Part 1 - Proposal Submission
Name of Project: Algovera
Proposal in one sentence: Unlocking more value by bringing generative art NFT algorithms on-chain
Description of the project and what problem is it solving:
Not many ML algorithms are used in Web3 at the moment. However, one current use case that has a number of active projects is generative models for art NFTs. Typically, these models are run off-chain with the outputs of the models published on-chain as NFTs. This is sub-optimal in terms of value generation. By publishing a generative art algorithm model on the Ocean marketplace, NFT enthusiasts could also buy or provide liquidity for the algorithm tokens (as well as buying the outputs of the models). This project will provide a Proof of Concept (PoC) for bringing generative models on-chain, and show artists that more value can be unlocked by publishing their models (and datasets) on Ocean.
Grant Deliverables:
- [ ] Publish 1X generative art algorithm on the marketplace and distribute tokens amongst contributors
- [ ] Start a DAO (e.g. on DAOhaus) to govern development of the algorithm, distribution of tokens etc.
- [ ] Documentation on stack of DAO tooling to be used to coordinate and govern algorithm development in a decentralized working group
- [ ] Host 1X vote on Snapshot among algorithm token holders
- [ ] Interview 2X artists about bringing their existing generative models or datasets on-chain
- [ ] Continue running weekly data science hacks (where we go through the whole process of training a ML model using new Web3 tools, publish it and share ownership)
- [ ] Continue running weekly community calls introducing data scientists to the idea of decentralized AI
- [ ] Publish blog post on bringing generative art models on-chain with Ocean
- [âś“] New Website (full rebuild)
- [âś“] New Discord (full rebuild)
- [âś“] New Notion (full rebuild)
- [âś“] Pitch deck
Which category best describes your project?
- Build / improve applications or integrations to Ocean
Which Fundamental Metric best describes your project?
- Total Value Locked
What is the final product?:
A generative art algorithm developed, owned and governed by a community of contributors. A PoC for artists to unlock more value by publishing their models (and datasets) on Ocean.
How does this project drive value to the “fundamental metric” (listed above) and the overall Ocean ecosystem?
The project will increase adoption of the Ocean protocol platform by NFT enthusiasts and artists. Opensea has 16,000 unique wallet addresses interacting with its smart contracts [https://dappradar.com/ethereum/marketplaces/opensea]. We assume that 10% of these (1,600 users) have purchased NFTs created by generative ML models. We assume that we can attract 5% of this market (80 users) to purchase NFTs created by generative models on Ocean. As well as buying the outputs of the models, If 25% of these users (20 individuals) also choose to invest in Ocean tokens to either invest in or provide liquidity for algorithm tokens (with an average investment of 2000 $OCEAN), this results in a Total Value Locked (TVL) of 16,000 * 10% * 5% * 25% * 2000 $OCEAN = 40,000 $OCEAN demand. A conversion rate of $1.10/$OCEAN gives $36,363 TVL. We assume a probability of success of 60%.
ROI = $36,363 / $20,000 * 0.6 = 1.1
Funding Requested: $20k USD
Proposal Wallet Address: 0x81e80B838e04644a93e4F3C701038889d3a1b95e
Have you previously received an OceanDAO Grant (Y/N)? Y
Team Website: www.algovera.ai
Twitter Handle: @richardblythman
Discord Handle: richardblythman | Algovera.ai#3425
Project lead Contact Email: richardblythman@gmail.com
Country of Residence: Ireland
Part 2 - Team
Core Team
Richard Blythman
- Role: Data Science and AI lead & Community Manager
- Relevant Credentials (e.g.):
- Background/Experience:
- Video Intelligence Researcher at Huawei Technologies
- Research Fellow (Computer Science), Trinity College Dublin
- Machine Learning R&D Engineer at FotoNation, Xperi
- Kernel Fellow, Gitcoin
Hithesh Shaji
- Role: Business Strategy Lead & Full Stack Developer
- Relevant Credentials (e.g.):
- Background/Experience:
- MSc Computer Science, University of Bath
- Consensys Blockchain Developer Bootcamp (Completion December 2021)
- Kernel Fellow, Gitcoin
- UG Cert Innovation and Entrepreneurship
Keaton Kirkpatrick
- Role: Operations Lead & Full Stack Developer
- Relevant Credentials:
- Github: https://github.com/keatonkirk
- LinkedIn: https://www.linkedin.com/in/keaton-kirkpatrick-63b327100
- Background experience:
- Merchant Account Operations Team Lead at ActBlue
- Consensys Blockchain Developer Bootcamp (Completion December 2021)
- Udemy: Ruby certified developer, Blockchain Development A-Z
- Member of Boston Blockchain Association
Part 3 - Proposal Details
Project Deliverables - Category:
IF: Build / improve applications or integration to Ocean, then:
Code for training generative art models on private datasets will be made available on Github. Algorithm will be published on the Ocean marketplace.
If the project includes community engagement:
- Running the campaign on social media for how many weeks?
We will continue to run several weekly sessions: (i) data science hacking sessions, (ii) community discussions on decentralization of AI. These sessions are recorded, uploaded to our YouTube channel and publicised on Twitter and LinkedIn.
Project Deliverables - Roadmap
- Any prior work completed thus far?
Website (and Notion): We performed a full rebuild of our website including design and content. You can check it out here. We also created a Notion with further information.
