Celia Quillen (5:27)

Hi Mike. Thank you so much for having me. It's great to be here.

Mike Carruthers (5:30)

So I have a sense, because this is what I thought for a long time, I have this sense that people think of AI as kind of a really fancy search engine. Like you go to ChatGPT, you type in something, and you get results back, much like you would do with Google, even though, as you're about to explain, it isn't really a search engine. But don't you think that's what a lot of people think it is?

Celia Quillen (5:56)

Absolutely. I think many people think of it as a search engine, which, especially in the early days, might have been a detriment to them because at the time, many of these tools could not browse the web for you and do that search. Now they can. But I do think that that is kind of the first and foremost thing that people of them as.

Mike Carruthers (6:13)

Yeah, see, that's what I thought. And I've, you know, I've done that. I've typed in things and gotten answers. But then I looked in your book and thought and saw that you can do so much more. And so what much more can you do? If you could give me like a little shopping list of things that AI can do that people may not realize. I don't want to go deep, I just want to go wide.

Celia Quillen (6:37)

Brainstorming. It's excellent at coming up with new ideas and helping you kind of creatively come up with new suggestions for how to accomplish tasks or how to name something. It can help you create and generate effectively anything that is text based. So if you needed to create a quick guide to a place that you're traveling, instead of searching one thing, you can search many things at once effectively by having it create a guide. You could have it craft a shopping list according to your groceries that you have to accomplish for the week. You could turn it into your personal secretary, articulating out all the things you have to keep in mind and remember and have it jot it down and summarize it in a nice concise way for you to refer back to later. There is no limit to how it can help you really anything that is text based, it effectively acts as an intern, as a super advanced intern that can do anything for you. Another use case that many people enjoy is meal planning. So you can put in your preferences for what your dietary preferences are, what you want to eat, how many meals you need prepped for what your calorie intake you roughly want for each meal is. And it can generate a list of ideas, recipes, grocery lists from there, and really help you outsource a lot of the menial tasks that take a lot of energy and mind space in your day to day.

Mike Carruthers (8:02)

One of the things that's so amazing to me about it that I discovered after reading your book is that it isn't just that you ask it to do something and it does it and that's it. You can refine and say, no, that's not what I was looking for. Or here's a document. Read this and tell me now. Now tell me something that you didn't know before. I mean, it's so malleable. It's like clay. You can make it do almost anything.

Celia Quillen (8:30)

Absolutely. And it is always better when you collaborate with it as well. You can refine it to make it more personalized, you can give it feedback. It's almost like a co collaborator. And it really is. It's a magical tool when you first use it, because realizing that it can unlock this, it's like having a second intelligence in the room with you.

Mike Carruthers (8:51)

What is it doing when it's doing all these things? And I know that's a big question and I'm asking for a little answer, but give it a shot.

Celia Quillen (9:01)

Yeah. Key thing to understand. Unlike a search engine, these tools are not going and browsing a massive database of information, plucking out a paragraph or two and then pasting them back to you. They're not doing what a Google search might, where it's browsing all of the URLs on the web and finding the content that most matches your search. Instead, it is a predictive tool. So it's predicting the best possible answer word by word, or more technically, token by token. So fraction of a word to give you an answer based off of the words you put into it. It's basically like a very, very advanced autocorrect. How it's able to do this is these tools were trained off of a massive amount of data. So think about the contents of the entire Internet. For example, the computer is able to run a lot of complex calculations and notice different patterns within language. So it basically became very good at mimicking human language patterns, assigns weights to different words. So for example, it might know that the word kitten is more frequently associated with cat than puppy. And over time, all those different things informs the model that allows us to interact with these tools, the large language model. But all you need to know is effectively it's not retrieving information, it is predicting the best possible answer word by word, based off of this massive training model that it has developed.

Mike Carruthers (10:31)

There are different AI tools. ChatGPT is the one that, you know, I think a lot of us that comes to mind immediately is the one to go to. But. But there are others. What are the differences between them? Are they more or less interchangeable or not?

Celia Quillen (10:48)

Certainly I find them to be more or less interchangeable. So you have ChatGPT, kind of the leader, the most well known. But there are countless tools in this arena, if you're looking for a more general purpose AI that can do kind of everything that I mentioned before, ChatGPT, Google Gemini, Claude AI is another one. Perplexity. Microsoft copilot Meta has an AI now that also has a chat interface. All of these tools can do all of that. Text based generation. There are some subtle differences between the different tools in terms of the extra functionalities they have, but as these companies behind them are developing and trying to compete with each other, they are frequently introducing the same features and functionality that ChatGPT might have just released so they can keep up. So some examples are attachments. In the early days of ChatGPT and these other tools, many did not allow you to say, attach a PDF to your question or query. Now ChatGPT can do that. Claude AI can do that. Google Gemini, you can connect it with your Google Drive. Microsoft Copilot works in Sync with your Microsoft 365 suite, so that you can now communicate with those PDFs. Other examples, as we mentioned earlier, web browsing. So all of these different tools, with the exception of Claude that I mentioned earlier, can now actually go and browse the web, generate a response summarizing what it finds, and then link back out to the source that it got that from. So all of these tools have subtle differences, but by and large I try and encourage people to, if you have one that you like, and especially if it's free and you're just getting started, experiment, play around with it. In no time, if some competitor launches a new feature, it probably won't be long until they launch their own version of the same thing. Because these tools are evolving so quickly and they're all trying to stay competitive with each other.

