 Okay, well, I hope you got surprises for you, things you didn't know. So this is likeable science. It's actually likeable science on Wednesday with Ethan Allen and me. He's the host guest and I'm the guest host. We're going to work this out. And we keep spotting these really interesting things that come around in the MIT newsletter and some of the others, the Flipboard newsletter on technology. There's a lot of stuff on the web that opens your eyes and makes you wonder why you didn't catch that before. So this has actually been the news a couple of times in the last day or so. So this is not a scoop or anything, but something remarkable happened with Google, not Amazon, Google, on deliveries, on drone deliveries. What happened? So they're company wings basically, or wing. It's commercial permit basically to deliver goods via drone, have goods flown in locally and to actually open a business that does this, delivers goods from businesses to homes on a commercial basis. Starting out, I guess, just in a couple of places in Virginia. They don't even really identify what suburb is of Charlottesville and Blacksburg or something operating in. But it's apparently gotten underway that they've tested this in Australia. Yeah, they did a test version of it in Australia. And they had like 3,000 drone runs. And it worked good. So they decided to approach the FAA for approval to do it in the US. And the FAA gave approval. That's the news piece. Yeah, and apparently there are restrictions. The drones are supposed to stay below 400 feet, I think. They can't fly at night. They've got to have, they can only fly when it's clear weather. You know, sooner or later there's going to be accidents. So a plane taking off or landing or coming low is going to point into a drone. Yeah, you're already about the accidents that might happen. But helicopters can't come lower than 500 feet. So this gives 100 foot margin. That's a 25% margin. So it's not likely that helicopters are going to stumble into them. Besides, a helicopter would win that battle. I just cannot envision exactly how this works because it's out of science fiction. Well, but the beautiful thing is, as I said, it's a very interesting trend now, right? We've been all shifting our buying habits and buying more and more stuff from Amazon because you can put on your computer, buy it, and it's there within a few days, right? You don't have to go out and get it from your local store and stand in line. Now all the local stores will be able to start offering that same kind of service. It's a democratization of the Amazon leverage. Yeah, you're right. So either Amazon will have to start setting up more and more headquarters in different places. Or else the local store will gain an advantage back or some ground back. I'm sure Amazon will be on Google's tail on this because they want to do it too. But it strikes me that the local store, the mom and pop store, the small store, they're going to want the best technology they can get. The fastest, the safest, all that. Where are they going to go? They're going to go to Google or Amazon, and they're going to have to pay Google or Amazon for that technology. That technology could be the drone, it could be the software, or you know what, it could be the delivery system in general. So they could lend them, say Google leases drone equipment and software to a given mom and pop and then the mom and pop run it. Or alternatively, Google can say, look, you want something droned, press the button on this page, okay, and we'll send some, we'll take the drone over right away and you load it up and we'll take care of it. So this is not as if they're giving it away. No, no, they're going to make money off it, you can bet. The nice thing is basically these drones being relatively small are emitting less pollution than would a car truck on the road go out to make that same delivery. Although arguably- The batteries, lithium or something, it takes a lot of carbon or puts a lot of toxins and that can- There are going to be some issues to deal with there, yeah. But, I mean, those batteries will get better. They're good enough now to be used commercially. You know, back a day, maybe only two, three years ago, a drone could get across the field and would run out of gas. Now apparently there's much better ways to do it. And furthermore, you got charging station kind of recharge. So the drone comes back, it flies by itself into the charging station, charges up and it's ready for the next delivery. And the owner of the store, you know, you don't have to think about it because it's ready. This is what I suspect the model will be, is that there will be a central drone service that will be the intermediate between the business and the customer. And basically that way the business doesn't have to worry about it, other than somebody has to step outside and sort of pull a package while the drone grabs it and takes it. Yeah, that's true. You know, and when you say that, it evokes in my mind. So I was saying that Google or Amazon or whatever, one of the big guys enters the field, they might, you know, lease or, you know, provide a service to the store to deliver to the customer, okay, or even to the customer. This has yet to be worked out. But what about those young, sprightly entrepreneurs who say, we will be drone central, we will be your drone delivery service. We will invest in the equipment. We will work the software. We will cause the delivery. And you and mom and pop, all you have to do is call us. We're not Amazon. We're not Google. We're just, you know, think tech