Fed Gov Today’s Post

Transcript

Jim Dempsey, welcome. It's great to talk to you. Thanks for your time today. Two factors that everybody in the ocean community is concerned about, both government and industry authenticity, reliability of information. How is that process getting better at ascertaining both of those things? What's the evolution of that look like in your view? Yeah, So one of the great things about the explosion. Big data is that there is more data to compare with what you're trying to analyze, and you can get a better sense of how. Authentic and how authoritative the data is that you've collected. As an example, at our company Sayari, we collect official records from government databases all over the world and can compare them to each other. So we start off with a very authoritative source and from there combine those records with others to get the most authoritative information we can. The word that you use there are things interesting is compare. I think there are. It doesn't seem to me that that's obvious, and yet once you say it, it makes complete sense that you have the opportunity now, both technologically and access to information wise, to be able to do that, right? The idea of entity resolution is very important in intelligence analysis generally, but especially innocent. It's very important because it gives people the opportunity to say, is this information that I've collected in one region the same as information I've collected in the other? How is it different? And why are those differences occurring? It strikes me that automation tools are important in the ability to parse all of that information, to do those comparisons that you just described and so on. Is that fair read on my part? Absolutely. Artificial intelligence, machine learning are critical tools to. Triaging this information as it comes in before it goes into entity resolution models that give people a much higher level of confidence that they're dealing with the information they intend to deal with and not misinformation. You've used that term entity resolution a couple of times. What does that mean in the context of the things that you've just described? Right. So if you collect information about an entity, it could be a business, it could be a person, it could be a plane, and you collect. Information about that same entity somewhere else. You want to be sure that the information is really about the same thing, that those two data sources are talking about the same entity. And so they are techniques that you can apply to determine that you are in fact looking at the same person, company, lane. That's the right role for automation tools. I imagine that automation tools are great to an extent, but at some point you do need to have human eyeballs on this information to determine. Whether what the automation tools have found is in fact correct is that I I think so, and especially as the world and the intelligence community start grappling with how to use large language models and and artificial intelligence, it's critical to have. Experienced analysts helping those decision engines make the right call. What do you see as the future of this space? The especially regarding authenticity and verification of data. What does this look like a year out, two years out, five years out in year? Gosh, I wish I could predict the future like that. If I did, I might be able to retire early. But it's really, I think it's going to be an evolutionary process as more and more data is used to. Build large language models to then run against existing data holdings. We will see an evolutionary. Appreciation of these tools in the in the workforce. It's great to have you here, Jim. Thank you so much, Francis. I really enjoyed talking with.