Mike Carruthers (12:50)

If I were to ask an AI tool like ChatGPT give me a snappier title for this, or give me a like, how does it know what snappier means? Or does it.

Celia Quillen (13:04)

So it doesn't really know anything, I guess is the one way to put it. It's not actually intelligent, but it is artificially so. So snappier, you know, kind of how thesaurus works, really. Snappier is associated with other words like catchy, zingy, you know, all these other things. And in its TR it's associated snappy characters in a text it might have been trained off of. Maybe a character is described as snappy in their dialogue. At some point in time it starts to associate those words with different kinds of output. So it's hard to explain it exactly, but it does have a very advanced way of understanding human language patterns and human language context and meaning. So if you ask it to be snappier, it will probably develop a response back to you that is more concise, maybe a little bit sassy, maybe it has a rhyme to it, depending on what else is in your prompt. But because it has become an expert of mimicking human language patterns, it effectively is very good at understanding the meanings of words that you use. Now if you used a gibberish word that had never existed before, it's like, can you make it more smurgablat? It would have no idea what you meant, but it might come up with something that's, you know, similarly silly sounding as smurgablat.

Mike Carruthers (14:22)

You and I are getting a great primer in artificial intelligence and how to use it to make your life better. My guest is Celia Quillen. She's author of the book AI for Life. So my wife had some friends over and I asked them, have you ever shopped at or heard of quints? Because I had just heard they were going to advertise here and I didn't know much about them so I wanted to know what other people knew. And all of them except, except one I think said, oh, I love quints. Their clothes are fabulous. Well, since then I've gotten two cashmere sweaters, a couple of polo shirts, and I wish you could see them here. You can hear that's the sound of my Mongolian cashmere sweater. And their cashmere sweaters start at $50. They also have these iconic 100% leather jackets, pants for every occasion. Quince has everything for men and women and their clothes are really high quality. The best part? Quint's Items are priced 50 to 80% less than similar brands. By partnering directly with top factories, Quint's cuts the cost to the middleman and then passes the savings on to us. Quince only works with factories that use safe, ethical and responsible manufacturing practices. Look, just go check out Quint's, see how great their clothes look and how amazing the prices are and buy something. It's a great way to support this show and I know you're going to love it. Indulge in affordable luxury. Go to quints.com sysk for free shipping on your order and 365 day returns. That's Q U I N C-E.com sysk to get free shipping and 365 day returns. Quince.com sysk.

Unknown (16:11)

There are many ways to entice a potential new customer with flashy branding and calculated marketing strategies. But at the end of the day. The success of a business comes down to the ability to close the sale. And for businesses like Thrive Cosmetics, Allbirds or Skims, it's what goes on behind the scenes. Making, selling, and for shoppers, buying simple that makes the difference for them and for millions of businesses. That difference is made with Shopify. Businesses that want to grow, grow with Shopify. Because if your goal is growth, your commerce platform better be ready to sell wherever your customers are scrolling or strolling. And Shopify will make sure you're ready on the web, in your store, in their feed, and everywhere in between. Nobody does selling better than Shopify. And with shop pay, they boost conversions up to 50%, meaning way less carts going abandoned and way more sales going. It's no wonder Shopify is home of the number one checkout on the planet. Upgrade your business and get the same checkout Allbirds uses. Sign up for your $1 per month trial period at shopify.com realm. All lowercase go to shopify.com realm to upgrade your selling today. That's shopify.com realm.

Mike Carruthers (17:31)

So, Celia, how do these AI tools like ChatGPT, how do they handle things like, like if you were to ask it, come up with a clever way to say, how does it know what clever means? And does it come up with something clever? I mean, does it make an attempt to be clever?

Celia Quillen (17:51)

Yes, that's a, that's a good way to put it. It is definitely attempt to. You'll see this with the earlier models from ChatGPT. It is slowly, these tools are slowly getting better at it. But if you ask it to, for example, make something funny, make it funnier, it will have a little bit more trouble with that. It will kind of give you a response back that sounds like it could be a premise for a joke, but then it just, the punchline falls flat. And that's an important thing to understand about these tools is they don't have the same human experience that we have. They don't really know anything. They're just guessing at the best answers. So I believe at some point I asked it to tell me a knock knock joke about the dentist. And it was like, like, knock knock, who's there? Dentist. Dentist. Who? Cavity. Another. And it, you know, that doesn't make any sense to us. But it was in the structure of a joke.

