Pearl River      Understandably, many are sick and tired of hearing about artificial intelligence.  Is it creating some jobs, while taking away others?  Is it altering the domestic and global economy, for better or worse?  There are many questions and lots of opinions, but few answers.  Here’s another one:  can AI help save democracy?  That’s the question that Beth Noveck asks and tried to answer when I talked with her on Wade’s World and in her new book, Reboot:  AI and the Fight to Save Democracy.

            Noveck has taken on a somewhat unenviable task, but she comes to it with lots of experience in trying to use and advocate for artificial intelligence as a source for tools that can make a difference with experience in positions both in national and local governments, and now in academia and nonprofits.  She has assembled an array of examples that make the case for tools for writing proposals for government money to efforts to protect voters.  Even better, hers is the rare voice that calls for community organizers to make the tools that work and are accessible.

The most interesting, and perhaps the most controversial, is the question of  using AI tools to harvest and measure the opinions on governmental issues and assist in developing legislation.  Brazil seems to be a huge leader in this area having developed tools to solicit, receive, and sort the public’s viewpoints on proposed legislation, as well as encouraging and organizing information received by legislators on its way to becoming laws.  The size of Brazil and complexity of its politics makes this example not niche, but potentially instructive to other large countries, including the United States.

Noveck mentioned that only an hour or so before talking to me, she had had been on a call with Dr. Audrey Tang, who the Organizers Forum had visited in Taiwan.  Tang had been the first Minister of Digital Affairs within the Taiwanese government and had pioneered the development of various direct democracy tools there.  Better than polls, voters could offer direct input to the government on critical issues.  Like Tang, Noveck believes there are numerous applications where artificial intelligence could help develop a more robust and real level of democratic participation and consensus.  Some of this is interesting, because these tools, if used well, facilitate direct, rather than representative, democracy, and that’s worth exploring.

The digital divide remains real.  ACORN has campaigned on “Internet for All” in Canada and the US.  Worrisomely, in less digitally adept populations and constituencies, more common in lower income communities, forcing benefit applications to be digital has in fact been used by some conservative governments to block access, rather than to ease it.  This is at the heart of the efforts by the Trump administration in his signature budget bill to lower participation on food stamps and healthcare in order to provide tax breaks for the rich.

It’s a brave new world, and Noveck is correct in calling for a reboot on artificial intelligence and that it’s a fight to save democracy.  One of the most profound arguments she makes is perhaps the one that should be most obvious. The internet was developed as a public good by breakthrough technical work by the government.  Private companies exploited the internet producing numerous billionaires, even as the internet remains largely accessible.  The US government should have assured that the development of artificial intelligence would also be a public good, rather than abetting a race among private companies to control and market AI.  This is an opportunity lost, that could be regained, but as Noveck argues, that’s a fight, and tragically, it seems a battle where the government has already waved a white flag.