drones. Yeah, it's a wild west thing out there. You have to invent everything. Yeah, and people I'm sure are thinking like that, as we speak, I'm sure there are groups thinking, and we do this for some mid-sized city to make that work. You know, it's happening. Yeah. It's really, you know, and if you're a technology nut, this is really beautiful. And it's a disruptive thing. It means all of a sudden, drones were pie in the sky. Bezos figured it out. Remember what about five years ago? Right, he said. It was pie in the sky, now it's a reality. I still can't imagine, you'll have to help me on this. So the drone is at the drone charging station, which is like on the roof of the store, maybe, or behind the store, you know, a place where you know, the drones can get charged, okay? The store loads up the thing, whatever it is, food or devices or anything, shoes, you know, a clothing, and they're strong enough to carry things, okay? And it punches in the address. Okay, now the drone is ready to go. How does the drone, it has to stay under 400 feet, okay? What does it do? Does it fly down the street? Does it, you know, fly over the cars between the buildings? What, how does it navigate? Well, I mean, I'm sure, again, it's got, it's basically got the same kind of GPS that our phones have, right? Although probably more sophisticated, but I suspect, no. I suspect it goes up, sort of, up to some altitude just shy of 400 feet, takes a straight line, boom, goes right down. So it's navigating. Why should it just in turn? I mean, that's not wasting time, you know. Well, it's watching for buildings. You know, buildings more than a phone, it wants to avoid a collision because that's the end of the drone. And probably, you know, the FAA is not going to like seeing that. Right. And then furthermore, you have a lot of drones. And there's a question in the article that came around. Right, how? It was at the end, like, how many can you have before they get crazy? But, you know, right now, even though they're not crazy, even though it's just a handful, they must have systems for avoiding other drones, right? So it's not just avoiding buildings. You avoid other drones. And if a drone is flying, I don't know, when they fly 60 miles an hour, maybe. And another one is flying at 60 miles an hour. Well, that's 120 miles an hour. The drone has to be quick enough to avoid a collision with any number of other drones. As well as birds. Birds. Flying objects, whatever they may be. So, I mean, there's so, we haven't realized it, but this must be already very sophisticated. In Australia, and now here, people have high expectations for accident avoidance. So, what this tells me is the technology has gone a long way since we last looked. And the technology is ready for prime time. Otherwise, we wouldn't have got the FAA approval. Everything's going to change. Yeah. Well, it would be interesting to see, because just a few years ago, the majority of people interviewed did not want drones basically flying near their homes. There was a majority of people really objected to that. I'm guessing that's changing, I know. You remember that YouTube clip that was going around, and this can't be five years ago, maybe three or four years ago, and it was about a guy who had a little store on the side of a fishing lake, and the fishing lake was all frozen. And he delivered beer using a drone, the drone. Go to coordinates of a fishing cabin on the frozen lake. You know, a six pack would go, and it would drop it off, and the fellow fishing would have his beer, and then the drone would go back. And the FAA didn't like that, because that wasn't licensed at the time. But now, it's different. Now that can happen. Exactly, exactly. Great stuff. Okay, all right, this one, that was great prospect. We should follow that, because you know, that you can tell that the human condition now demands it, and that commerce now demands it. So we'll be hearing a lot more about that. Exactly, exactly. Okay, let me have for a second story, Ethan. So we also are reading about another sort of amazing technology jump, the much different area. But you're in tests, right? People don't like to go in and do urine tests, right? It's no fun. It's inconvenient. You have to go wait in doctor's office, send you off into a little bathroom, da, da, da, da, you got pee in a cup. You know what, let me digress for a moment and say that if you had a specimen of some kind, body fluids, what have you, for medical evaluation, you can send them in a drone. Apparently they're actually already doing that in some areas, yeah, now. But this is an interesting thing, because they now have fairly sophisticated but simple little pee strips, you know, that will check for a number of different diseases, as well as pregnancy, the presence of certain compounds, for instance, illicit drugs or whatever in your system, and do multiple tests on a little single strip. It's a color metric business. So now somebody basically figured out that if you get a standardized background sheet and give people that, and they can just put their pee strip on their standardized background, take a picture of it, and the picture is uploaded to software that looks at it and says, yeah, yeah, we think this, it was on this kind of phone, under this kind of lighting condition, therefore this reading means, and it gives a diagnosis right away. Is the phone making the diagnosis or just sending the picture? I think it's just sending the picture to some software that lives up in the cloud somewhere that does that. So it would have to be a very accurate