Jasmine Farrell’s Favorite Algorand NFT Projects, & the Vision for ZestBloom 2.0

Algorand Editors
Algorand
Published in
8 min readAug 11, 2022

Below, Jasmine Farrell, the Co-Founder and CEO of ZestBloom, shares more about ZestBloom–including its main features, planned improvements, and vision for ZestBloom 2.0–followed by a few of her picks for the most interesting and promising NFT projects being built on Algorand.

What is ZestBloom?

ZestBloom is a development company focused on the applications of AI and blockchain. Both technologies have disruptive potential, but are difficult to integrate. The Zestbloom team and I are focused on pioneering infrastructure to train and run useful neural networks with on chain components.”

While that might not sound like it’s related to NFTs, the practical applications being explored by us definitely are. ZestBloom is one of a small handful of companies focused on the intersection of smart contracts, blockchain data processing, and deep learning.

After raising over $1.5 million in a seed round we’ve developed a patent-pending NFT search engine. In addition, we’re also developing an IPFS caching service, creator tools for minting and airdropping, a white label marketplace, and embeddable sales tools.

Our ultimate goal is to offer a suite of solutions to give artists more control and fewer responsibilities.

ZestBloom Key Features

Authentication: Every NFT managed by ZestBloom will have a content-based digital signature that prevents others from stealing any part of the work. Nodes constantly index the Algorand blockchain looking for misuse and copycats, which are robustly detected even if only a small section of the work is stolen. Cleaning out copycats concentrates attention and bids on only original content.

Search Engine: Most NFT marketplaces exist independently and have no incentive to list works on other sites, except for aggregators like OpenSea. However, these aggregators are cluttered and difficult to search because they only use broad categories (art, music, etc.), and only the title/description terms are indexed. Content-based image search is a solved problem allowing users to navigate similar images by content, much like modern image search engines. ZestBloom brings all of the functionalities of Web2 search into the Web3 metaverse.

Royalties, collaborations, and resales: Any royalty and resale system requires an enforcement mechanism to prevent attempts to bypass fees or siphon rewards from original artists. NFTs can be stolen, trivially edited, reuploaded with a different description and title, and appear in different collectors’ search results to siphon bids. Furthermore, bad actors can fake bids on these pirated works to confuse real collectors into purchasing derivative works. With content-based authentication and search, ZestBloom enables enforceable protection of creators and safe collaboration.

Fractionalization: Auctions limit liquidity for artists and price out the average collector for high-quality art — Fractionalized NFTs (fNFTs) significantly lower the barrier to entry for the masses. High gas fees on Ethereum significantly decrease the utility of fractionalization because fees often cost more than the fNFT. Because transaction fees on Algorand are negligible, fractions can trade cost-effectively for pennies. Artists often don’t know how to price their work, and an ambitious reserve price may result in no bids. Fractionalization offers an alternative sales model with more accessible prices. Thus, having multiple owners of one NFT also reduces the risk for the artist because it increases the likelihood of return with more potential buyers.

ZestBloom 2.0

We are relaunching the ZestBloom marketplace in August with collaborative smart contracts from up to 10 creators at once, royalties and resales, and basic auction/bidding functions. Decentralized content authentication and fractionalization are the next priorities — planned to launch in early 2023. Our aim is to directly represent artists and their works, requiring as little effort from the creators and collectors as possible. Rather than copying the expensive and exclusive marketplaces on Ethereum, ZestBloom continues to advance NFT technology by adding intrinsic value to Algorand’s NFT ecosystem.

Vision

We began our journey by building an NFT marketplace, including all of the necessary web3 components from smart contracts to on chain data pipelines. We believe blockchain technology has democratizing potential to reduce the competitive advantages of data exclusivity and will allow independent groups to train increasingly useful models on a collaborative, decentralized database. However, this potential cannot be realized without significant development focused on solving the challenges of integrating modern AI models with on-chain data sources. We plan to continue innovating integration solutions through two main strategies:

1. Web3 services and applications of AI

2. Cross chain data aggregation and processing infrastructure

To achieve our first goal, we’ve built an NFT search engine that uses state-of-the-art models to provide industry-leading text and image search. Smart text search dramatically improves user experience by allowing users to more quickly find a particular NFT or collection based on a natural language description without an exact collection or trait name. Image search allows users to identify the original asset ID and chain of origin from a screengrab or real-world image.

