Photo Credit: pexels
In the 2020 election, artificial intelligence (AI) has already surfaced as an issue, thanks to all the worries and concerns in the mainstream media about AI-powered “deep fakes”. And, in some states, AI-powered facial recognition technology has also emerged as a potentially divisive campaign issue, especially as federal authorities have warned that they may be willing to use sophisticated facial recognition technology to pick out the faces of anarchists, rioters and looters in crowds. That raises all sorts of concerns about government surveillance, and made many people worry that the “machines” are just getting too smart these days.
A new Stanford AI research project
But there are plenty of benevolent and helpful ways that AI could be used to power election politics. Take, for example, the latest research project from Stanford University called Stanford Cable TV News Analyzer. The project uses computer vision and facial recognition technology to see which faces of politicians and talking head pundits are on the screen at any time, and also uses sophisticated natural language processing to read through the transcripts and texts of cable TV news programs, in order to see what people are talking about any time. Think of the project as a cable TV-obsessed machine capable of watching multiple TV stations at once, all while taking careful notes of who’s speaking, how much airtime each issue is getting, and what ideas and policies seem to be trending.
The upshot of all this, says Stanford researchers, is that it will be possible to detect biases and trends in media reporting. It will also be possible to see which politicians seem to be breaking through the clutter, and which ones seem to be languishing in the background. And, finally, it will be possible to detect biases in how different TV networks cover certain politicians. Some of this might be obvious – such as the fact that FOX News is more likely to offer a sympathetic portrayal of Donald Trump than CNN. But some of this might be less obvious – such as the fact that careful repeating of a term like “Chinese virus” across all TV networks might actually begin to shape your views and perceptions of the COVID-19 pandemic. Six months ago, when people suggested that the pandemic might have been bioengineered in a Chinese lab, most people thought it was an outrageous conspiracy theory. But today, is that still the case?
Potential drawbacks to AI
While projects like Stanford’s Cable TV News Analyzer could lead to much more transparency in the world of politics, there are, of course, some issues to consider. Consider, for example, that the Stanford project uses facial recognition technology, and is directly integrated with the “Celebrity Recognition” API used by Amazon’s very controversial “Rekognition” technology sold to law enforcement and government officials. It’s easy to see how political operatives could co-opt this technology and use it for nefarious purposes. At the very least, it’s possible to say that the political propaganda of the future is going to be much more sophisticated than it is today. Today’s politicians will already say anything to get elected. Now, imagine them taking a quick look at your face and knowing exactly what to say to you in order to convince you to vote for them. A little spooky, eh?
The future of AI is now
It’s probably too late now to put the AI genie back into the bottle now. AI is everywhere we look today, including on social media, in the form of natural language processing, data mining and facial recognition technology. Right now, the bots control what we see in our feeds, and what we see in our search results. Let’s hope that, one day in the future, they will not control whom we vote for in an election.