Jun 13, 2018 09:07 UTC
Sep 21, 2018 at 11:41 UTC
CogX ’18: Blockchain Plus AI. What Could Be Done Today?
Today, blockchain platforms are trying to get to the point where they can work reliably for human-driven use-cases, without AI being involved at all. When they’re more mature and stable, they could look to integrate AI into bits and pieces.
In the words of Joshua Gans, author of ‘The Disruption Dilemma’, how one analyses a breakthrough technology is by considering what it reduces the cost of. AI reduces the cost of prediction. Blockchain, just a little more ‘boring’, reduces the cost of verification. The breakthrough is that you can timestamp a particular piece of generated data and have it verified.
The way convergence is being developed, a lot of people wonder whether this will disrupt and break down the data monopoly that has been built by organisations like Google and Facebook and Twitter.
Samantha Stein of Startup Battlefield says “If you could be paid for your data instead of other people using that service for free, you of course would choose that, because the cost that you would end up paying for using a service versus how they’re monetising off of your data, it would be a greater win for you. The convergence of AI and Blockchain will take us there.”
If user data about our interests that is already being collected on us could be collected by us through specific algorithms, then instead of them having the power to monetise it, we’ll have the decision making power to show our data and monetise it for third-party use.
Comparing it to the current data breach incident of Facebook and its privacy policies, Joshua Gans said “Facebook’s problem is not a data problem. It is a network problem.” What we need is identity portability between platforms. Everyone owns a personal identity and everyone must have the provision of deciding whether to disclose that or not.
Ben Livshits of Brave Software argued that we have had data portability for a few weeks now. You can download pretty much all of your data of the past few years from Facebook.
“The question is: where do we put it?”
The problem is that there is no decentralised network or application solution at the moment which would make it particularly viable to download tons of information. Developers need to discover something that is significantly more appealing to the end user who doesn’t ultimately care whether the framework down the line is centralised or not, as long they have the features they require.
AI is hungry for data. People are very innovative in finding ways to screw up the system regardless. What this convergence will do is lower the friction in a lot of the developments and hence lead to business and consumer value.