The robustness of our model allows users to identify NFTs they may see in advertisements, clothing, or posters based only on a picture with a smartphone. Furthermore, the model can be used to determine NFT originality; the model correctly identifies near-duplicate images, videos, and gifs (with >99% accuracy) on cropped, rotated, or adversarially altered content. Both types of searches run in milliseconds, even with millions of NFTs in the database.

In addition to search and copyright enforcement, our model can help recommend and curate media to create a seamless browsing experience for users. Our image search algorithm has the ability to recommend similar NFTs from other collections, flag inappropriate content, and even appraise NFTs when combined with sales activity.

Our second goal will build on the solutions we’ve already implemented in the development of our search engine. Our cross-chain data aggregation and processing represent basic infrastructure will dramatically accelerate development of AI-based services that are necessary to compete with web2 platforms.

The first infrastructure service we plan to offer is an IPFS API for marketplaces and explorers. Pulling content directly from IPFS is slow and costly; we solve this problem by caching content from IPFS as it is minted. Further, we can automatically provide useful metadata about the content using our labeling model to categorize media types (e.g. photography vs animation), flag copies, moderate inappropriate content, and add trait information. These additional services increase the utility of our caching infrastructure beyond just access to the content.

Our dynamic database of NFT content and sales activity is also useful to other companies that want to train new models. We plan to turn our full pipeline into a service that provides data for models to continuously improve as new NFTs are minted and added to our database. Although our models currently function as oracles, we believe that there is significant utility to executing smaller models directly on chain so that smart contracts can utilize more complex computations without relying on off chain components. No company has yet successfully demonstrated a completely on-chain, useful neural network because of the technical challenges we are currently solving.

If you’re interested in testing out the NFT search engine it’s live here

My Favorite NFT Projects on Algorand:

Alien Tourism D4 from Delta Crew: Jared Blitz, Keegs Aldridge, and Derek Weir aka D4

This is an interesting project because it’s a few artists that came together to create Delta Crew. Each has their own style and unique approach to their work but all bring something unique to the table. D4 is the artist behind Alien Tourism creating comical designs with funky wardrobe and colorful backgrounds totaling 4000 pieces filled with special edition and collaborative spinoffs — all non-binary. Jared and Keegs run the show with marketing and getting exposure for the project. Their first launch of Alien Tourism was with ZestBloom’s shuffle created as an iframe to be displayed on their website instead of a NFT marketplace. It was a huge success and we anticipate more artists taking more ownership of their works by embedding them onto their personal sites.

Tim Ung

Tim is one of a kind architect designer that builds galleries using Mona and mints awe inspiring designs of plans on ZestBloom. I believe that he is one of the few architect professionals in Algorand that has crossed over into the NFT realm with his stunning talent. His roadmap includes 500 generative NFTs with a chance to win one of his metaverse galleries he’s created in Mona.

Peter Wu+

Peter Wu+’s genesis NFT series Creatures of Prometheus is part of an evolving project extending the artist’s concentrated examination of our relationship to technology, evolution, and otherness.

Peter Wu+ is exploring this question through his project EPOCH Gallery. Each exhibition at EPOCH situates works by multiple artists in atmospheric virtual environments, textured with signs of ruin and decay.

ECHOES is an experimental collaboration between EPOCH and LACMA’s Art + Technology Lab featuring past and present Art + Technology Lab grant recipients and organized by artist and EPOCH creator Peter Wu+. The exhibition’s virtual landscape is modeled after the physical excavation area of LACMA’s east campus, where the forthcoming building for the permanent collection is slated to debut, as well as neighboring locales around Wilshire Boulevard in Los Angeles. ECHOES is generously funded by the Algorand Foundation.

Kate Daudy

The Evolution Project is a collaboration between internationally acclaimed visual artist Kate Daudy, digital artist James Viggers and Kostya Novoselov (Nobel prize winning physicist). It’s a clever concept of breeding two trees to form a new evolutionary tree connecting each tree’s lineage. The visual DNA is then identified, codified and incorporated into the new NFT artworks. ZestBloom created the breeder application from the inspiration of The Evolution Project — the original breeder NFT. These tools will allow other artists to create similar pieces to that of Kate Daudy’s Evolution Project.

Here are some other cool projects you might be interested in checking out:

Briixel https://linktr.ee/Briixnft

Diego Nossa https://linktr.ee/nossacamera

Leon Tocua https://lynkfire.com/leontocua

Hypnagogist https://linktr.ee/hypnagogist

Lola Dupre https://loladupre.com/

--

--

Algorand Editors
Algorand

Algorand is building the technology to power the Future of Finance (FutureFi), the convergence of traditional and decentralized models into a unified system.