As someone who works with computers every day (my job is as a software tester), AI is a topic that I’ve kept an eye on over the years. Initially, AI was an SF trope (HAL, SkyNet, Data on Star Trek TNG), but in recent years, there’s Siri and Alexa and all the other digital assistants.
But AI is also the algorithms that decide which Google results you see, or which posts appear at the top of your Facebook feed, but those are so invisible that no one thinks about it unless there’s a news story about conservatives complaining that Facebook filters them out of people’s news feeds. Also, AI often is filtering out applications to jobs, which is even more problematic. Last year, Amazon pulled their AI filters because it was filtering out women, since the training data for the AI was based on past discriminatory hiring. If your past involves hiring mostly white men, the AI is going to start filtering out non-whites and women.
This last is one of the things that leans into this new book about AI, the companies actively working on the code, both in North America and China, and where it is going.
The Big Nine, by Amy Webb, is divided into three distinct sections.
1) The companies (and governments) working on AI. Primarily, three companies in China and six companies in North America. Ms Webb goes into the history of their work and where they are heading.
2) Issues that need to be fixed in this development. This includes the mono-cultures within the companies (their employees mainly come from the same sorts of backgrounds and schools, with a lack of training outside of coding). She points out the problems with datasets built by people who don’t think of diversity, and the need for programmers to consider things outside of their schooling, like ethics for example. The results of this ranges from the innocuous (photo tagging AIs suggesting ‘gorilla’ for a photo of an African-American face) to the problematic (digital assistants that can’t understand accents). And in the future, will someone be saying, like Ian Malcom in Jurassic Park, ‘Your scientists were so preoccupied with whether they could, they didn’t stop to think if they should’
3) In the third sections Ms Webb indulges in a thought experiment, presents three futures for AI that are ‘good’, ‘continuing on the same path’ and ‘negative’.
I will say, though, her ‘good’ scenario was to me ‘not as bad as it could be’. It was still a little unnerving as a vision of the future. But mind you, I am a software tester who doesn’t even have a cell phone, so take that for what it may be. It left me wanting to say ‘get off my lawn!’.
Thanks to the publisher and Netgalley for letting me read this