picture, it has to be a high-resolution picture. Right, but again, I think there's AI behind this, so it just sort of learns all the time, and it knows what brand of phone was sending the picture, and probably learns about lighting conditions, and the thing is, that because basically even in their tests now, they've gotten very high compliance rates from people who are told, you should do your analysis regularly for whatever your incipient condition is, and they don't like to do it, and they have a very low compliance rate, 20, 30% maybe if they're lucky. With this kind of business, they've been tested, they've been, little test runs gotten about 70% compliance from these patients. When it's a matter of just, I'm just at home, yeah, I can do this in two minutes, you know, no fuss, no must. Well, the jury has some questions, okay, here's some questions. You said it's a standardized P-strip, and presumably you pee on it, and then you put it down on a surface, take the camera, the camera has some enhanced, I mean the phone has some enhanced camera, it takes a really accurate picture, and then you just send that back. Maybe that could be in the app, maybe it is in the app, so you take the picture, the thing automatically sends it to your favorite laboratory, which is actually an AI laboratory, and the AI laboratory is matching it as AI does against results of other cases. Billions of other cases. Billions of other cases, and making diagnoses of that. Okay, but you said it was, and maybe there's technologies on the beginning here, but it was a multi-purpose P-strip, and it could handle a certain number of issues if you will, diseases, conditions, but it can't handle everything. There is a million things that could happen in your pee that could tell you a million things about yourself, I mean, including really an enormous percentage of all the health issues in the world could be identified in your pee, and if it isn't now, it will be. So how can you make a P-strip that has all those tests built into it? Well, that is the amazing thing, is it used to be all sort of signal purpose, and it might be one for a pregnancy test, and it might be one for this test, and one for that test, and now they'll do just tiny little bands on the same strip, basically, with different chemicals or structures, little nanoscale structures on them that will change in certain ways in the presence or absence of certain kind of indicators, you know, certain proteins, certain metabolites, so yeah, so that, I mean, that in the technology is growing very rapidly and maturing very rapidly too, that detection, that sensor in, you know. So really it's a strip that reveals the chemical composition of the pee, rather than coming up with a diagnosis about the pee, right? Right, the strip is just, it's not a smart strip, it's just looking and different bands or spots on it are picking up different aspects of what's there. Yeah, presumably the AI behind it wants everyone to be using the same strip, you know, so... Right, well, otherwise it's... Right, otherwise it's very affected, yeah. I mean, it strikes me though that if you had a bladder infection or something, this, the P-strip is not going to be, it doesn't have a strip to say, oh, this is, you know, this is for bacteria, for a bladder infection. Probably does, I suspect they do. Well, I was thinking that it would have a strip for the kinds of chemical compounds you would find in the pee that would suggest, you know, or confirm the presence of, you know, an antigen infection agent. So it wouldn't be looking for the actual microbe. No. Looking for the chemical elements of the microbe, yeah? Yeah, some signature of it, yeah. Signature, yeah, that's very much what they're doing, and the great thing is, because they can get these higher compliance rates now, they can spot incipient disease states earlier. And, you know, there are something this article said, I mean, there are something like, what, 30,000 people in the U.S., in the U.S. alone, 30 million adults are affected by sort of diabetes, or pre-diabetes conditions. Oh, diabetes, yeah. You should be a diabetic condition from your pee strip without having an invasive needle stick. Right, and then you can take corrective action before this becomes very serious. Well, maybe some of this is self-diagnosis, because, so, okay, you take, by the way, the pee strip isn't going to be bigger than your cell phone, right? Right. It's not going to be this long, it's going to be like three or four inches long. I'm going to put it on the table and take a picture. So, yeah, I mean, so you send it in to a company, which is not necessarily a diagnostic laboratory company, it's a software company, right? And it has, this is like Ancestry.com. And now it sends you back a report. Right. Well, you know, this isn't a secret between that company and the doctor. If there is a doctor at this point, may not be. Yeah, I don't know. I mean, you say it isn't a diagnostic lab and it is, in some odd way it is, though. Yeah, yeah, yeah. You know, it's odd how the lines are blurring now, right? Yeah, right, the line is blurring. It's not a fancy lab with chemical reagents and everything. Oh, that's built into the pee strip. All this is, is AI looking at it, literally comparing your result against 10 million other results and basically saying, we think there's a 97.6% chance that this is, you know, this condition and a 47.8% chance that it's another condition. And that sounds a lot like Ancestry.com, right? So now it comes back to you and says, you know, Ethan has a 97% chance you