These are interesting times to be working with data – from the issues that Facebook has seen with sharing information without consent and their ongoing response to that whole thing to the breaches and other illicit accesses to information. From data trust issues to IT trust issues in a much broader sense. There have been a number of things happening that point to a very far-reaching change in how people are perceiving information.
Not the least of these is the fact that people have been leaving Facebook in droves over the data issues… NOT. It’s incredible to see that the general population is interested in the outrage associated with the use of personal information without consent, but when it comes down to brass tacks, people are not interested in giving up their apps that are doing the dastardly deeds. What’s more, they don’t seem to want to leave services, but they are trusting them less at the same time.
You can see a huge change in so many different areas.
New York Times CEO Mark Thompson gave a keynote presentation that talked about fake news, and how, in his view, relying on automated procedures to suss out fake news (or even biased news) may prove to be quite a problem. Not a problem to figure it out, but a problem to free speech and free press. It comes down to possibly having automated routines (AI) to review posts and such and take action on them to block, or limit access. This plays into data a bit (more in a minute) because it’s using an automated process to pull information that is “better” or “worse” and I think could easily be setting the stage for data filtering requirements that are both automated and potentially take away from data just because people don’t like the results.
Walmart is opting out of a database maintained to catch fraudulent transactions. Why? In short, “their data is better” – they want to use their own internal systems because they like (and trust) their databases more than the standardized or more managed systems. All of these are my terms (managed, better, etc.) but they are not interested in participating in the general system, but rather want their own, out of data trust or completeness issues.
There is the concern, too, that when AI is depended upon to deliver the “real” information, trust decreases. Some of this is ignorance, some is worry about who built/trained the AI and some is just “that’s new technology, ew!” -ism. From a brief summary of Salesforce’s CEO Marc Benioff earning’s call:
Benioff also warned that when AI technologies are indistinguishable from humans, trust will also be an issue.
But it points to the data trust issue once again.
There are a bunch of things that are colliding here. First, there are issues of trust around data security. Second, issues of trust around how information is presented. Last, there are issues of trust around the generation of information from data. All of this is coming together, sort of like watching a slow-motion train-wreck. You want to look away, everything in you says it won’t really happen, that somehow the accident will be avoided.
I suspect the fact is that instead of assuming all of this will NOT come home to roost on the data industry is just silly. “Fake News” will become “Fake Data” when people don’t like it. If it talks to them in ways they like, it’ll be spot-on. If not, they’ll ignore it. If told that it’s a problem, they’ll ignore it. If it backs up their own requirements and needs, well, it’ll be gospel.
This is a problem. We’ve got to get to a point where we are able to provide real information, confidence, tools, analysis, and protection for information and data. This is at a stage where we can drive this process a bit. I don’t know how or what is needing to be done (it’s a really big, complex thing) but I think we have to be careful to take the time and effort to figure it out. Otherwise, we run the risk of sidelining so much of the information and tools we could be and should be providing.