The AI ​​world needs more data transparency, and web3 startup Space and Time says it can help Tech Crunch

Spread the love


As AI expands and things on the Internet become easier to change, there's a greater need than ever to make sure data and brands are verifiable, said Scott Dykstra, CTO and co-founder of Space and Time, on TechCrunch's Chain Reaction podcast.

"Not to get too cryptographically religious here, but we saw that during the collapse of FTX," Dykstra said. “We have an organization that has as much brand trust as my personal life savings in FTX. I trust them as a brand.

But the now-defunct crypto exchange FTX was allegedly manipulating its books internally and misleading investors. Dykstra sees it as similar to making a query to a database for financial records, but manipulating it in their own database.

And it extends beyond FTX, to other industries as well. "There is an incentive for financial institutions to want to change their records ... so we see this all the time and it becomes more problematic," Dykstra said.

But what is the best solution for this? Dykstra thinks the answer lies in verification of data and zero-knowledge proofs (ZK proofs), which are cryptographic measures used to prove something about a piece of information — without revealing the source data.

"It has a lot to do with whether bad actors have an incentive to want to manipulate things," Dykstra said. Whenever there is a high incentive, where people want to change data, prices, books, financials or more, ZK proofs can be used to verify and retrieve data.

At a high level, ZK proofs work by having two parties, a prover and a verifier, confirm that a statement is true without providing more information than whether it is correct. For example, if I want to know if someone's credit score is above 700, a ZK proof — prover — can confirm that to the verifier without revealing the exact number.

Space and Time aims to be a verifiable computing layer for Web3 by indexing data both off-chain and on-chain, but Dykstra sees it expanding beyond the industry to others. As it stands, the startup is indexed from major blockchains like Ethereum, Bitcoin, Polygon, Sui, Avalanche, Sei and Aptos and is adding support for more chains to power the future of AI and blockchain technology.

Dykstra's recent concern is that AI data is not truly verifiable. "I'm very concerned that we can't really effectively verify that the LL.M. has been implemented properly."

Teams are currently working to solve that problem by creating ZK proofs for machine learning or large language models (LLMs), but that could take years to try and create, Dykstra said. This means that the model operator can tamper with the system or the LLM to perform problematic tasks.

There should be a "decentralized, but global, always-available database" created by blockchains, Dykstra said. "Everyone should have access to it, it shouldn't be monopolized."

For example, Dykstra said, in a hypothetical scenario, OpenAI itself wouldn't own the database of the journal where the journalists were creating the content. Rather, it should be owned by the community and maintained by the community in a way that is readily accessible and uncensorable. "It has to be decentralized, it has to be on-chain, there's no way around it," Dykstra said.

This article was inspired by an episode of TechCrunch's podcast Chain Reaction. Subscribe to Chain Reaction on Apple Podcasts, Spotify, or your favorite pod platform to hear more stories and tips from entrepreneurs building today's most innovative companies.

Connect with us:

  • On X, formerly known as Twitter, Here.
  • Via email: chainreaction@techcrunch.com





Source link

Leave a Comment