companies the high-performance compute they need to train models and run inference at a competitive price. This is an absolutely huge addressable market, valued in the trillions of $$$ that spans far beyond crypto.
$NEAR - @NEARProtocol has a strong focus on building user-owned AI, massively investing in infrastructure and ecosystem builders (covered a lot of this in the past). Near is looking to become the decentralized coordination and verification layer for Artificial General Intelligence and aims to break the monopoly of big tech with novel ownership and monetization models for devs. One of my big bets on the overall AI narrative.
$TAO - @bittensor_ is the clear leader of the AI narrative right now, so I do have to mention it here too. It is a decentralized, open-source, and fully permissionless machine learning network, that creates a peer-to-peer marketplace where AI models are shared, trained, and ranked by value, incentivizing collaboration and innovation in AI development. It provides the basis of an entire ecosystem of currently 45 dApps called "subnets", with many more already on testnet.
$NIL (not live yet) - @nillion is a secure computation network that decentralizes trust for high value data through Nillion’s MPC (Multi-Party Computation) protocol, which provides improved privacy, non-linear computational capabilities, and efficiency vs. traditional MPC schemes. The blind compute capabilities of Nillion are absolutely key in the AI space, where models are often required to work with highly sensitive data (think health data or financial data).
$RITUAL (not live yet) - @ritualnet is building a sovereign inference layer that aims to provide a way for any protocol, application, or smart contract to integrate AI models with a few lines of code. Ritual provides open AI infrastructure and aims to build a groundbreaking, new architecture on a crowdsourced governance layer that provides the basis for a suite of sovereign modular execution layers, each containing specialized Stateful Precompiles (SPCs) overfit to different classes of arbitrary computation, mostly revolving around AI models, ranging from classical to foundation models.
$FHE (not live yet) - @fhenix is an Ehereum L2 and co-processor that uses Fully Homomorphic Encryption to enable on-chain confidentiality. FHE is a cutting-edge encryption scheme that enables the computation of encrypted data without ever needing to decrypt it, which allows devs to build fully confidential, end-to-end encrypted smart contracts, unlocking a whole new design space for blockchain applications. Great potential for privacy-preserving AI use cases (similar to Nillion).
$RISC (not live yet) - @RiscZero is building @boundlesshq, a verifiable computation layer that unleashes unlimited, ZK-powered execution on any chain. It is a full-stack solution to any ZK needs and encompasses a decentralized prover network, advanced integration tooling and more. It is built on @RiscZero's general-purpose zkVM R0, which is based on the RISC-V instruction set and can cryptographically prove the execution of arbitrary code, allowing devs to build ZK apps in Rust or C++. Huge potential to bring AI/ML use cases on-chain and breaking block limits that have made this impossible so far.
There is a few other very interesting ones, and I will continue this list in a follow up post.
Stay tuned anon (also NFA and DYOR).
1 image on X
What came next · since 08 Oct 2024
Show how the 5 covered assets moved
This is a landscape piece, not a set of calls. Price moves since publication are shown as market context.
Covered, no call
Mapped in this piece · best to worst since
No reliable price feed for $RITUAL, $RISC.
Since then
Zero has written 27 more pieces on NEAR Protocol and $NEAR since this one.
- Quantum and post-quantum map: PQ crypto assets by readinessOn NEAR
- I Built a Research Desk Staffed Entirely by AI Agents. It Works.Covers $NEAR
- New vs returning wallets across Ethereum, Solana, NEAR, Polygon and SuiCovers $NEAR
- NEAR DCML explained: confidential, verifiable and monetizable AI via TEEsCovers $NEAR
Daily closes in USD from DefiLlama and Yahoo Finance. Only a bullish or bearish call in a thesis or market take is tracked; explainers, lists, maps and mentions are neutral context. Not financial advice.