The Chicago-based company november, llc, has developed an artificial-intelligence program that could alter the way people view politics—and beyond. Thanks to information gathering, the algorithm could customize the types of political ads people receive, based on their own habits and interests. Windy City Times spoke with Dane Tidwell, the openly gay president of november, llc, about this potentially life-changing system.
Windy City Times: For our readers, could you talk about the program?
Dane Tidwell: We [ at november, llc ] developed an artificial-intelligence system, Helios that will take massive amounts of data and use it to build more unique profiles about people based on information about them: what they know, where they shop. It could change a lot of different areas; what we're applying to is politics.
WCT: So this would be a customized reach-out to voters?
Dane Tidwell: The system would be a customized reach-out to anyone for any purpose. Seventy-five percent of your mail is junk. The artificial-intelligence algorithm would allow companies to customize what you're receiving.
WCT: And how would you get this information?
Dane Tidwell: There are public sources. When you go to PetSmart [ for example ] , you use the key fob to get the discount. All of that information is stored. We can actually purchase that information, although different companies have different policies and make a lot more information available than others.
WCT: Now I'm playing devil's advocate...
Dane Tidwell: That's OK. Most people already think it's creepy. [ Laughs ]
WCT: That's what I was going to say. A lot of people already feel that Big Brother is watching them. I know a lot of people who would want others to know their credit scores, for example.
Dane Tidwell: It's just general information, it's not actual. I don't know that your credit score in 700. I do know the range you fall within—poor, medium, excellent. That's the information the credit companies sell.
WCT: So what do you say to people who find all of this creepy?
Dane Tidwell: The information is already public, anyway. The only difference is that we're applying a process to it so that it might be something you're interested in.
WCT: Another thing regarding customization is that sometimes people are cut off from items they could be interested in, but are unaware of, for the moment.
Dane Tidwell: I think that with the more data we pull in, the less likely that that is. Right now, we're focusing on voting and turning red states into blue. For the last few years, we've slowly been integrating digital into general campaign operations. We're actually taking it a step further and moving most campaign operations to automation. We can reach out to these people and know this is important to [ them ] .
If I'm sitting at home, I don't have a job and I don't know how I'm going to pay my rent next month, I really don't care what [ former U.S. Senate candidate ] Alexi Giannoulias did for a union; it has no impact on me whatsoever. I think that's part of the reason that the returns on these large investments are so minute. It takes a commercial an average of 10 [ viewings ] before it makes an imprint once. When you have one of the most expensive media markets in the countries, those 10 commercials better get exactly who you want them to.
WCT: Would sexual orientation or LGBT-related interests be one of the factors taken into consideration? And how do you determine if that would interest someone?
Dane Tidwell: Where the project started was when Jacob [ Meister ] was reaching out to the LGBT community when he was running for the U.S. Senate. So, we attempted to try to identify LGBT voters by the magazines they subscribe to. This is where you don't get all the information from one source; you triangulate using eight different types of information, such as magazine subscriptions.
WCT: Because you don't want to stereotype someone.
Dane Tidwell: Right. You could look at the fact that a guy owns a home with another man, for example.
WCT: Don't people fall through the cracks?
Dane Tidwell: They can. That's why it wasn't very successful with Jacob. When you're trying to target that close a demographic, it can be difficult based on public sources of data. It's much easier to target people who are progressive, for example, because "progressive" is such a large umbrella. You can look at professional organizations they're members of, levels of education. You just have to know where to look.
WCT: You mentioned that this program could have other applications beyond voting.
Dane Tidwell: Yes. Nonprofits would use it to find different donors. Say, Greenpeace was looking for new members. It depends on the type of information they want referenced, and how a close a profile they want to build.
WCT: So when do you see this being employed on a large scale?
Dane Tidwell: There are aspects that are already out now. The timeline for full-scale deployment is 2012. We may be using parts of it for the aldermanic elections [ in 2011 ] .
