Your browser does not support the HTML5 canvas tag.

14 January, 2018

Who and what will AI serve? US and China give very different answers

Artificial Intelligence is a frighteningly powerful new tool — and weapon. Who and what will it serve? In the U.S., tech giants Facebook, Google, and Amazon and their corporate agendas; in China, the needs of the public and the economy. These two models should be thought through now.

by Jim Carey

Part 2 - Silicon Valley: A hotspot of AI development and unethical business

Elon Musk may worry about the future of AI but this is not a feeling shared by many of his colleagues in Silicon Valley. When it comes to discussing the AI research sector in the U.S., it is useful to focus on the most well-known of Musk’s fellow tech behemoths who are leading the way. There are three companies that deserve the most scrutiny: Facebook, Google, and Amazon — which, working with agencies like the CIA and NSA as well as in partnership with other private companies, now dominate the AI game.

Although these giant companies score very low in public trust, they’ve nonetheless been placed in charge of the nation’s AI, essentially by default. When it comes to trusting names like Facebook, Google, and Amazon with this kind of technology, if their current business models are any indicator, we should all be wary.

Let’s start with the example of Facebook; a company run by Mark Zuckerberg, a man who is so lacking in empathy that he has no problem doing things like running psychiatric tests on unaware customers or attempting to steal land from indigenous people.

Facebook has already launched several projects to develop AI, led by their Facebook AI Research (FAIR) program. Included in this group’s research are projects such as the collection and analysis of Facebook users’ personal photos to better understand “the popular visual concepts illustrating various cultural lifestyles.

While FAIR’s smaller projects don’t always make the news, they do on those occasions when they begin producing results that show the research has surpassed programmers’ expectations. Such was the case in July of last year when Facebook’s AI began speaking a language the humans couldn’t understand, causing that particular program to be shuttered.

Facebook also applies AI to its user data in other ways, such as the system revealed late last year that will track users’ online activity and behavior to evaluate their mental health. While that may sound altruistic to some, having Facebook leading the charge should cause some concern.

After all, Facebook is the company that admittedly ran a (borderline illegal) experiment to manipulate the mood of its users by filtering the content they saw. This gross breach of trust with its users was later justified by Facebook, whose lawyers said the experiment was all within the guidelines of the user agreement, which comes in at over 15,000 words and readers may remember as that book-length document (with links to several more specific agreements) that tech companies are fully aware nobody possibly has the time to read.

It is also important to remember that Facebook serves as a giant marketing platform that derives its value from its ability to provide advertisers with users’ personal information. While it’s known that the platform is filled with easily-purchased and highly-targeted advertisements, there is also suspicion on behalf of users that Facebook could use phone and computer microphones to listen in on conversations for further ad-targeting.

While Facebook denies these claims, the other two companies on our list are a lot more open about their spying, while also engaging in business practices similar to those of Facebook.

Facebook’s AI competitor, Google, is also advancing its programs, including one that perfectly mimics human speech. Innovations like these are the reason many experts think that Google is at the head of the pack on developing its AI.

Much like Facebook, however, Google also is guilty of collecting massive amounts of data on its users. Also, like Facebook, some of the means Google uses in this data collection rely on AI — such as the Google ‘voice search’ feature, which has been shown to record conversations even when not in use.

Google also recently launched its photography software, which they claimed would “instantly recognize faces of special interest to its owner and, when it spots those faces, takes candid pictures of them.” Many people were uncomfortable when Google first unveiled this new technology — including Elon Musk, who criticized the software on Twitter, saying it “doesn’t even *seem* innocent.

If all of this isn’t bad enough, now Google is also creating devices that utilize AI to do even more blatant and comprehensive data-gathering. In this new arsenal is the popular Google Home, an AI assistant that responds to voice commands, which has been shown to listen in on users even when not in use.

This unwanted feature of Google Home is also built into products from the last company on our list of AI giants: Amazon. Much like Google Home, Amazon Echo and it’s AI program ‘Alexa’ also listen in on users.

Also following Google’s strategy, this data is then combined with data taken from other methods, such as monitoring search histories or product purchases, and added to massive databases kept by these private tech firms. This stored data is later fed back into Amazon’s AI, which then uses the data to drive Amazon’s “algorithms for demand forecasting, product search ranking, product and deals recommendations, merchandising placements, fraud detection, translations, and much more,” according to CEO Jeff Bezos in a letter to shareholders.

If the massive storage of data and unwitting exposure to things like experiments “educating” these advanced computer learning systems for private companies isn’t enough, things get even murkier when you examine the relation between these three tech giants and the U.S. government.

Source, links:


[1] [3] [4]

No comments:

Post a Comment