Increasing Data Transparency in the AI World: How Space and Time Startup Can Assist
In a world where AI is becoming more prevalent and online data is increasingly susceptible to manipulation, the need for data transparency and verification has never been more crucial. Scott Dykstra, CTO and co-founder of Space and Time, emphasized the importance of ensuring the verifiability of data and brands in a recent episode of TechCrunch’s Chain Reaction podcast.
The discussion was sparked by the collapse of the crypto exchange FTX, which was found to have manipulated its books internally and deceived investors. Dykstra compared this to manipulating a query to a database for financial records within the organization’s own database. He pointed out that this issue extends beyond FTX and affects various industries, noting that financial institutions often have an incentive to manipulate their records.
Dykstra believes that the best solution to this problem lies in the verification of data using zero-knowledge proofs (ZK proofs), which are cryptographic actions that prove something about a piece of information without revealing the original data itself. ZK proofs can be used to verify and retrieve data whenever there is an incentive for bad actors to manipulate it, such as in financial records, prices, or books.
Explaining the concept of ZK proofs, Dykstra described how two parties, the prover and the verifier, can confirm the truth of a statement without disclosing any additional information. For example, a ZK proof could confirm whether someone’s credit score is above 700 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. The startup has already indexed major blockchains like Ethereum, Bitcoin, Polygon, Sui, Avalanche, Sei, and Aptos, with plans to add support for more chains in order to power the future of AI and blockchain technology. Dykstra envisions this technology expanding beyond the industry and benefiting other sectors as well.
One of Dykstra’s primary concerns is the lack of verifiability in AI data. He expressed worry that it may never be efficiently possible to verify if a large language model (LLM) was executed correctly. While there are teams working on building ZK proofs for machine learning or LLMs, creating such proofs can take years, leaving room for operators to tamper with the system or LLM in ways that could be problematic.
To address these challenges, Dykstra believes there needs to be a decentralized and globally accessible database created through blockchains. He emphasized that it cannot be monopolized by a single entity. For instance, he highlighted the importance of decentralized ownership and operation of a database for journalists creating content, rather than being controlled solely by a platform like OpenAI. Dykstra emphasized that this decentralized database must be readily available and uncensorable, requiring it to be on-chain.
The quest for data transparency and verification in the AI world is ongoing, with Space and Time at the forefront of this mission. By leveraging zero-knowledge proofs and blockchain technology, the startup aims to establish a verifiable computing layer that ensures the integrity of data and brands. As the need for data transparency continues to grow, initiatives like Space and Time offer hope for a more trustworthy and accountable AI ecosystem.
Connect with us:
On X, formerly known as Twitter, here.
Via email: chainreaction@techcrunch.com