Wynd Network raises $3.5M from Polychain, Tribe and more for decentralized AI project
Wynd’s Grass app rewards users for helping it scrape the web and uses blockchain to verify AI models are being trained with unbiased data
Grass and Adobe modified by Blockworks
Polychain Capital and Tribe Capital led a $3.5 million seed round in the developer behind Grass, a decentralized web scraping network focused on the creation of decentralized AI models.
Bitscale, Big Brain, Advisors Anonymous, Typhon V, and Mozaik also took part in the seed round for Wynd Network. No Limit Holdings led a pre-seed round in December 2022.
Grass, developed by Wynd Network, is a browser extension app that scrapes the web with individuals’ unused internet bandwidth and rewards users in so-called Grass Points. Wynd then sells its scraped datasets to companies in need of web data, including AI developers.
The company’s model demonstrates one possible application of blockchain tech for AI development, with investors showing heightened interest in finding any overlaps between the two sectors.
Read more: New LimeWire AI-powered music platform uses Polygon, Algorand
AI models are often trained with massive amounts of online data scraped or collected from websites. ChatGPT’s base model, for instance, was trained with petabytes of data collected by Common Crawl between 2016 and 2019.
According to Chris Nguyen, Wynd Network’s chief technology officer, some web scrapers have run into issues whereby websites block data center IP addresses or falsify data to keep competitors from scraping them.
By using browsers operated by individuals, Nguyen said Grass can bypass some of these roadblocks. Wynd does not have access to users’ data, Nguyen said.
Users can earn Grass Points for running the extension and referring other users. This approach has fueled public speculation about a future token, a reflection of the appetite among decentralized protocol users today.
The Wynd team is also using blockchain to verify where its scraped data comes from.
With off-chain models, special interests could theoretically pay web scrapers to bias the model in their favor, Nguyen argued. He said Wynd can verify its scraped data is legitimate without having to put much data on-chain.
“There’s not a single blockchain in the world that could support a tiny fraction of the amount of data that is required to train an AI. But what we’re able to do is cryptographically link a data set to the request. To make the request itself, all I have to do is push a Merkle root on-chain,” Nguyen said.
Wynd is working with two open source-focused AI companies and is in talks with companies outside of AI as well, according to Nguyen.
Potential use cases that merge AI and crypto have generated buzz as of late — as well as investor activity. Ritual raised $25 million to develop decentralized AI. VC juggernaut a16z crypto’s big ideas list for 2024 highlighted AI as well.
Those sentiments were echoed by Tekin Salimi, the founder of crypto investment fund dao5.
“I think investor sentiment is generally very bullish around crypto and AI,” said Salimi. Salimi added that the recent performance of popular AI tokens like Render (RNDR) and Bittensor (TAO) — as well as the anticipation of AI emerging as a major force in the broader market — are driving investors towards startups that integrate AI and crypto.
Get the news in your inbox. Explore Blockworks newsletters:
- The Breakdown: Decoding crypto and the markets. Daily.
- 0xResearch: Alpha in your inbox. Think like an analyst.
- Empire: Crypto news and analysis to start your day.
- Forward Guidance: The intersection of crypto, macro and policy.
- The Drop: Apps, games, memes and more.
- Lightspeed: All things Solana.
- Supply Shock: Bitcoin, bitcoin, bitcoin.
