So often we find that after a mass shooting or an act of terrorism, the post-mortem analysts discover previously unrecognized hints in social media. The nightclub shooting in Thousand Oaks, the mass shooting in Las Vegas, and the recent shooting at two Muslim mosques in New Zealand are examples of terrible violence, resulting in innocent victims being mauled and killed, but the perpetrators telegraphed their intentions on social media.
We know after 9-11, there were so many clues, not necessarily entirely on social media, but in the behavior of the perpetrators. But the clues were ignored, much less integrated. And we see the same neglect in social media. Of course, part of the problem is the sheer volume of postings and interaction on social media. Another challenge is how quickly can we find potentially significant items on social media, how soon can we analyze them, and then how expeditiously can authorities act to pursue leads?
Much of my own work has been inspired by our failings and missed opportunities to anticipate, inoculate against, and pre-empt this terrible violence. That’s why I have devoted much of my time to creating a way to do all this in real time. But to do so requires the creative use of artificial intelligence, coupled with massive amounts of computer power.