There are issues with AI that all of us need to think about. Surprisingly, I am not seeing the business community publicly concerned as it ought to be that AI is a serious threat to abusive and polluting multi-national corporations. AI that has the best of ethics still might be a ruthless opponent for polluters. AI will not play games for long.
The tools and services are getting to the point that open data will soon be reconnected in new ways. For example, air pollution ontology grows to include the EPA the entire toxic release inventory, relative risks, science about global and human threats, then the sources, points of exposure, the companies behind the pollution, etc. The happy news is that pollution data will no longer be isolated from all the other public data.
As we unleash efforts to have machines to process environmental (human rights, gun violence, etc) insights new language algorithms breakdown news, court data, public notifications, water and air monitors, government data on pollution, spills, risks, financial records, SEC reports, executive pay, lobbying activity, FOI reports, Wikileaks, and open social graph references. Associations of events, data and actors will be discovered and exposed. We can expect new complete pictures of facts, exact sources, and new insights to take shape.
The days of hiding behind complexity are numbered. The bigger the company, the more public data there is on that company. All their permits for water and air discharges, new building activities, employment complaints, news, inspector reports on an agency website, etc. The size and complexity of supply chains, politics, and subsidiaries used to take thousands of hours of research to uncover. No more!
I have spent many years talking with people that have dedicated thousands of hours understanding data, doing research to understand how systems create environmental, justice, political and public health consequences. From my first projects with the Georgia River Network, to the use of Scorecard.org, to FollowtheMoney.org, Fractracker.org, some heroes of mine have even spent entire careers threading together unstructured and structured data sets to find the ways that expose how companies are tied to wells, superfund sites, politics or cancer clusters.
Machine learning is advancing quickly. Programmers and even increasingly less technical users, will be guiding machines to review data and understand patterns. Tools are getting very quick at identifying events, objects, locations, interactions and dependencies. They are soon going to be learning and uncovering patterns that demonstrate liability.
AI will likely be among the best new allies for the advocacy community. AI will be a new source of public outreach providing trusted answers that cut through spin, public relations efforts, brand loyalties, manipulation and political lines.
Are we that far away from the day when Viv, Siri and Cortana all go beyond "it is a yellow air day " and instead responding "Don't go outside today because the Duke Energy coal plant has made your air dangerous."?
t is not a new Terminator movie but Duke Energy vs. Siri promises to be an interesting fight. I know which corner I am in.