Mike Carruthers (18:40)

Right, Right.

Celia Quillen (18:41)

So, so, but it, but for the most part, if there is, you know, I think humor is the one that really falls flat on. But cleverness, you know, if you ask it to create 10 clever headlines, maybe one of them you pluck it out after you read it through as a human is clever. The other nine might not be so much, but one of those ten is bound to get you a little closer to where you want to be.

Mike Carruthers (19:03)

And how do we know how accurate are these things? Like, when you ask it a question, how confident should you be that the answer you got back, assuming it's a factually based answer, is correct?

Celia Quillen (19:18)

That is a super good question. And the answer is you should be always a little skeptical with everything, especially if it's a super niche topic. Because these tools do create. They're just predicting the next best answer. That prediction could be wrong. Just kind of like how the weather prediction might be wrong for the day if you're looking at your weather app. But these tools are progressively getting better and better and trained and on more data sets which are making them less likely to what we call hallucinate mathematics. They used to be terrible at predicting the best answer to a mathematical question. Now the more advanced models, especially those that are called deep reasoning models, where they kind of, when they generate a response to you, it takes a little bit longer because it's actually running through its response, through itself over and over again to see if it's really getting to the best answer, those will be more likely to be correct. Now, when it comes to general questions about things that are likely, there's likely a lot of information about it on the Internet, for example. Therefore, it's likely trained on a lot of that data, like why is the sky blue? It's probably going to give a pretty solid answer to that. If you ask it about a very niche disease that just appeared in the world that hasn't been heavily researched, it's probably going to give you an answer, but it may not be correct. The best thing I tell people to do is that when you use these tools as a jumping off point because they can get you really far, and then do a quick Google search, see if there's a secondary source that backs that up. You'll find that a lot of the time it is totally right. It's wrong enough of the time that you should always do that extra step of fact checking.

Mike Carruthers (21:07)

If I Google, if I just search in a search engine, the best baked lasagna recipe there is, I'll get back results of 10 million lasagna recipes, none of which may or may not be the best. If I ask AI to give me the best, does it somehow put in a judgment that this one or these four are the best, or will I just get 10 million baked lasagna recipes.

Celia Quillen (21:33)

You will get one lasagna recipe back. It will not necessarily be the best lasagna. Best is arbitrary. But you will get a recipe back that is most likely to result in a lasagna. That's the best way I can put it. So if you truly want to have the vetted best ranked lasagna, lots of great ratings, Google might still be the best way to go. But if you're looking to just get started, I want to try and cook something quick. I do this all the time. How do I give me a quick recipe. I have chicken, asparagus, parmesan and tomato sauce in my fridge. What can I make with this? It will generate back a recipe that will result in something tasty. It may not be the best thing you've ever made, but it will get you there. And that is again because it's taking in all of the data it has of all the lasagna recipes that are in its training data, which is likely a massive amount because they are effectively trained on the whole variety of things on the Internet.

Mike Carruthers (22:28)

One of the first things that struck me that I noticed that makes it so different an experience than dealing with a search engine is if you ask for a baked lasagna recipe, you get one because you ask for one. So why not ask for 10 if you really want to like sort through them, but it takes you very literally. Whereas a search engine will give you as many baked lasagna recipes as it can find, seemingly. But AI gives you exactly what you asked for. You asked for one, here's one that's correct.

Celia Quillen (23:01)

You asked for one to get one. Now you could ask for five and it'll give you five different ones. I do say in the case of a recipe, if you're looking for a recipe, most of these tools have a certain, you know, text based cut off. After they generate a certain number of words, they'll finish the answer. That way they don't go on for forever and you'll have 10 pages of content to read. So if you're looking for recipes, for example, sometimes you're better off asking for one and then saying, okay, that sounds okay, but could I have another? And then it generates another one. But I like it. Honestly, instead of having to scroll through endless pages of search results and then click through them, scroll past all of the ads and all of the recipe blogs tend to have a lot of extra context up front that has nothing to do with the recipe. ChatGPT will give you exactly what you want. Nicely packaged, short and sweet. And if you don't like it, you can ask for another one and it'll generate it just about as quickly.

Mike Carruthers (23:56)

So tell me something that you've done with AI that I wouldn't have thought to do that was really, really cool.

