Amit Katwala
speaker
59 appearances
2 recordings
2 series
first heard Feb 2025
last heard Mar 2025
Amit Katwala’s voice in public audio — every appearance, attributed to the second.
Trend
recordings per month · last 12 monthsNo recordings in the last 12 months.Older appearances are listed below; set an alert to hear about the next one.
Appearances
However, in order to get certification from the American Polygraph Association requires a 12 week course, I think it is. So we're talking about a few thousand dollars, a few months, and then that's it.
Those ways of... I don't know if there's ever been kind of a direct comparison done with the same subjects, but a lot of other methods of detecting lies have been very thoroughly tested, but they all suffer from the same sort of fatal flaw, which is you can't be sure. And with something like lie detection, you kind of need to be sure, right?
If you are running someone in a murder case where the crime is capital punishment, then you need to be 100% sure that... they are going to be susceptible to the particular type of lie detector you're trying to use on them. And there's no evidence to suggest that everyone will display exactly the same range of physiological responses to telling a lie.
And actually, you might expect that a psychopath or a serial killer might not actually show the range of emotional responses that you would expect. So maybe those people are actually less susceptible to lie detectors than your ordinary man on the street because they already have... a kind of different range of emotional responses. And actually, that's what we're looking for with the polygraph.
Yeah, I think it's used because it's convenient. So if you can get a confession from someone using the polygraph, that's much cheaper than... having to go out and collect all the evidence and then take the case to court and all that kind of stuff. It is slowly being superseded by different forms of lie detection in some fields.
I think there's kind of new inventions that are cheaper to run than the polygraph that don't require a trained examiner and all this kind of stuff. So we might see it shifting away from the polygraph towards new forms of lie detection in the future, but it's convenient. People believe it works and, you know, it gets results easily.
Actually, a lot of the time it gets the results that the people running the test are looking for. Whether it actually gets to the truth is a kind of different question.
That's right. The way it's portrayed in film and TV is quite wrong, actually, because it's kind of portrayed as this back and forth between the investigator and the suspect. But yes, actually, the way that the test was originally designed was, yeah, these very, very slow...
yes or no questions repeated multiple times over the course of several hours the suspect gives yes yes or no responses because the the point is you want to minimize the difference between the control questions which are irrelevant to the crime and the target questions which are about the crime so you want to you want the suspect to be sitting as still as possible to reduce interference so you don't want them to be spinning out long sentences and things like that you want to
keep the differences to a minimum so that when they do lie, it shows up on the chart. Or at least that's the theory anyway.
Yeah, that's right. It's just somehow become sort of embedded in the justice system. And I think a lot of people have debunked it. It's been debunked by numerous academic studies, government reports, expert analysis on several occasions, but it just sort of refuses to go away. And I think that's probably something to do with human nature, right?
drawn to this idea that a machine can reveal the truth, right? Especially in a time when truth is so difficult to find. I think this idea that there is a machine that can do it for us is still quite compelling.
I appreciate it.
It's not just a faster version of our current computers. It's a fundamentally different device that basically takes advantage of some weird properties of quantum physics to do things that current computers can't do. So instead of using bits, which are ones and zeros, like a normal computer, a quantum computer uses what are called qubits.
So these can be one, zero, or somewhere in between, a state called superposition. Now that's quite complicated. I like to think of it in terms of flipping a coin. So if a normal computer could be heads or tails, a quantum computer can be heads, tails, or also a spinning coin where you don't know quite where it's going to land.
It means it's going to be much better at simulating the natural world, which also follows those rules, right? The natural world does not follow the rules of ones and zeros. It follows these kind of quantum rules. physics rules.
So if we can build a computer that simulates those quantum physics rules, then we're going to be much better at simulating biology, chemistry, physics, you know, designing new medicines, finding materials for better batteries, things like that.
So Microsoft have been trying to do this for almost 20 years, and they've really embarked on a very different approach to the other companies in the field that are doing this, you know, Google, Amazon, etc, IBM. And this was really like a kind of Hail Mary play.
Like when they started doing this, it was really like, no one was really sure if it would actually work or not, if it was even a real thing. So some experts are sceptical as to whether the claims that Microsoft have made are actually going to come to fruition. But certainly they've made a big splash.
Showing 41–59 of 59 · page 3 of 3
← Previous