Welcome!
Good morning, everybody. I was on WTAG this morning with Jim Polito who is back from his convalescence. We got into a discussion about AI and how it is not as trustworthy as people might think.
For more tech tips, news, and updates, visit – CraigPeterson.com.
—
Automated Machine Generated Transcript:
Craig Peterson: [00:00:00] Good morning, everybody. Craig Peterson here. I was on with the now recovered. Mr. Polito. Yes, indeed. He was in the hospital, this whole COVID thing. Oh, nasty. We got into some depth here on facial recognition. It is gotten pretty bad here when governments using it and misusing it. There are people in jail that just shouldn’t be there. It’s just not so nice.
We talked about that and a little bit about the weather in Canada. So here we go with Mr. Polito.
Jim Polito: [00:00:35] You know what? Let’s put it all aside. Let’s not even play his intro cause he requires no introduction.
I’m talking about our tech talk guru and great friend Craig Peterson. Good morning. Sir
Craig Peterson: [00:00:50] Hey, good morning, Jim.
Jim Polito: [00:00:52] You got to get to facial recognition, but I got to tell you something that Tommy B mentioned. Of course, you being a Canadian, you understand the jet stream and how the jet stream dips from Canada in the winter. That brings us the cool weather. He said first of all, that the jet stream is flat across the country, with no troughs.
He said Canada is warmer than usual right now. Could you define what warmer than usual and what it means for Canada to be warmer than usual in January?
Craig Peterson: [00:01:26] I imported an article on Montreal and which is not that far North and the different sounds snow makes in the winter. It was saying, when it hits 30 below, you get a crunch. Eskimos, I don’t know if you know it, but they have what is it like 10 or 12 words for snow, depending on what kind of snow. It is
Definitions it’s different. There’s an injun word called a Chinook. And if you are out in Alberta, which is in Western Canada, a chinook is where you get a bubble of warm air. I can remember going to school as a kid being all bundled up because it was, whatever degrees below zero.
And that. Back in the Fahrenheit days. And then as she would come in and you could see the temperature go up 50 degrees in the matter of less than an hour, but cold weather, warm weather in Canada depending on where you are. Yeah, it’s not warm.
Jim Polito: [00:02:33] It’s all relative. Tommy B has talked about the phenomenon of the Chinook and just how weird that is.
It’s almost like. A bubble of oil floating through the water or vinegar. It’s the weirdest thing. Yeah, he has, he has described it, right? Yeah. Tommy said that you actually see it coming and wow. Pretty cool. All right. Let’s get to look. We know that. Facial recognition. Technology has been an issue with race in that individuals say it miss identifies people of color.
Now we’ve been told, let’s start with the Las Vegas casinos, which were really the early adopters of this, other than the government that Hey, no facial recognition technology works. It makes certain measurements. That’s why knows who you are. You can put a beard on and glasses. It’s still going to know who you are, but you’re saying that in criminal cases and this just isn’t just race or minorities. This is everybody that facial recognition technologies had issues.
Craig Peterson: [00:03:49]I think we’re falling into a trap here, unfortunately, and it’s a very common one. People seem to think that computers somehow are better and less error-prone than people.
The bottom line is that computers are at least no better than the programmer. That becomes a real problem here because it’s people, you watch even a shy guy movie, and the computers say that and such, and therefore it must be the way to go. It’s almost like watching Fauci, right?
He might be correct when it comes specifically to one extremely narrow area, which is a problem with almost every Ph.D. I’ve ever known in my life, is just too narrow. But when you start to consider other factors you’re wrong. What’s happened now is the police department is believing the results that are coming back from these facial recognition systems.
We know that they were used for instance, out West in the peaceful demonstrations that happened out in Portland, in Seattle, when they were burning when they were demonstrating against this terrific president
Jim Polito: [00:05:06] the peaceful demonstration. I love the guy Baghdad, Bob the CNN reporter in front of the fire.
Craig Peterson: [00:05:12] This is mostly peaceful. So they were using it to try and track people. But we’ve got a real problem here. It really was exposed the last year with this company called Clearview. Do you remember these guys?
Jim Polito: [00:05:26] Yeah.
Craig Peterson: [00:05:27] What happened there is. They were scraping the internet. They were going onto the internet. They were finding people, posting pictures of themselves, their friends or families on Facebook, et cetera. They were pulling all of this into a database and then they have the computer do some analysis on it. Folks, AI-artificial intelligence is not intelligence.
