would be a national dashboard. A dashboard, a user interface that you could put on your phone or your computer in which the people in charge would say these are our priorities and here's the priority in one sentence. And if you click on this link you'll see all the things we're doing about it. But more importantly for every priority we'll be tracking metrics. And if we can't track it maybe we shouldn't do it. But for everything we say we're doing we're going to show you the dashboard and show you the history and the trend. So we're not just going to talk about inflation. We're not going to just talk about gas prices. Every citizen will have it right on their dashboard and they can just — oh gas prices, there's the trend. Okay, trend looks good, etc. That's transparency. Because right now the citizens don't have the control of their government that they should because we just don't feel like we know what's going on and what's working and what isn't. Without a dashboard you don't know if the government is going in the right direction or the wrong direction. So that's transparency.

All right. Instead of draining the swamp, which I thought was a good starting thing but it doesn't mean anything. What exactly does that mean? It means whatever you want it to mean, which is why it worked in the first place. But then when you observe that no swamps appeared to be drained in his first term — you observed that, right? I personally saw no swamp getting drained. But it's also such a general thing that you don't know if it happened or didn't happen. Like would you recognize it if you saw it? I don't know. Instead of draining the swamp you should say we'll put it on the dashboard. I'll put on your dashboard who gets what funding from whom. Because right now if you as a citizen wanted to find out who are the top donors to whatever I guess you could find it but you'd have to go look for it. It should be right on the dashboard. If you see Congress vote for a bill, don't you want to say oh this bill is about let's say energy. Wouldn't you want to say all right they voted this way and the bill passed and then here's all the people who voted for it and here's how much money they got from the energy committees or the energy lobbies, right? That's transparency. Now we might not change our opinions because of that because if it turned out that Republicans are taking more money from big oil companies you would say to yourself well actually that makes sense. Like the oil companies are not going to fund the Democrats if the Democrats want to put them out of business. But you should have that information. You should know if people are voting in lockstep with their funding, right?

So you could easily imagine a Trump taking all of his negatives and turning them into positives simply by using transparency as an overall theme. That's all it takes.

All right, I saw the experiment. Rob Morris was tweeting about this. They did a test where they provided mental health support to 4,000 people using GPT-3, the AI. So real humans were getting mental health support from an AI. Now the AI was not allowed to operate by itself so there's a human who would use the AI but then the human would decide if the AI's wording got to the patient just so there'd be some control. And what they learned was that initially when the person getting the advice believed it was coming from a person it seemed to work. In other words the person was getting some benefit. But as soon as they were told that it was a machine the benefit disappeared because what the person wanted — the human who was asking for some mental health advice — what they wanted was empathy. And when they realized that a computer can't provide empathy they realized they weren't getting what they wanted.

Now here's the interesting part. I don't think they exactly could measure whose mental health got solved and who didn't. So they couldn't really see that anybody improved or didn't improve. They could only ask the people if it was a good experience I guess. What else could you do? Just ask the person, talk to it, was that a good experience? Because you can't really see if they got cured or anything in a few conversations. And here's what I think. In my experience all it would take for the AI to go back to full benefit, actually more benefit than the human, all it would take is to program the AI to say "I care. I care." Now would that be a lie? Would it be a lie if a computer said I care and then acted in a compatible way to that statement?

Here's the problem. It's not that different than the way you care. It really isn't. When you care you have a physical feeling that you might associate with the caring, right? Because if you're about a stranger died on the other side of the planet your body doesn't register anything. If somebody close to you has a tragedy your body registers it and then you say that's caring. Is that caring or is that just your body having a reaction? What does caring mean? Ther