As AI proliferates and things on the web are easier to govern, there’s a necessity greater than ever to be certain that data and types are verifiable, said Scott Dykstra, CTO and co-founder of Space and Time, on TechCrunch’s Chain Response podcast.

“To not get too cryptographically religious here, but we saw that throughout the FTX collapse,” Dykstra said. “We had a corporation that had some brand trust, like I had my personal life savings in FTX. I trusted them as a brand.”

However the now-defunct crypto exchange FTX was manipulating its books internally and misleading investors. Dykstra sees that as akin to creating a question to a database for financial records, but manipulating it inside their very own database.

And this transcends beyond FTX, into other industries, too. “There’s an incentive for financial institutions to want to govern their records … so we see it on a regular basis and it becomes more problematic,” Dykstra said.

But what’s the most effective solution to this? Dykstra thinks the reply is thru verification of information and zero-knowledge proofs (ZK proofs), that are cryptographic actions used to prove something a couple of piece of knowledge — without revealing the origin data itself.

“It has loads to do with whether there’s an incentive for bad actors to want to govern things,” Dykstra said. Anytime there’s a better incentive, where people would want to govern data, prices, the books, funds or more, ZK proofs will be used to confirm and retrieve the information.

At a high level, ZK proofs work by having two parties, the prover and the verifier, that confirm a press release is true without conveying any information greater than whether it’s correct. For instance, if I desired to know whether someone’s credit rating was above 700, if there’s one in place, a ZK proof — prover — can confirm that to the verifier, without actually disclosing the precise number.

Space and Time goals to be that verifiable computing layer for web3 by indexing data each off-chain and on-chain, but Dykstra sees it expanding beyond the industry and into others. Because it stands, the startup has indexed from major blockchains like Ethereum, Bitcoin, Polygon, Sui, Avalanche, Sei and Aptos and is adding support for more chains to power the longer term of AI and blockchain technology.

Dykstra’s most up-to-date concern is that AI data isn’t really verifiable. “I’m pretty concerned that we’re probably not efficiently ever going to have the ability to confirm that an LLM was executed accurately.”

There are teams today which are working on solving that issue by constructing ZK proofs for machine learning or large language models (LLMs), but it might take years to attempt to create that, Dykstra said. Which means that the model operator can tamper with the system or LLM to do things which are problematic.

There must be a “decentralized, but globally, at all times available database” that will be created through blockchains, Dykstra said. “Everyone must access it, it might’t be a monopoly.”

For instance, in a hypothetical scenario, Dykstra said OpenAI itself can’t be the proprietor of a database of a journal, for which journalists are creating content. As a substitute, it must be something that’s owned by the community and operated by the community in a way that’s available and uncensorable. “It must be decentralized, it’s going to must be on-chain, there’s no way around it,” Dykstra said.

Connect with us:

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