Celia Quillen (24:05)

A couple years ago, I planned a whole trip to Puerto Rico with it. I did not know where I wanted to go. I just knew I wanted to go on a vacation. I knew what the criteria that I wanted were. My husband and I wanted to go on a nice romantic trip together, somewhere warm that we had never been before. That, you know, maybe was a little bit tropical, but was in a certain distance. I didn't want to go to Hawaii. It was too far away. And it generated a bunch. I asked it to generate, you know, 10 recommendations based off of all that criteria. Generated 10 recommendations. Then I had it do some more research on. Okay, tell me more about Puerto Rico. Once we had landed on that trip, within a basically 10 minutes, I had it make a table breaking down all the different activities that we could possibly do there based off of the price per person, based off of a description like, describe it, give me a price per person, give me where it is relative to the hotel we're thinking about staying at. And it just accomplished so much research tasks that I would have taken me hours and hours and hours on Google, popping through different travel blogs trying to find responses. So that's one thing, just everything related, travel planning and related to that. The vision capabilities on these tools, which is basically being able to take a picture and ask questions about it is huge. You can snap a picture of a menu at a restaurant and ask. I'm on this kind of diet. I'm looking for lower sodium. I'm trying to be gluten free, but I don't know what has gluten in it. It's one of those fancy menus that has menu items that I've never heard about listed on the menu. Could you describe what this dish probably is? Or go and describe in layman's term what this fancy menu means? Or translate menu entirely in French into English. It can analyze data. Now, ChatGPT has a data analysis feature, so I personally track all of my finances in a spreadsheet. I can attach it into ChatGPT with all the different inputs of how much money is in what account and have it create graphs and charts and analyze, you know, what's going up, down, what I should be considering. There are so many things it can do. I could go on forever.

Mike Carruthers (26:17)

What I have found so interesting about it is a. It is so easy to use. And the more you use it, the more you want to use it. And the more you want to use it, the more things you find it can do. It's just remarkable.

Celia Quillen (26:33)

Absolutely. Yeah. And it's really great. I mean, it is truly a collaborative tool. It is like engaging with a second person in the room. So if you're trying to sort through a problem and you're. That's another use case. If you're, you know, an interpersonal problem with a family member or friend and you just need something to bounce ideas off of and get advice from, it's surprisingly good at delivering mature, thoughtful, somewhat objective advice as long as you ask for that. And it's very, very collaborative and creative. It's just. It's an expert at mimicking how humans communicate.

Mike Carruthers (27:10)

I remember you said in the book something about some of them may tell you it can't do something and you can correct them and say, no, you do have that ability.

Celia Quillen (27:18)

That's correct. Sometimes it will say, sometimes these tools will say, I'm sorry, I can't help you with that. I don't have the ability to browse the web. But if you know it can browse the web because either you've. Because you've tried it probably in the past before, then it will say, okay, I'll go do it now. And then it'll come back to you and you'll know it did it because you'll be able to see a link that you can click through on, how.

Mike Carruthers (27:39)

Are these things and at what rate are these things like ChatGPT and the other ones getting better?

Celia Quillen (27:46)

The level of progress they've made in the past two years, two and a half years since they released is insane. So they will be able to do more and more advanced things as time progresses.

Mike Carruthers (27:57)

Well, I love how you make this so easy and so accessible because, you know, AI at first glance might seem a bit intimidating and overwhelming, but you make it so not that. So thanks. Celia Quillen has been my guest. She's an expert in artificial intelligence and she is author of the book AI for Life. 100 plus ways to use artificial intelligence to make your life easier, more productive and more fun. There's a link to her book in the show notes. Well done, Celia. Thanks.

Celia Quillen (28:29)

Thank you so much, Mike. It was a pleasure. Your data is like gold to hackers. They're selling your passwords, bank details and private messages. McAfee helps stop them. Secure VPN keeps your online activity private. AI powered text scam detector spots phishing attempts instantly. And with award winning antivirus, you get top tier hacker protection. Plus you'll get up to $2 million in identity theft coverage, all for just $39.99 for your first year. Visit McAfee.com, cancel anytime terms apply. Imagine what's possible when learning doesn't get in the way of life at Capella University. Our game changing flexpath learning format lets you set your own deadline so you can learn at a time and pace that works for you. It's an education you can tailor to your schedule. That means you don't have to put your life on hold to pursue your professional goals. Instead, enjoy learning your way and earn your degree without missing a beat. A different future is closer than you think with Capella University. Learn more at capella.

Mike Carruthers (29:31)

Edu Uncertainty is a constant. We live with every moment. We never know what's going to happen. And sometimes that's fine. We don't always want to know what's coming. But not knowing can also be problematic. I mean, some things we would really prefer to know before they happen. Nevertheless, we live with uncertainty. And joining me to help make sense out of all the uncertainty that lies before you is David Spiegelhalter. He is Emeritus professor of Statistics at the University of Cambridge and author of the book the Art of How to Navigate Chance, Ignorance, Risk and Luck. Hey David, thanks for being a guest on something you should know.

David Spiegelhalter (30:19)

No, it's a great pleasure to be here.

Mike Carruthers (30:21)

So uncertainty is just part of life. It's there. There it is. So why do you study it?

