TL; DR
Full story
“Trust me, brother.”
It’s not a reference that exactly inspires confidence.
Of course, if your idiot cousin Brian tries to pitch you his idea for a dog grooming/ice cream delivery service that:
“Saves costs because you can use the same van for both companies.”
It is expected that ‘trust me bro’ will form a large part of his ‘market research’.
…but when it comes to building an artificial superintelligence, “trust me bro” is not a sufficient ethos/approach to answer the question of where you get all your training data from (whether it is truthful and whether you actually own it).
That’s the problem Space and Time (a web3 startup) wants to solve.
The idea is that by creating a decentralized, uncensorable third-party system that says “yes, this data is clean and paid for” (or vice versa), it sets a standard and incentivizes AI companies to compete with the new rating system .
For example:
Suppose the first handful of AI companies to adopt this verification have an average of 20% of their data verified by Space and Time…
That opens the door for other AI companies to jump on board, surpass these ratings and claim that their data verification “exceeds industry standards.”
Creating a race to the top, throughout the sector.
It’s a great idea! And one that must be realized and distributed at some point.
Will it be space and time that do this?
No idea!
Creating and establishing accepted industry standards (also known as “making pickups”) is difficult.