Timnit Gebru, a leader and pioneer in the ethics of Artificial Intelligence at Google, was working with colleagues on a new research paper. The paper addressed some of the dangers inherent in the Google approach to AI, including data center energy consumption and the impact on marginalized groups.
AI is well known to be a crude and voracious consumer of energy, with a culture of throwing massive amounts of energy at whatever problem is faced, rather than thinking intelligently and wisely about treating the environment with respect.
AI is already noted for either ignoring or discriminating against minorities. The reason is simple. Like most tech, AI is mainly designed by male white geeks. They build a model based on male white geekdom and then scale that model for the rest of humanity. It is also true that in many cases the data the male white geeks feed and train the AI with tends to be from white males, so the AI’s DNA becomes very male, very white. Within these historical datasets are the prejudices and discriminatory practices of male, white society, so the AI swims in a sea of prejudice and discrimination.
Timnit Gebru, an Eritrean-American computer scientist, was brave and brilliant enough to challenge Google’s male white geekdom, and in response, Google did a Trump on her, firing her and then peddling lies and misinformation. This is what we expect from Google because this is how monopolies behave.
I used to be a big fan of Google. It had a search engine that put the searcher before the advertiser. They even had a motto: “Don’t be evil”. The money called and Google changed. Google used to place its ads in the right column of the search results page, but while that was good for the searcher, it was not so good for the advertiser. Google realized that they could make a lot more money if they could pass the ads off as search results by placing them first in the list of search results. Google thus began passing off its ads as search results.
Google is an advertiser and it wants your data. Google is the biggest collector of health data in the world but this is not for the good of your health. Health data creates a healthy bottom line. When push comes to shove, Google will always act in its own best interests, not yours.
It’s funny how so many go on about freedom of this and freedom of that and yet they willingly and freely hand over their most sensitive personal data to an advertising company. Because Google is an advertising company. A company that seeks cleverer and cleverer ways to manipulate you to buy things. A company that hides its true intent behind “free” services.
Nationalize AI. Make personal data a human right. Governments, you have maybe 10-20 years max before AI Google and AI Facebook get out of reach. Politicians, stop sleepwalking. This is complicated stuff but you must get your heads around it.
AI is the super-information architecture of the world we will all soon live in. Our personal data is the most powerful and precious resource in this new world.
To think that our personal data and the AI engines are being designed and controlled practically exclusively by male white advertising geeks is truly scary.