Smith concentrated on the name of the app and it sprang obediently to life. Craig Bailer had been right. Once you got the hang of it, launching, manipulating, and closing individual modules was surprisingly easy.
TVMonitor glowed green as the host gave a few general numbers and asked a lengthy question about recent budget cut proposals, then took on a reddish hue when a congressman started talking about the relative value of certain changes to Medicare, suggesting they weren’t going to be as effective as he claimed. When the subject of the relatively small impact of recent defense cuts came up, it went back to green.
The concept was simple. LayerCake searched objective facts on government sites, Wikipedia, Snopes, and hundreds of others, then compared them with what was being said. If they matched, the TVMonitor icon glowed green. If not, red. Of course, when the discourse turned to something that transcended facts—abortion, for instance—then the personal values of the user would rise in importance. To the degree that facts were stated, though, they would still be vetted.
And what about those personal biases? Smith launched a subsystem of TVMonitor that split the icon. Now the right side was monitoring the brain waves of viewers who had opted in, evaluating their reaction to the debate. He focused on the pulsating colors, fascinated by how poorly coordinated they were. People’s reactions seemed completely independent of objective truth. It was something that Dresner hoped his little feedback loop would correct.
These were just the kinds of data the Merge’s political apps fed off and why politicians were adopting them at light speed. The ability to see and react to people’s perceptions as well as to see when the system was calling you a liar was an incredibly powerful tool—one that would soon be mandatory for anyone hoping for a successful career as an elected official.
As Smith saw it, the whole thing was one giant psychology experiment. Could Dresner do what he’d set out to do? Not just change the way society functioned, but the way individuals thought and reacted to the world around them? Would people trust LayerCake? Or were they more comfortable wrapped in the warm blanket of their own biases? He himself admitted to harboring a few sacred cows and he wasn’t sure how he felt about Dresner going after them—even if he was justified in doing it.
Smith took another pull from his beer, the dull throb in his head subsiding a bit as he surfed through the channels to a Fox panel on the question of the day: privacy issues raised by the Merge. Interestingly, the TVMonitor icon faded. Was it because it hadn’t had time to vet this particular show or because of the potential for bias? Was Dresner’s search system capable of attacking its creator?
“This is different,” a woman on the right said emphatically. “This isn’t just about the user opting in, it’s about whether the people around that user have. How do we know that the facial recognition software isn’t recording everyone we see? What if I’m walking by when you go into a strip club? Or light up a joint? Will that be calculated into your value as a human being? What if I’m shopping next to a woman buying a pregnancy test? Is that information going to be immediately packaged and sold to her health insurer, BabyGear.com, and potential employers? And why not take it one step further? Why not just correlate everything someone sees to how they feel about it using their brain waves? I’m guessing that’s something Target and Amazon would be interested in.”