David Spiegelhalter (30:28)

Oh, I think it's fascinating. I mean, it's just part of being human. It's all around us. We don't know what's going to happen and we don't know what's happening. We don't know what's happened and we don't know why. And we live with all the time and we're happy with that. I mean, I always ask people, well, would you want to know how a football match was going to end if you'd recorded it? And would you want to know what you're going to get for Christmas? And even. And of course they say no. And even would you like to know when you're going to die? Which of course I can't tell them. But very few people actually say they would like to know. Some do, but. But very few. And so I, I always think of uncertainty as a relationship between ourselves and the outside world. Somebody's called it the conscious awareness of ignorance, which I really like because it really Puts that's something we own, our personal ignorance and that can vary so much from person to person.

Mike Carruthers (31:24)

And so why doesn't the conversation end there? Some things are uncertain. You'll never know until you know and have a nice day. That's it.

David Spiegelhalter (31:35)

Well, I'm a statistician, so, yeah, I spent my entire career, in a way, analyzing uncertainty, using data and evidence to try to reduce our uncertainty and even try to quantify our uncertainty to actually get a handle on what we don't know. And I really strongly believe that we should, when possible, try to put our uncertainty into numbers. I mean, we all use words like, like could, perhaps, likely, and everything in our everyday language, but actually those can be very dangerous. And I always refer back to 1961, when Kennedy became president and found out about the CIA planned invasion of the bear pigs at Cuba to topple Castro's government with 1500 Cuban exiles. And he commissioned the Joint Chiefs of Staff to do an intelligence report on that. And they thought it was not a very good idea at all and thought that it was about 30 to 70 chance of success. In other words, 70% chance of failure. But in the report that went to Kennedy, the numbers got taken out and it got replaced by, bizarrely, the term a fair chance of success. And they meant not very good. But I don't think that's how anybody else would interpret it. Of course, that's not the only reason he approved the raid, which of course was a complete fiasco. But people have learned from that. And now in intelligence, people really try to put numbers on their judgments. In the UK, you know, if you're in MI5 or MI6 and you use the term likely, you mean between 55 and 75% probability. And if, you know, anyone assesses a 30% chance of success, they have to use the word unlikely to communicate that number. So there's a sort of lookup table between words and numbers.

Mike Carruthers (33:19)

Somewhere in this discussion has to come human nature, because numbers don't tell you everything. For example, it is much safer to fly than it is to drive in a car. But a lot of people don't live that. They know it intellectually, but flying scares them to death. And so the probabilities don't mean much because they ignore them.

David Spiegelhalter (33:46)

Yeah, I know. I mean, this has been studied for ages. I'm not a psychologist, but, you know, I think it's well known that, you know, people do have an anxiety about, you know, losing, not having control of their lives. They feel particularly bad when you know, something risk might be imposed upon them by something else. They're particularly scared of things that they don't understand very well or have really dread results giving people cancer and so on. And that makes people anxious about certain things where actually maybe they're not the main things they should be anxious about. When people started doing this research in the 1980s, microwave ovens came very high on the risk index in terms of people's fears and anxieties, and that's gone down. There's some fashions in risk in what people are anxious about and what they are, but it is true. There's, you know, our feelings come into how we feel about risk and uncertainty all the time. We can't just wave our fingers at people and say, oh, how silly you are that you're worried about this and you're not worried about and you're not worried about something else. You know, I don't think we should do that. I think we should acknowledge people's concerns and. But try to give them the best evidence in the clearest way so that they can make an informed choice. We shouldn't be propagandizing people, trying to manipulate their feelings, either to reassure them or to frighten them. And that's why in my work, I really push this idea of trustworthy communication. Why should anyone trust scientists, experts, authorities and things if they feel they're being manipulated? And that means actually giving people information that's balanced and that does talk about the benefits and harms possibly of vaccines, for example. And, you know, perhaps doesn't. Don't use the phrase vaccines are safe and effective, but say, yeah, they're safe enough and effective enough to use in some people, in some circumstances talk about.

Mike Carruthers (35:37)

Luck because luck seems to throw a monkey wrench into everything. I mean, you know, people say, you know, it's really stupid to play the lottery. Your chances of winning are just so ridiculous. Well, tell that to the guy that won $300 million. You know, turned out to be a pretty good deal.

David Spiegelhalter (35:53)

Yeah, I like the idea of luck. I mean, I don't believe that luck is some external force that operates in our life, But I do think it's quite useful as a description looking backwards at things that happened to us that were out of our control and were unpredictable and yet had a big impact, like winning the lottery or being knocked down by cars. So, you know, these are all things that could happen to us. And philosophers have identified, in fact, the most important type of luck, which is what called constitutive luck, which is just who you're born as, actually Maybe even before that, there's existential luck, which is the luck of being born at all, because it's extraordinary that any of us are born. The chain of circumstances that are required for that. But given you are born, who you're born as in the world is incredibly important for the rest of your life. You know, who your parents are, the country, time of history, etc, essentially your genes, your, you know, your early environment or so you've got no control over these at all. And yet they're staggeringly important for the rest of your life. So that's the biggest bit of luck luck that you have just who you are. And then of course, you've got what's called circumstantial luck, which is being at the right place at the right time, meeting your wife on a train as somebody I know did, and, or, you know, being like my grandfather, being at the wrong place at the wrong time and being blown up by a shell in, in 1918 in the western Front. And then you've got outcome luck, which is just how it happens to work out for you at that particular moment. Like, you know, I know a friend of mine was on a plane that crashed. Seriously, nearly everyone on board was killed. Terrible circumstantial luck to be on that plane, but really good outcome luck because he survived. Similarly, my grandfather actually, the shell, you know, blew him up, but only not severely enough. He was taken off the front line and survived the war, otherwise I wouldn't be here.

