I don't attribute malice actually. Sorry if I came across like that. I believe this whole shift is occurring because of systemic structures and incentives rather than individuals motivations. A rational employee is going to do the best they can at their job within the framework of their company, the economy, and the regulatory environment. The same can be said of those in the C-suites. And the same can be said of the investors. I don't believe there is any conspiracy or individual malice. I believe it is a very boring, rational outcome of the overall operating environment we've created.
In that context, I believe that it's important to prioritize "citizen" issues over "consumer" issue, and security over convenience. Google has done some awe inspiring work, much of which I believe Google could have accomplished without requiring a user to agree to a single ToS governing all Google products, without demanding access to all user data. Statistical methods have come a long way, and much smaller data sets coupled with more explicit interaction with its stakeholders/customers (out in the open, explicit conversations vs deep dives into data without notifying the sources of the data) can still enable almost everything Google is doing. The stuff that would not be enabled within that framework could still be accomplished by Google explicitly paying folks to provide data sets (paying people to read scrips or have conversations explicitly recorded for instance) for the express purpose of building up data sets to work against for deep learning related to NLP for Google's "assistant". Lord knows Google has the cash to afford such an approach. But, since Google's current corporate structure requires it to highly prioritize profits, it just takes the data from private conversations.
The argument then is, well, that's what happens when you utilize a free or subsidized service. Yes, that is what you get. But it is not the only way we could be going on about this business. We just have our priorities all mixed up, because of the systemic incentives.
In that context, I believe that it's important to prioritize "citizen" issues over "consumer" issue, and security over convenience. Google has done some awe inspiring work, much of which I believe Google could have accomplished without requiring a user to agree to a single ToS governing all Google products, without demanding access to all user data. Statistical methods have come a long way, and much smaller data sets coupled with more explicit interaction with its stakeholders/customers (out in the open, explicit conversations vs deep dives into data without notifying the sources of the data) can still enable almost everything Google is doing. The stuff that would not be enabled within that framework could still be accomplished by Google explicitly paying folks to provide data sets (paying people to read scrips or have conversations explicitly recorded for instance) for the express purpose of building up data sets to work against for deep learning related to NLP for Google's "assistant". Lord knows Google has the cash to afford such an approach. But, since Google's current corporate structure requires it to highly prioritize profits, it just takes the data from private conversations.
The argument then is, well, that's what happens when you utilize a free or subsidized service. Yes, that is what you get. But it is not the only way we could be going on about this business. We just have our priorities all mixed up, because of the systemic incentives.