Discord: We performed a full rebuild of our Discord and opened it to the public (54 members in 3 weeks). We welcome members from the Ocean community to join here. Similar to Raid Guild, members can now choose from a list of 16 developer and non-developer roles.
Data science in Web3 hacking sessions: We held our first hacking session last week (weekly on Thursday) where the aim is to go through the whole process of developing a useful decentralized AI algorithm owned by a community of individuals rather than one centralized party. A group of 6 individuals attended the first session including an artist with experience in generative models and a data scientist with no prior Web3 experience. You can watch the recording on our YouTube channel here. The workflow and practices for data scientists are quite different e.g. using new private AI technologies. Hacking sessions may include (1) crowdsourcing additional data, (2) developing new model architectures and training procedures, (3) training on private datasets, (4) using cloud compute, and (5) creating a web app to demo it. Another major component of the hacking sessions is to experiment with Web3 tools to coordinate and govern decentralized groups with no one leader. This may include creating a DAO (e.g. on DAOhaus) for the working group with contributors as members, (2) creating a shared treasury (Gnosis safe) to be spent as the DAO sees fit (based on voting), (3) minting and distributing algorithm tokens controlled by the DAO. A major part of the hacking sessions is to onboard AI enthusiasts to Web3. Just using Web3 tools does a lot of the heavy lifting. With the above idea newcomers would create and use digital wallets, use a shared treasury, take part in voting, and more.
Community discussions: We have held two community discussions (weekly on Tuesdays) on the decentralization of AI, whether it’s a good idea and how Web3 might help. These discussions were also opened up to the Kernel community. Groups of 13 and 7 individuals attended these sessions. You can watch the most recent recording on our YouTube channel here. Our third discussion asks how we might coordinate and govern AI algorithm development in our decentralized working group. Some initial ideas: (1) Create a DAO on DAOhaus for the working group with contributors as members, (2) Algovera could put $1000 in a shared treasury to be spend as the DAO sees fit (based on voting), (3) Algorithm tokens that we mint and distribute will be controlled by the DAO
VisioTherapy: VisioTherapy has evolved into Algovera. Our previous startup in the Web2 space involved collecting data and developing machine learning (ML) algorithms for motion analysis in physiotherapy applications, and our first idea was to build something similar but using Web3 values. However, we realised that: Not only are there no motion analysis ML algorithms in the Web3 space, but there are few of any ML algorithms. Hence, it made sense to focus less on one specific ML algorithm and more generally about the practice and workflow of ML in Web3. VisioTherapy is being continued as an Active Project within Algovera. More broadly, our motion analysis algorithms under development can be applied to other projects in the Web3 space such as HeatDAO, who currently use expensive motion capture systems to turn dancers’ motions into NFTs.
Onshore Ocean: We also worked on all of the Onshore Ocean projects to date, but parted ways as a result of being accepted into the KB4 accelerator. We very much continue to build on the foundations established, especially related to using the ocean.py library and running algorithms using C2D. The tutorial notebooks will continue to be used and improved alongside our hacking sessions to onboard data scientists into Web3.
Pitch Deck (and Kernel Block 4): We are halfway through the Kernel Block 4 accelerator. We just presented our pitch deck to a number of investors at the halfway expo week. We are continuing conversations with investors. You can check out our pitch deck here.
Team Additions (and full-time crypto!): We welcomed Keaton this month as an Operations Lead and Full Stack Developer. Richard also finished up in Huawei to go full time on Algovera. Hithesh has been full-time since VisioTherapy.
- What is the project roadmap? That is: what are key milestones, and the target date for each milestone.
Oct 11: Launch Discord
Oct 12: Launch Website and Notion
Oct 19: Community Discussion: How might we decentralize AI Part 1
Oct 25: Create pitch deck
Oct 26: Community Discussion: How might we decentralize AI Part 2
Oct 28: Hacking session: kickoff, and process of publishing datasets and algorithms to Ocean marketplace
Nov 2: Community Discussion: How might we coordinate and govern AI algorithm development in our decentralized working group
Nov 4: Hacking session: Start a DAO (e.g. on DAOhaus)
Nov 5: Host Snapshot vote among algorithm token holders on type of generative algorithm
Nov 8: Publish placeholder for generative art algorithm on marketplace
Nov 9: Community Discussion: Community design session using Miro
Nov 11: Hacking session: TBD
Nov 16: Community Discussion: TBD
Nov 18: Hacking session: TBD
Nov 22 - 26: Interviews with 2 artists about bringing their existing generative models or datasets on-chain
Nov 23: Community Discussion: TBD
Nov 27: Hacking session: TBD
Dec 3: Publish documentation on stack of DAO tooling to be used to coordinate and govern algorithm development in a decentralized working group
Dec 6: Push code for generative art algorithm to Github
- Please include the milestone: publish an article/tutorial explaining your project as part of the grant (eg medium, etc).
We will publish one blog post (targeted at artists) on bringing generative art models on-chain with Ocean.
Please include the team’s future plans and intentions.
- Any maintenance?
- Foreseen or possible additions?
We plan to build a web app to demonstrate our created model in future. Also, with a PoC in place for putting generative art models on-chain, we hope that we can provide a consultancy for artists to bring existing models and datasets on-chain. This would enable increased collaboration between artists using existing models and datasets.
With a growing community of contributors, we also plan to develop algorithms and apps in different domains including face anonymisation and motion analysis.