got this, right? And that, and the other thing, you know, it's going to evaluate a number of conditions and give you a percentage. And then it's going to take all those conditions and put them together and give you a, you know, like a landscape of how your health is. Okay, so at that point, it's up to you because this is between you and you up to this point. And you can say, hmm, a 97% of, I got a condition, I think I better go see my doctor. You don't have to, you can suffer or you could go see your doctor and say, look, I got the 97%, you better take a look at this. So it's a tip-off, right? Right. It's a, it's a direct, and it's probably not as scary as some of these other tests which tell you, you know, it could have false positives on them. This way it'll be more sophisticated than that. Well, I'm sure, again, there are rates of both false positives and false negatives, but they are hopefully very small. Their aim with these is always to make them as good as a standard lab test, you know. And once they do that, again, just like the previous thing, it's democratizing the whole process. You bet. It's putting this fancy technology in the reach of pretty much anyone on the smartphone. It would cost a dollar-half, you know, maybe really cheap, right? And if, you know, 300 billion people use it, it would be, you know, a dollar and a quarter. Right. But a couple of things come to mind. Number one is, you know, I mentioned Ancestry.com. Well, you've got swabs in your mouse. Have a piece of DNA. Now that's not a swab, and you send the swab in, and they have a way of getting your DNA off the swab. But if we could figure mechanically how to get that DNA from your mouse, you know, onto a piece of strip kind of device, and have your phone read that, you can have Ancestry.com with respect to not only Ancestry, but diseases. And, you know, predictions of disease right there on your phone, tell a lot more than just a piece of strip. And I am guessing there are probably a thousand labs around the world working on almost precisely that issue right now. As you speak, I'm just guessing off the top of my head, but I'm betting that, because yeah, that's, it's had such potential again to be a great diagnostic tool as well as a, has a lot of appeal to a lot of people, you know. Now I'm thinking of blood, right? And as, what do you call it, the fluid that goes with blood. Plasma. Yeah, whatever. I mean, serum, serum. Okay, you know, you could do the same kind of thing with that. You'd have to get the blood on some kind of piece strip affair. So the camera, the phone could take a picture of it and, you know, use the same kind of AI type analysis. The problem there is that to get blood, it's invasive. You're going to have to go stick yourself. Right. Which people hate to do. It's, you know, it attracts from the whole thing. And I'm wondering what you think about the possibility, A, is that if we do the P strip well enough from P, we don't need blood. Is it possible? That's a really hard medical question. We absolutely need blood to make certain diagnosis. There are more things every day that are being discovered that you can analyze from either P or sweat from cases or saliva. So non-invasive ways of getting bodily fluids. It's being found more and more. There are markers for a lot of things in all those fluids. The question is, yeah, can you, can we develop the technology to spot those markers reliably, flag them unambiguously? Yeah. And there's no reason we can't, I mean, realistically. Then of course I'm thinking of skin. You know, I mean, you could look at skin with a camera, I think, and learn a lot about the skin anyway. That's not going to necessarily tell you what's under the skin. But the skin and, you know, a few cell layers into the skin, you could do that, I think. Alternatively, you could have some kind of device that will microscopically peel off a layer of skin that you wouldn't even know it was gone because it's, you know, it's just one cell thick or something. That's, that might help you a little bit in determining, you know, things you need skin to determine. And finally, the next one that strikes me is that, you know, these vaccination machines, that's a whole new subject vaccination. It's in the press all the time. I know. Don't get me started on that. The vaccination machines where the kid comes up and you're vaccinated by like a puff highly charged air, you know, and the drug is or the vaccine is in the, is in that highly charged puff of air and it enters into his body and through his skin, it's shot, you know, between the cells in some way. It's invasive. And he really, he feels the puff of air, but not more than that. So why can't that be in reverse? Why can't that be a reverse puff of air where, where you can get a little blood out? Okay, by reversing that process. What do you think? That would be interesting. It's not invasive. Right. Right. It would be interesting to have something that could draw, could draw blood relatively painlessly and without a puncture wound. Again, no reason that couldn't be done, I suspect. I mean, probably a little harder just that you have to get a good seal and then a pretty high vacuum developed really quickly. If you did that, yeah, no reason you couldn't pop a little droplet of blood out and suck that right up. Yeah. Then there were these physical, these physical manifestations of health. I mean, physical processes, for example, respiration, temperature, blood pressure, those three things and it's taken in every medical context around the world. It's your vital signs. Now, quicken, I'd