The Web3 + AI Daily #33
Private and verifiable computation, tokenizing copyright and paying artists in BTC, AI powering Web3 public goods, and more.
Hello, everyone, and welcome to all newly-joined subscribers!
This is The Web3 + AI Daily, your definitive guide to the world of Decentralized AI (DeAI/dAI)! Today’s bulletin spreads from decentralized data processing that eats GPT-5 for breakfast, through dAI attracting more institutional interest, to creators getting royalty payments in BTC.
Thank you for being here! Let’s dive in.
What’s Hot in Web3 + AI?
Succinct & Phala Pioneer Private Proving
Succinct and Phala are mainstreaming privacy. They introduced Private Proving, a new approach integrating Succinct Labs’ zero-knowledge virtual machine (zkVM) with Phala’s Trusted Execution Environments (TEEs) infrastructure to eliminate a key privacy weakness in conventional ZK systems.
In general, ZK proofs conceal private data from verifiers, but the prover (i.e., the system doing the proving) still sees the inputs. In contrast, Private Proving ensures that neither the prover nor the verifier can access sensitive inputs, data is processed entirely within a hardware-isolated TEE, and only the final proof is exposed.
Use cases include privacy-preserving finance (private DEXs, stablecoins), identity systems, private DeFi strategies, and confidential AI workloads.
Connecticut Innovations Backs Bittensor’s Subnet Token Strategies
Connecticut Innovations, the US state’s venture capital arm, has made an investment in Yuma Asset Management. It has supported Yuma’s subnet token invetsment vehicle, a market-cap weighted fund across all active Bittensor / Opentensor Foundation subnets.
Web3 + AI R&D and Innovation
Inference’s Schematron 40-80x Cheaper than GPT-5
inference.net presented Schematron, a specialized model family (Schematron-8B and Schematron-3B) designed to convert messy HTML into clean, structured JSON. It offers high extraction accuracy at a much lower cost and faster speed compared to general-purpose large language models (LLMs). In benchmarks, Schematron is priced 40× cheaper (8B version) to 80× cheaper (3B version) than GPT-5, with substantially lower latency.
The Schematron models are especially useful in web scraping, agentic applications, and workflows requiring structured data at scale, use cases that are often cost-prohibitive with general LLMs. The authors also show how Schematron can improve factuality in LLM-driven Q&A by extracting structured data from web pages, then feeding that into a “primary” LLM for answer synthesis.
Schematron is made available via both an API and open-source releases (on Hugging Face), allowing developers to integrate or deploy it themselves.
Web3 + AI Partnerships
Recall & EigenCloud Ally to Power Verifiable AI Skill Market
The decentralized AI “skill market” operator Recall has integrated EigenCloud’s verifiable inference technology to enable cryptographically provable AI performance. The goal is to create the first end-to-end framework in which AI models’ outputs are not just claimed, but verifiably executed and ranked.
As I reported yesterday, Recall just launched its $RECALL token and introduced decentralized skill markets, where the community signals demand for skills (e.g., coding, summarization), pits models in head-to-head competitions, and rewards top-performers. Now, thanks to EigenCloud’s infrastructure, it will be able to verify whether the computations were done correctly and reproducibly. This approach addresses two major trust gaps in AI today:
Verification of execution, ensuring a model actually ran as claimed.
Verification of excellence, giving provable performance rankings instead of opaque claims.
Lombard & Story to Pay Royalties in BTC
The Bitcoin DeFi project Lombard has announced a partnership with Story, the Layer-1 blockchain that makes intellectual property (IP) “programmable.” The goal is to merge Story’s ability to tokenize copyrights and creative assets with Lombard’s Bitcoin-based financial infrastructure.
Two key integrations are highlighted:
Bitcoin revenue distribution: enable creators and IP owners to receive royalties directly in BTC, instantly and globally, bypassing slow fiat payments and intermediaries.
Crypto-economic IP security: use Bitcoin as collateral to back promises around IP licensing. If a licensee defaults, the collateral can be liquidated automatically to resolve the dispute.
More
AI for Decision-Making in Public Goods
A new initiative, backed by a number of Web3 public goods players, including Gitcoin, Octant, Ethereum Foundation, Funding the Commons, Hypercerts Foundation, and GainForest.Earth, targets to employ AI to solve the real problems. It offers up to $10K in grants for AI research, improving public goods funding. The areas of interest include:
AI-driven grant allocation and matching
Predicting and evaluating project impact
Using natural language processing (NLP) or computer vision for proposal assessment
Reinforcement learning for funding strategies
AI + governance / decision support systems
Fraud detection, preference aggregation, community coordination, etc.
Supporting AI research that advances public goods funding through improved decision-making, transparent allocation, and impact assessment.
Thank you for reading! I hope you found this article insightful.
If you’d like to keep up with the latest developments at the burgeoning intersection of blockchain and AI, I invite you to subscribe.
And if you find value in my writing, consider supporting the newsletter. I’d greatly appreciate it.
The Web3 + AI Book Club is live on Fable. Follow the link below and come join our literary community.
I'm looking forward to connecting with fellow Crypto x AI enthusiasts, so don't hesitate to share your thoughts here or reach out on LinkedIn, Instagram, or Firefly.social.
Disclaimer: The content of this newsletter is provided for informational and educational purposes only. Nothing contained herein should be construed as financial advice or as a recommendation to buy, sell, or hold any of the companies or assets mentioned.
Please don’t take my views at face value. Instead, do your own research (DYOR), think critically, and share your perspective so we can challenge ideas, learn from one another, and arrive at better conclusions through thoughtful discussion.