December 6, 2024
They have an audacious vision: Your Knowledge, Your AI
Sahara has been quiet for months, but just announced a massive $43M funding raise this week.
That’s a serious chunk of change, making it one of the largest funding rounds in the Crypto AI space. And a strong signal of investor interest.
They’ve secured some top-tier backers—Pantera, Polychain, and Sequoia are all in. But what really caught our eye were the advisors, including the co-founder of Nous Research and a founding member of Midjourney.
We’ve got a soft spot for teams with big ambitions, so we did what we do best: dug into everything we could find on Sahara and distilled it into a quick, 5-minute read.
Their CEO, Sean Ren, appeared on Show Me the Crypto (episode 134). We sat through the full hour, so you don’t have to.
Sahara AI is building the AI infrastructure layer to turn decentralized AI into a reality.
Their plan? Ambitious, to say the least. They’re working on:
Users contribute their personal data to build domain-specific knowledge bases, developers use that data to train and deploy AI agents, infrastructure providers host Sahara nodes, and businesses get tailored AI solutions.
But let’s be real—you don’t realize a big vision by trying to do everything everywhere all at once.
You start with a sharp product wedge and grow from there.
Sahara’s initial product is the AI Marketplace, and it’s a key pillar in their go-to-market strategy.
During discussions with their seed investors, it became clear that this marketplace is a preview of how they plan to achieve their ultimate vision. It’s their first step, even before the launch of the Sahara Chain testnet.
At the marketplace, users can complete tasks that help train AI. These tasks range from solving math problems to summarizing videos—essentially, it’s all about data labelling and annotation.
The submissions are then used to train AI models through reinforcement learning. In return, task completers are awarded points, potentially leading to tokens down the line (though that’s purely speculation for now).
Enterprise clients like MIT, Microsoft, USC, and the Motherson Group are already tapping into this decentralized data marketplace to train their AI models. It’s set to open to the public soon.
The marketplace offers a robust suite of tools, including a data toolkit (for collection, labeling, QA, and more), a distribution engine, and a project management tool.
Sahara concluded that the current challenges for AI required a new level of data labelling and annotation.
CEO Sean Ren explains:
“We tried to build our network to allow 200,000 data labellers to work on these enterprise-level tasks. These data labellers come from all over the world, mostly from Africa and APAC and they have different expertise, language, and cultural backgrounds.
Some of them had never worked on data labelling tasks before, but that’s the exact challenge. How can you make a product that looks very user-friendly, onboard first-time data labellers, and produce quality data for big enterprise clients like Microsoft, Amazon, Snapchat, etc.?”
One of the big issues in AI right now is copyright. Earlier this year, several U.S. newspapers filed a lawsuit against OpenAI for copyright infringement—and that’s just the tip of the iceberg.
When OpenAI trained ChatGPT, it didn’t acknowledge the contributions from Reddit, Wikipedia, and GitHub users—the massive data troves it used to train its model.
Sahara sees this as a fundamental flaw. They recognize that these user contributions are incredibly valuable, and it’s only fair that users benefit from them. Sound familiar? We’ve touched on this before when we wrote about Vana, the Robinhood of user data.
Sahara believes that if you contribute value to the AI process—whether through the knowledge you share in the AI marketplace or by uploading your personal data for your own AI agent—you should own it.
This is where “Provenance” comes in.
In Sahara’s context, Provenance means adding a user watermark to your data. Messages, emails, content—everything will carry your unique watermark. That way, anyone developing AI applications or training models with your data must pay you for it. Provenance also extends to the models and applications.
This way, Sahara ensures that the original contributors are recognized and compensated accordingly.
Sahara is still in its early stages, but the momentum is building. Back in May, they opened signups for their testnet and attracted 25,000 applicants.
The Sahara mainnet is set to launch sometime in Q4 of 2024. In the meantime, you can join their waitlist here.
CEO and cofounder Sean Ren has an extensive background in AI. He’s an associate professor at USC, specializing in natural language processing and AI, and has won multiple awards for research and innovation in AI.
COO and cofounder Tyler Zhou was the investment director at Binance Labs before founding Sahara.
You can explore the rest of the team’s backgrounds here.
Cheers,
Teng Yan & Joshua
This research is intended solely for educational purposes and does not constitute financial advice. It is not an endorsement to buy or sell assets or make financial decisions. Always conduct your own research and exercise caution when making investment choices.
