Blockchain may be nearing the peak of its hype cycle but much of the hubbub is just that — top financial firms are still just testing distributed ledger technology, trying to figure it out and understand its potential. Few understand how the world is changing like David Shrier.
David is the managing director of MIT Connection Science and the author of a new book called Trust Data: A new framework for identity and data. You can get it at Amazon: a.co/7yBpXBj
David Shrier presented at the 2016 Tradestreaming Money Conference held in NYC last November. The audio you're about to hear was part of his talk on moving beyond the blockchain hype.
Subscribe: iTunes I SoundCloud
Below are highlights, edited for clarity, from the episode.
Understanding blockchain and its potential
What is blockchain? Blockchain's a ledger, a receipt. A lot of financial services firms are essentially record keepers. An order ticket is just the front end of a ledger process. Blockchain is interesting in that it can take a lot of costs out of the complex systems we have to govern today. But we're interested in even bigger problems. One of them is the big data problem. Sensitive personal information is getting hacked frequently and the problem is only getting bigger.
On the other end of the spectrum, data is more valuable when it's shared. Wouldn't it be great if 30 of the largest trading partners could see their aggregate risk exposure to an individual security without revealing any individual position? But we're not there yet. We've taken our old way of doing things in the analog world and made them digital -- this is the analog-digital gap.
[caption id="attachment_12736" align="aligncenter" width="649"]
MIT's Dave Shrier presented at the 2016 Tradestreaming Money Conference[/caption]
Digital identity in an analog world
Our current expensive model of identity is still totally based on analog paradigms. Our parents attest to our birth and eventually you get a birth certificate that can be easily forged. We've had genetic testing for decades but we still rely on this analog attestation model.
Our existing models are centralized, easily forged, and costly to manage. The US Treasury would love to see more access to the unbanked and underbanked while banks complain that regulation is too expensive to service these people. The mobile phone and behavioral biometrics are really hard to duplicate. If you extract all a user's patterns of behavior using a phone (like what angle they hold it at or who they call frequently), you could extract a digital fingerprint that could be thousands of times more secure than a password.
OPAL Enigma
The Trust Data system we're proposing addresses the problem that we need better systems as we move from an analog world to a digital one. While we want to share more data, we need to tighten the privacy around it. MIT's OPAL/Enigma is one of these systems.
OPAL stands for open algorithms and it's a very powerful idea. With OPAL, you bring the code to the data rather than bringing the data to the code. Instead of centralizing data, this is more secure because you can leave the data where it's stored and machine learning allows questions and answers to come in and out securely.
In terms of security, Enigma would need over 1000 private keys or passwords to be able to hack into it. Enigma's data is distributed and encrypted but you can still make useful sense out of it while it's encrypted.
MIT's Dave Shrier presented at the 2016 Tradestreaming Money Conference[/caption]
Digital identity in an analog world
Our current expensive model of identity is still totally based on analog paradigms. Our parents attest to our birth and eventually you get a birth certificate that can be easily forged. We've had genetic testing for decades but we still rely on this analog attestation model.
Our existing models are centralized, easily forged, and costly to manage. The US Treasury would love to see more access to the unbanked and underbanked while banks complain that regulation is too expensive to service these people. The mobile phone and behavioral biometrics are really hard to duplicate. If you extract all a user's patterns of behavior using a phone (like what angle they hold it at or who they call frequently), you could extract a digital fingerprint that could be thousands of times more secure than a password.
OPAL Enigma
The Trust Data system we're proposing addresses the problem that we need better systems as we move from an analog world to a digital one. While we want to share more data, we need to tighten the privacy around it. MIT's OPAL/Enigma is one of these systems.
OPAL stands for open algorithms and it's a very powerful idea. With OPAL, you bring the code to the data rather than bringing the data to the code. Instead of centralizing data, this is more secure because you can leave the data where it's stored and machine learning allows questions and answers to come in and out securely.
In terms of security, Enigma would need over 1000 private keys or passwords to be able to hack into it. Enigma's data is distributed and encrypted but you can still make useful sense out of it while it's encrypted.