Mike Carruthers (37:49)

Well, we're glad you are.

David Spiegelhalter (37:51)

Yeah, exactly. So am I. But I very nearly couldn't be. Just like all of us, it is amazing that any of us are here at all, really.

Mike Carruthers (38:00)

What about coincidences? Are, are they a thing or. It's just, that's just, there's just so many things going on that there's bound to be.

David Spiegelhalter (38:09)

Yeah, I, I, again, I, there's no, I don't think there's some external force. I mean, people do talk about a force like synchronicity, which Jung invented, or seriality and things. I don't believe that. I think, think they just happen because there's so many opportunities for them to happen. And, and also I think it's very. They happen because randomness, true randomness, is very clumpy indeed. You know, just because things are random, they're not equally spaced. So, you know, I know that if I've got 20 people and I ask them the last two digits of their cell phone number, there's an 87% chance that I'll get a match in that group of 20 people. And some people would think, oh, wow, what a coincidence. I think, no, it's utterly predictable that that'll happen. So you know that the clumpiness of randomness leads to coincidences occurring. And they're quite difficult to interpret. And that's why anybody who talks about a coincidence, they really need to get some expertise in there to try to evaluate that. Our human intuition about it is terrible, absolutely awful. About probability in general, we're really bad at it.

Mike Carruthers (39:18)

What are some other things like that, like what you just said about if you had 20 people in a room and they. The last two digits of their phone number, things like that, Those kind of probabilities in everyday life that people probably have no idea are true?

David Spiegelhalter (39:34)

Well, the standard one is if you have 23 people, then there's 50, 50 chance that two of them will share a birthday. And that is something you find quite difficult. So on a soccer field, there's 23 players, 11 in each team and the referee. So in half of all soccer games, there's two people on the pitch with the same birthday. And that I think people do find quite difficult to join me. And it's shown. People have shown it again and again. So it works. And you can prove it mathematically fairly easily. But it is something that doesn't correspond to our intuition. But there are other things that go the other way. That's something that we think is, whoa, that's surprising. It happens so often. The other thing that actually happens is. Is surprisingly rare is shuffling a pack of cards. Now, if I shuffle a pack of cards properly and give it a good shuffle, I could be absolutely confident, bet you all the money in the world that nobody in the entire history of humanity has ever had the cards in that order before. And then I can shuffle them again into a good shuffle. And I say, nobody in the entire history of humanity has ever had the cards in that order before. So no, and not just now. It's unbelievably unlikely that anyone ever. And no two people in the air in the whole history of humanity have ever had a pack of cards in the same order after a good shuffle. And that's. People find that really difficult believe because it's not intuitive just how many shuffles there are. And you know, you can do the sums. It's sort of 52 for the first card and 51 choices for the second card and 50 for the third card. And you multiply all those numbers up together and you get a very good number. In fact, the number is about the Same as the number of atoms in our galaxy. It's, you know, it's got 78 zeros. I think beyond it, you know, it's an unbelievably massive number. And that's not intuitive at all.

Mike Carruthers (41:33)

Wait a minute. You're saying that if you shuffle a deck of cards, the number of the different number of orders that you could come up with is a number with 78 zeros?

David Spiegelhalter (41:46)

Yep. As many atoms as there are in the whole galaxy. And so the shuffle you come up with after a good shuffle, nobody will ever have done before, you can be almost certain of it. You can't prove it, of course. It's not logically true, but statistically, it is just unbelievably unlikely.

Mike Carruthers (42:03)

Yeah, that's amazing. That's an amazing fact that.

David Spiegelhalter (42:07)

Yeah, yeah, it is quite surprising.

Mike Carruthers (42:10)

What that reminds me of is when people say, there are no two snowflakes alike. Well, how would you ever know that? How would you ever.

David Spiegelhalter (42:18)

How would you prove it? That's true. It is difficult because you can't prove. You can't look at every snowflake. I don't know every shuffle that's ever been done, but you can do the sum and you can work out just how incredibly unlikely a match would be.

Mike Carruthers (42:31)

What about the phenomenon? It often happens in weather and whatnot. It's called the perfect storm, where multiple things come together to create total havoc.