rather ocean it, ocean it across the street here downtown, invented a blanket for use in hospitals, which could take your temperatures, lie down on the blanket, it could take your temperature, and it could measure your respiration, the movement of the blanket. That was pretty good. I'm not sure how widely it's been distributed, but I thought it was a pretty good idea. Then where they got stuck, and I don't know if they have solved the problem, is on blood pressure because on blood pressure, you really have to have a cuff on in order to determine blood pressure. On the other hand, these days, you could go to any number of places. I mean, you've got many models on Amazon about blood pressure cuffs that will give you an accurate medically accepted reading of your blood pressure. And again, that technology is developing and fairly soon, you'll be able to have a little wristwatch size thing each way, and it will automatically tell you if your blood pressure is changing in some way that's potentially unhealthy, you know? Yeah, so I mean, you could have this little wee cuff, little wee cuff on your wrist, and it could tell you blood pressure, it could tell you pulse, it could tell you temperature, and it could tell you respiration, maybe. And it could be sampling your sweat all the time too. Thank you very much, because sweat is part of it. Right, so it comes your own little health monitor, basically. For 24 hours a day, seven days a week, it's sitting there monitoring your health, feeding all the state up in the cloud. On the way to the AI company. Which is feeding suggestions back about how you might want to change your diet or get more exercise. Or whether you have a disease. If you fit into an AI pattern, these various factors, you know, incidentally Jay, it comes on my phone, incidentally Jay, you're having a heart attack. I hope you do something about that really soon. But you could match that up, that little band, okay, with the pee thing. And then you could maybe take a look at the skin, or maybe the band could take a look at the skin. Yeah, exactly. Respiration and skin cell, who knows. I mean, I don't think it'd be a big problem to develop this. And also, you know, if you could get blood, that would be good. And so, all of that considered, you know pretty much what any doctor knows. When you come into his office, and the AI guys can make not only one diagnosis, but many diagnoses, and a whole health picture diagnosis, plus a prognostication of how you're doing in this life. Exactly, and really then tell you, yes, it is time for you to go see a doctor now, or no, don't worry that you're feeling bad right now. It's really nothing serious. You'll get over it in a few hours, you know. Yeah, try that and see him in lawsuits again. I'm sure this is not, you know, not yet over the legal hump. No, no, all these things have, are going to get the attorneys involved. You know, the drones are, these tests are everything, even the... Okay, we got one more item. Right. So let me ask the engineer. We got a couple of minutes more to discuss the third item. He says yes. Okay. So facial recognition. Okay. Boom, here's another one that AI plays a huge role in, right? You get a picture of your face. You compare it to a database. For instance, Hawaii has everyone, all Hawaii driving and all Hawaii residents faces in a file somewhere, you know. And so there's no reason you can't cap into that database in theory and compare the picture you just take of some person on the street and instantly find out who that is. The operative piece being the street is fair game. The street is the public thoroughfare. And just like the government can take pictures as much as it wants on the street, so you can, I think. I mean, I don't think it's a privacy issue there. Yeah, that's the thing, the law has not caught up with this issue yet and has not said is it unreasonable search and seizure or unreasonable invasion of your privacy to have your picture taken and compare it to some database and be identified, huh? I've seen some of the, you know, the work, so to speak, that recognition does. And you can have a side shot. You can have a shot only part of the face. You can have a side shot at night in bad light or too much light. You can have a side shot from the way high, you know, overhead or down below. And that stuff will still find you. It's amazing. It sort of remodels your face into an acceptable configuration to be read by AI. Yeah, and even with public databases then, you can find a match for that face. So that was the story of these people that set up some security cameras, hooked into some of these public databases, and pretty soon were able to say, ah, hey, there's Dr. So-and-So is going into and out of his office here and boom, you know. I mean, they knew this even though they had no clue as to who this was. They were able to figure out who it was. That's the privacy issue there. You know, for AI, you have to have millions and millions of faces. You know, in AI, it's not one picture you're comparing it against. You're looking for a match of many pictures. All have some resemblance to this face. And so you've got to have all those pictures. Somebody has a big picture bank somewhere. I mean, when I say big, I mean millions and millions and millions of pictures of millions of people, that's the way it works best at all. And so we're going to get that from, who has that? The person who's been collecting those pictures probably violated somebody's privacy, don't you think? Yeah. This article points out that in China, there is a