Michael Arrington (TechCrunch and CrunchFund founder) who recently took part of the Y Combinator’s (an American seed accelerator) bianually demo days (where around 90 companies were launched) noticed that the new trend around companies is A.I. and that no matter the industry your company’s in, it has to have an A.I. component also.
“Every single company is an A.I. company now. Just like every company used to be a mobile company. Doesn’t matter what you do. You’re heavily involved in artificial intelligence.” said Arrington during the 24th March 2017 “Gillmor Gang” show.
Take the Google Home product, for example. The biggest concern around the artificial intelligence part of it might be the fact that data is being handed to Google nowadays too easily and more clearly than ever: through an almost all the time opened microphone in people’s homes.
All that consumer’s data is being sent to the cloud. Having in mind the idea that everything is now being sent to the cloud, even if it’s a search that you type in or a question for your personal A.I. assistant, user privacy is the most sensitive subject around A.I..
But the reason an A.I. product needs data stored into the cloud is so that it can make better sense of the user’s input, to be able to process it better and based on the analysis done on that data by the A.I. companies, deliver to the user better results. Better speech recognition, better translation, better audio and video detectors, better self-driving cars and so on.
So, we’ve moved on from asking our search engines 3-5 years ago, search-questions like “time now in New York“, “what’s my IP“, “who was Dostoevsky Fyodor” to currently asking our A.I. applications and devices instead, that sort of questions and more, through a microphone.
A good way of keeping the user safe would be the freedom of choosing what data to share but the results are proportional with how much data the user shares. Because currently, we all know that in order to have a good translating A.I. application, the user has to feed it enough data, train it enough for days so that it can actually reply good and accurate results.
And if you think about it, the user always shared data with companies through their products whether it was anonymously or by choice. This back and forth process, this exchange of information is what actually got the user to experience a better life and generate income for companies or otherwise we wouldn’t even be talking about artificial intelligent products, let alone using them today. These companies always needed the user’s data and vice-versa.
“We’re the ones that are being trained, not the computer.”
Steve Gillmor (head of the show) told a story he experienced at a highschool dinner table where an expert in languages and accents, noticed Steve’s Washington D.C. accent even though he went there only during the holidays – what the expert was actually tracing was his father’s accent who lived in Washington D.C. and went on talking that people were sort of being pushed towards specific accents from news TV anchors and weather men in those days.
But now, startups are showing up to teach people languages through their A.I. products which made Steve believe that “we’re the ones that are being trained, not the computer” and that is a very interesting and opposite point of view if you think about language teaching products, machine learning products which need the user’s input to be fed into so it can deliver better results.
Now, think about how much troubling this data cloud sharing is when the government is involved. Ever since WikiLeaks changed the game of exposing worldwide governments and their actions by showing the user how they use the law to force companies hand over all that data, the user got more aware and more careful how it uses the world wide web.
If only a language and accents expert, a person, used to be able to trace your accent, now it’s even more easier for governments to trace you based on your accent via this kind of A.I./machine learning startups. Just because there are cases when governments ask companies to hand over their users’ data.
Not everyone wants to share their online and offline habits with their government. Companies generate income based on the user’s data while governments lead do spying their citizens when demanding that data from the companies. Basic human freedoms are at stake here. And the question always remains if these governments always know or not what’s best for their citizens.