Machine learning is not learning like the way we learn and the results that come out of these systems just can not be trusted. The really bad thing about this is they’re using the results. And so they’ll take a picture that they got off of the ATM or some other monitor and they’ll take that picture, they’ll plug it into the Clearview or some of these other systems. It says it’s Joe blow and they believe it. And they go and arrest Joe blow.
There are cases now where they’ve thrown Joe blow in prison or jail, I should say. Then they try and prosecute him and he’s trying to defend himself. It turns out that in some of these cases, nobody even looked at the pictures, they just trusted the computer.
Jim Polito: [00:06:37] We’re talking with our good friend, Craig Peterson, tech talk guru. Craig, you know what this kind of worries me about, but I do think it’s because there has been too much trust in the system. But there are systems now that try to duplicate the work of a, so we all know, you get an x-ray like, believe me, I was just sick. I had plenty of chest x-rays okay. A radiologist reads that x-ray goes through it and yeah, this is okay. Here’s what I see. There are just like facial recognition. There are these systems to try to duplicate the work of a radiologist and they haven’t been successful. They can be used as a backup. They can be used as a check, but you still need a human being, a man or a woman in front of that image, looking at it.
There are certain things that the computers, when you do 3d technology and there certain things with a mammogram that sometimes a computer is very effective at finding a certain pattern of blood vessels and whatever.
But I don’t know the eyes of a real person to me make the difference.
I’ll put like money into a bill counter. Trust that, but that’s about it.
Craig Peterson: [00:07:58] Yeah, that’s it. Yeah. And AI has been used a lot in medicine lately. Some of this computer stuff doesn’t make sense. You mentioned one example, counting money, right?
Another example that seems to be pretty good is finding cancerous skin cells. So it’s basically just taking a picture of your arm. You can get an app for that by the way, and it’ll check it out. Some are other types of diagnosis, certain types of cancer in different parts of the body can be detected fairly well by computer.
And again Fauci narrow areas they are actually better than the human is. But right now, I don’t know. It’s in the far 90% of the time, a person’s going to do a better job.
Will that flip, he says. The type of AI’s that we’re looking at for the next 10, 20 years, they are going to get better in certain, very narrow ways. We’re not going to be able to see a true tricorder that gives a full diagnosis for many, I think decades.
Jim Polito: [00:09:09] Well, hold on a second. You dropped a star Trek reference thinking that I would not pick up that Dr. McCoy and his tricorder that I would not pick that up. I know you did that. Just to test me. You drop that in there.
Yes. He had the little thing he opened up and then he had the little thing he would hold over you. And it basically diagnosed everything.
Craig Peterson: [00:09:35] Yeah. That’s the one I remember too. Dr. McCoy said that he wasn’t a bricklayer. He, all he understood was this one narrow part of medicine. And yet we are at pastor Fauci.
Jim Polito: [00:09:52] Yeah. Wait a minute.
Didn’t he once operate on a Vulcan, remember that one, and the ship was under attack. And I, it was either Spock’s father or something. He operated on a Vulcan. Jim his blood is green. Jim, I don’t know what to do. On that note, this has been too hard. This is fascinating.
Obviously, folks, we have a great show from our tech talker, Craig Peterson on the weekends.
But Craig, how do folks find out more about all of your great work?
Craig Peterson: [00:10:23] Oh, and this year, things are much better. I am now publishing at least weekly, a little training you can take. It’s absolutely free. In addition to getting my newsletter and finding out about the live little training I do and the bigger ones, just go to Craig Peterson, song.com.
All of the stuff you need to know. I post there right in the homepage, sign up to that email list. I am not going to harass you. Just Craig peterson.com.
Jim Polito: [00:10:52] I can assure you that he will not. And he does this segment with us out of the goodness of his heart. Craig, always a pleasure, and we will catch up with you next week.
Craig Peterson: [00:11:04] Bye-bye
thank you, Craig guy, Craig Peterson. Everybody. Yeah. Yeah. Go to Craig peterson.com. No, he doesn’t try to sell you anything. Anyway, a final word when we return. You’re listening to the Jim Pollito show. Your safe space.
—
More stories and tech updates at:
Don’t miss an episode from Craig. Subscribe and give us a rating:
Follow me on Twitter for the latest in tech at:
For questions, call or text:
855-385-5553
