5 ms·

GOOD: Its concept of separating personal data into a separate store/service/db which promotes autonomy and data portability. Its documentation. Its developer co

GOOD: Its concept of separating personal data into a separate store/service/db which promotes autonomy and data portability. Its documentation. Its developer community. BAD: it doesn't support delegation (e.g. GNAP). It isn't multi-contextual. It costs money to host or requires IT skills to self-host, making this a solution only for elites. UGLY: Its aggregation of personal data of many users creates a honeypot. It's a partial solution--it lacks a governance/legal dimension that is required because data can't be owned (privacy isn't the opposite of sharing, privacy is about trust in how data is used after being shared from a personal data store, and there's no technology solution to that).

whearyou 4y ago
Well said, one question is what’s a realistic way to fold in the legal/regulatory dimension given we’re talking with technologists and not lawyers. I wonder if there any historical parallels to inform us, and what would that look like here?
- pbtrevithick 4y ago
  Due to (i) a laissez faire public attitude, esp. in the US, that "if this was a serious problem, someone would have done something about it", (ii) a pervasive misunderstanding of how important information privacy is to democracy and society and (iii) regulatory capture by big tech it, will be a long time before there is a regulatory solution (esp. in the US). As for historical parallels, they're not promising. It was decades after the excesses by the "robber barons" and their unfettered capitalism in the late 1800s before the US government built institutions and laws to push back by creating consumer protection (child labor laws, the food and drug administration, the social security administration, etc.) As an alternative to waiting for government, some of us are exploring a combination of contract law and a personal "identity agent" (an app running on your devices that includes a personal datastore). This agent makes your data available to a service provider's apps and sites under "human information license" that states a set of reasonable, non-exploitative terms. This approach requires voluntary participation on the part of the service provider. There may be sufficient incentives for some kinds of providers to participate, but admittedly it's an uphill climb. WRT the contract notion, what I envision is similar to JLINC's information sharing contract (https://sisa.jlinc.org/v1/3b39160c2b9ae7b2ef81c3311c7924f1c4d4fa9ca47cfe7c840c9852b50d68d5 https://sisa.jlinc.org/v1/3b39160c2b9ae7b2ef81c3311c7924f1c4...)

whearyou 4y ago

Well said, one question is what’s a realistic way to fold in the legal/regulatory dimension given we’re talking with technologists and not lawyers. I wonder if there any historical parallels to inform us, and what would that look like here?

pbtrevithick 4y ago
Due to (i) a laissez faire public attitude, esp. in the US, that "if this was a serious problem, someone would have done something about it", (ii) a pervasive misunderstanding of how important information privacy is to democracy and society and (iii) regulatory capture by big tech it, will be a long time before there is a regulatory solution (esp. in the US). As for historical parallels, they're not promising. It was decades after the excesses by the "robber barons" and their unfettered capitalism in the late 1800s before the US government built institutions and laws to push back by creating consumer protection (child labor laws, the food and drug administration, the social security administration, etc.) As an alternative to waiting for government, some of us are exploring a combination of contract law and a personal "identity agent" (an app running on your devices that includes a personal datastore). This agent makes your data available to a service provider's apps and sites under "human information license" that states a set of reasonable, non-exploitative terms. This approach requires voluntary participation on the part of the service provider. There may be sufficient incentives for some kinds of providers to participate, but admittedly it's an uphill climb. WRT the contract notion, what I envision is similar to JLINC's information sharing contract (https://sisa.jlinc.org/v1/3b39160c2b9ae7b2ef81c3311c7924f1c4d4fa9ca47cfe7c840c9852b50d68d5 https://sisa.jlinc.org/v1/3b39160c2b9ae7b2ef81c3311c7924f1c4...)