surveillance camera for every seven people. So I mean, you presumably can't really walk down the street in any major Chinese city without being immediately flagged and captured. Including small children. Pretty soon stuck into a database. You know that they'll get the guest name. They'll figure out which hotel you came out of, be able to collect your registration out, all those details you had to provide at the hotel. Yeah. That's the hard part. So you get a picture of somebody in the street, no big deal. Now you're going to find out who that is so you can identify him when there's a match. And maybe hotel records where he just walked out of the hotel. But that's hard. You have to have a plenary system about everything that's happening in the neighborhood to figure it out who he is. And if you don't know who he is, all you can say is, well, he looks like this guy, but that doesn't help you get where you need to go. But now the $60, that's what I'm interested in. So some guy in New York decided he wanted to do this. He wanted to make a facial recognition system just for fun. And he wanted to do it for $60. How did he do that? I confess, I did not read that particular article. But I mean, again, these surveillance cameras have just gotten so cheap. Camera was the cheapest part. Yeah. And then again, I don't know if you have to buy into a service that searches a lot of photos, public photo databases. You may not even need software on your machine. All you need is the picture or this part picture of the person you're looking at. Again, it's sort of like the urine test we were just talking about. All you're doing is capturing some data, digitizing it properly and sending it off to somebody else to compare against a billion other results of a similar sort of thing. The AI will give you within certain parameters how close you are to making a hit on a match. Yeah. And then you have cell phones, right, who can recognize a database you give it, but can also take pictures, good pictures, of people in a crowd at a party, what have you, you know, that you don't know. And you want to know. You want to know. You want this person identified, send it in, and the AI will send you back a report on who it is. Yeah. I mean, it's your call a month or so ago, a couple of months ago, there was the issue with FaceTime, a little bug in FaceTime that FaceTime could be activated by a remote user. So your FaceTime turns on without you knowing about it. And they can FaceTime could start listening in and watching you, basically. No surprise. Yeah. I mean, so, again, think about that kind of issue with what we're talking about here. Yeah. And you begin to see the power of this information to miss you, certainly. Yeah. You certainly ought to go out and do something about this. Well, to pour fuel on the fire, when this company, whether it's a big company or a little company, is responding to you and saying, oh, that's Ethan Allen's face, you know, and here's some stuff about Ethan Allen. It's keeping a record of the fact that you ask. It's keeping a record of what it gives you, because everything is on the record, everything. And the remarkable thing about databases, I don't think people understand relational databases, all this stuff is connected. All this stuff is effectively infinite in its ability to remember and recall things. And so your whole life is a composite of all these data points on all these different things and somewhere they all come together, probably at NSA. Yeah. I mean, the power now of sophisticated software to look and search multiple databases were originally designed to pick up all kinds of different information, not related in the least to one another. The power to integrate that stuff now is growing by leaps and bounds and really interesting ways, but certainly potentially frightening. I'm sure the ACLU is paying a lot of attention to that. Meanwhile, Xi Jinping has already developed a social quotient where it knows everything you've done and it assigns a social quotient of how good a person you are and whether you should get to ride the train or the plane or qualify for a loan, whatever it is. And so it's inevitable that that's going to be bigger and better and more sophisticated. It's going to govern the way people conduct themselves. They won't have a bad quotient. But the other thing I suggest to you is that it's also inevitable that our country, okay, with all the data it collects on everybody, is going to go in kind of the same direction. It's more than a credit report. It's a whole profile of everything you've ever done, what kind of person you are. We're going there in 1984 on Spades. There's a Netflix series, The Good Place. And they do that. People are all given point values on how good your life has been, basically, how well you've lived your life. Unless you qualify it to a certain level, you don't get in. Right. It's coming that way. And gee, I mean, it's really happening while we're not watching it. And as the years go by, it'll happen faster and faster. And then we're going to find out that what happens in China happens here. And we'll have a social quotient that we won't even know about. The old Christmas song, right? Santa Claus is coming, right? When you're sleeping, he knows when you're awake, he knows when you've been better or good. So be good for goodness' sake. I mean, that's all coming. Exactly. Coming true, you know, except it's not Santa who knows. And the thing is, I feel that when we do these shows and you start singing, it's about the end of the show. That's true. It's about the end of sooner.