David Spiegelhalter (42:43)

Yeah. Taleb uses the frame Black Swan, which is essentially a really, completely unpredictable event that occurs, and then people find an excuse for it. Perfect Storm's different. It's just a very extreme example. Example of something fairly familiar. And, you know, the story I use is about this boat, the MV Derbyshire, which is twice the size of the Titanic and disappeared without trace and in 1974. And so they. And they did. No idea what happened. And eventually they found, about 12 years later, they found the wreck. And there was an inquiry and two statisticians, you know, who I know, took part in the analysis of that, and they'd done simulations on wave tanks and things like this, and they worked out that the boat, it was in a typhoon, went into a perfect storm and suddenly, you know, a wave they reckon, you know, could have been 25 meters high. The entire weight of it went onto the front of the boat, the front hatch, and caved it in. And then there was a chain reaction and all the. All the. The, you know, the containers in the ship just sort of exploded. It just, you know, caved in and the ship went down in seconds. They didn't even have time to send a distress call. And, and that, and that was, they used this, you know, technical stuff of extreme value theory. But it was an example of a perfect storm in which an event happened that a had never. More extreme than anything that had ever been observed before. But with the appropriate analysis you can show that it's quite plausible that it happened.

Mike Carruthers (44:25)

What did you say about what's a black swan?

David Spiegelhalter (44:28)

Oh, a black swan. That was what Talib invented. A black swan is something that, it's like, you know, people thought all swans were white and then they went to Australia and found a black swan. And so it's something that, that you never even crossed your mind. A complete unknown. Unknown in Rumsfeld's language. You know, something never even crossed your mind that it was there and then suddenly it's there. And usually, of course people then very quickly find a reason to explain why it happened. But it's a, it's a qualitatively different event from something that has happened before. And again, I think it's useful to distinguish that from a perfect storm, which is just an extreme version of something that is familiar.

Mike Carruthers (45:06)

So what do we do with all of this? I mean, knowing what you know on a day to day basis, how does this fit into your life that you do that I don't do? Because you know what, you know, I.

David Spiegelhalter (45:18)

Kind of, I got a few principles, I do, which is try to look at the possible futures. You know, try to think of what might happen, what might go wrong and without getting too anxious about it. And then take some mitigating circumstances, you know, instead of buying insurance and stuff like that. When you go traveling. Taking risks but not being reckless. That's what I always say to kids when I talk about this stuff. Take risks, but don't be reckless. In other words, protect yourself against the severe downsides of what might happen, but go for it. Because, you know, people have found that, you know, people who later are considered lucky, well, actually they've got particular personal characteristics of being bold, being taking advantage of opportunities for having a positive mindset. Mindset of going for things, acknowledging it's not always going to work out. So I, I, I kind of think that I try to do that and I, but without being, I won't get on a motorbike, for example. You know, I always think I would love it, but I'm not going to get on a motorbike, but I will go off and have other adventures and which I, I feel are okay. So I kind of try to weigh things up. Not always Using, not using numbers, but just trying to think about, you know, roughly what are the risks that I'm concerned about.

Mike Carruthers (46:33)

Do you sense people are overly worried about risk or do we have it kind of right?

David Spiegelhalter (46:39)

Oh, that's very difficult. There's huge variation, you must know, in the people. You know, you'll know some people who are very cautious, rather anxious, always concerned, you know, one in a million chance. Well, there's always the one and things like that and others who are more gung ho and go for it and kind of ignore, you know, you know, are not too sensitive, you know, are bold, possibly even reckless. And so there's enormous variation in our reaction to it. I think what's more important than the. And that's how individuals are. People are different when it comes to sort of societies, when people are making judgments on behalf of us about what really risk society is prepared to take, then I think, you know, it is interesting just how much how cautious some, you know, some policies are compared with others just because actually people are really concerned about, for example, nuclear power or nuclear waste. And so vast amounts of money is spent to possibly have negligible effect in reducing the risk because the risk is so small in the first place. Place. And so, you know, I think when it comes to societal decisions, we should be expecting people to take, I wouldn't say a rational, but a balanced approach to actually looking at the important things and not, I personally feel, not following people's anxieties.

Mike Carruthers (48:04)

Yeah, well, you know, it's funny because I think of myself as a fairly cautious person, but I also know as I look back on my life, things have a way of working out for the most part that seldom. I mean, I have not been the victim of any huge disaster. And we just had the fires here in California. We have friends that lost their home, but they'll be okay. I mean, it will eventually there will be a path to recovery and things work out.

David Spiegelhalter (48:36)

This is the important thing which I, I always emphasize, which is resilience. And the point is that, you know, we can't stop everything stuff happens, we can't stop everything bad happening to us and we're going to die anyway. So the crucial thing is resilience, which is being able to deal with stuff as it comes along of, you know, recover as best as you can. And it's resilience to things you never even thought of. I mean, you can protect yourself to some extent against things you, you have thought of. But in the end, what one needs is reserves, both, I think, mental and, and monetary if possible, but certainly mental reserves to deal with unexpected things. The illnesses are going to happen, the, the, the dramas, the things are going to happen to the people around you, your people you love. You've got to be, deal with it. That's, that is what how things go. And so I, I don't know if you can train people into resilience. Right. But it's certainly a characteristic that is, I think, the most valuable thing to have in an age of uncertainty.

Mike Carruthers (49:40)

I still can't believe that thing about shuffling cards. I mean, that is just unbelievable. But it's all about this discussion of certainty and uncertainty. David Spiegelhalter has been my guest. He is emeritus professor of statistics at the University of Cambridge and author of the book the Art of Uncertainty, how to Navigate Chance, Ignorance, Risk and Luck. And there's a link to his book at Amazon in the show notes. David, thanks, thanks for sharing.

David Spiegelhalter (50:10)

Thanks very much.

Mike Carruthers (50:14)

You would think there is one place where you would be safe from all the troubles in the world and that would be in the shower. Yet there are hazards lurking in your shower. For example, did you know showering in a thunderstorm is dangerous? It's actually dangerous to be near any plumbing or wiring inside your house when there are thunderstorms going on outside. This is according to the Lightning Injury Research Program at the University of Illinois at Chicago. Lightning can hit a power line or the ground and then come right up through your water pipes. Using an old shower head can be trouble. A lot of showerheads contain potentially dangerous bacteria, according to a study from the University of Colorado at Boulder. That's because they're full of these little nooks and crannies that provide an ideal home for microbes. It's better to use a rain type showerhead or take the showerhead off altogether and go with a single stream of water. Yeah, I don't think I'm gonna do that using a dirty loofah or washcloth or brush. They may be good for exfoliating the skin, but those things are loaded with germs unless you wash them regularly. You should soak them once a week in a diluted vinegar soak or run them through the dishwasher. And not having a bath mat is asking for trouble. According to a report from the center for Disease Control and prevention, more than 230,000 accidental injuries occur in the bathroom or shower in one year. And almost 20% of those injuries are due to slipping and falling. So it is best to have those non slip strips or a non slip bath mat in your tub. And that is something you should know. You know, podcasting is a very competitive business. We're always competing for new listeners and one of the best ways to get them is for you to recommend this podcast to people. You know, it really helps us and then they become listeners and hopefully they'll tell people and our audience continues to grow. So please share this podcast with people you know and it's a great way to show your support. I'm Mike, her brothers. Thanks for listening today to something you should know.

Celia Quillen (52:39)

I'm Amy Nicholson, the film critic for the LA Times.

Mike Carruthers (52:42)

And I'm Paul Scheer, an actor, writer and director. You might know me from the League.

Celia Quillen (52:46)

Veep, or my non eligible for Academy Award role in Twisters. We love movies and we come at them from different perspectives.

David Spiegelhalter (52:55)

Yeah, like Amy thinks that, you know, Joe Pesci was miscast in Goodfellas and.

Celia Quillen (52:59)

I don't he's too old. Let's not forget that Paul thinks that Dune 2 is overrated.

Mike Carruthers (53:04)

It is.

Celia Quillen (53:05)

Anyway, despite this, we come together to host Unspooled, a podcast where we talk about good movies, critical hits, fan favorites.

Mike Carruthers (53:12)

Must sees, and in case you missed.

Celia Quillen (53:14)

Ems, we're talking and Parasite, the Home Alone from Grease to the Dark Knight. We've done deep dives on popcorn flicks, we've talked about why Independence Day deserves a second look, and we've talked about horror movies, some that you've never even.

David Spiegelhalter (53:25)

Heard of, like Ganja and Hess.

Celia Quillen (53:26)

So if you love movies like we do, come along on our cinematic adventure. Listen to Unspooled wherever you get your podcasts, and don't forget to hit the follow button.

Mike Carruthers (53:42)

A while back we had Ramit Sethi on as a guest and he's one of the smartest people you'll ever know when it comes to everyday money matters. And he was here talking about money and couples. As it turns out, he has his own podcast called Money for Couples, which if you're part of a couple, then I highly recommend you listen to this podcast. Because when you do, instead of fighting about money, you and your partner will discover how to start building a rich life together. Money for Couples is a podcast full of real life actionable advice like how to pay off your debt and still enjoy your life, how to build a shared financial vision, how to spend extravagantly on what you love and cut back on what you don't, and you'll learn from real world stories of couples facing the same money challenges as you. All of the episodes are helpful, but if I had to pick one or two there's one called we make $300,000 a year but spend like we make a million. That's a situation I think a lot of people can relate to. And another is called We've Saved for Retirement but have no Money to spend now. Money for Couples is the name of the podcast hosted by Ramit Sethi. And all you have to do is search for Money for Couples wherever you listen to your favorite podcasts.