Computers Can Predict When You're Going to Die… Here's How

Join our Patreon community: / itsokaytobesmart
SUBSCRIBE so you don’t miss a video! ►► bit.ly/iotbs_sub
↓↓↓ More info and sources below ↓↓↓
Predictive analytics uses math and historical data to make predictions about the future. It’s used in commerce, sports, politics, social media and tons of other places. And as it turns out, people have been using math to predict people’s death for centuries. Can it predict mine?
-----------
High fives to all our Brain Trust Patrons:
Ed Eyden
Jennifer Burton
Charles Horton, Jr
Holly, Brett, and Ashe Bullion
Jaap Westera
Mehdi Damou
Barbora Bei
Burt Humburg
dani bowman
David Johnston
Baerbel Winkler
Eric Meer
Karen Haskell
Join us on Patreon!
/ itsokaytobesmart
Instagram
/ drjoehanson
/ okaytobesmart
Facebook
/ itsokaytobesmartpbs
Merch
store.dftba.com/collections/i...

[@besmart]

I've got a few more good years in me… How do you feel about learning that computers can accurately predict your death and have been doing so for years? Let me know in the comments!

[@RichConnerGMN]

idk

[@AbdullahCumhur]

i got the 2nd reply ill edit it to answer the question when i watch the video :)) *EDIT:* Bruh it was all because the insurance people wanted more money 💀

[@gojohnniegogo]

I'd be a lot happier if it was used for our benefit and not so insurance companies had a nice a excuse to jack up insurance premiums. One thing I'm glad of is that the UK hasn't had people go bankrupt because of medical costs, as much as our current government are trying get this through.

[@michaelmayhem350]

🎼Shame on us, doomed from the start May god have mercy on our dirty little hearts Shame on us for all we have done And all we ever were, just zeroes and ones....

[@VictarisGX]

And here I thought we'd see a Futurama professor's Death-Clock reference. Well done!

[@dynad00d15]

**throws himself in front of a car** AH AH! YOU DIDNT SEE THAT COMING, COMPUTER! ow..

[@arth__BK]

Computer:" added to the database"

[@YasugoLiehu]

Me: “Computer, when am I going to die?” 'Puter: “Tomorrow” Me: “Oh, sweet, thanks. That's a load off my mind.”

[@resistanceisfutile520]

I wish

[@blobs2635]

Lol, I would be sooooooo scared if the computer said that!

[@nathantowns2043]

@@blobs2635so would YasugoLiebu. They’re just trying to be edgy

[@crystalclear6864]

😂

[@Daniel-sy6gs]

​@@nathantowns2043it's sarcasm, he's not trying to be edgy.

[@primarytrainer1]

love that the Hank Green doppelganger joke is still going lol

[@thyblackpanther]

What joke?

[@kagenekoUA]

@@thyblackpanther at the end (of the video)

[@ShirinRose]

I actually saw an Instagram comment the other week on one of Hank's videos asking him why he changed the name of his youtube channel from 'It's Ok to Be Smart' to 'Be Smart' xD

[@JohnThurner]

Doppelhanker

[@s.l.summers2958]

Watch the Scishow Quiz episode where Hank and Joe face off. They look so similar lol.

[@jeroenrl1438]

There always will be bias because somebody will decide what data to record. If favourite colour is important and nobody writes it down, it will be missed - and other (combinations of) factors will be found, thinking that will be enough. The problem with computers is that they can't be curious about types of data they don't have. They can't ask their subjects new and surprising questions, opening new ways of thinking.

[@Aaron.Thomas]

Of course not, but that was never the goal.

[@Cybored.]

not yet!

[@BattlewarPenguin]

6:50 finally, Charon bought something nice for himself with all those gold coins, he looks really happy in his new yacht Edited: Noo the ending, Joe stole his yacht

[@AllTheArtsy]

before watching: i mean this is the whole business of life insurance. those folks are literally betting on your life after the video: i mean, yeah

[@Andre-qo5ek]

pre-watch comment: i mean.. actuaries have been doing essentially this for forever... with the added part of BETTING when you will die, it is the definition of life insurance. post-watch comment: yup

[@CurtOntheRadio]

No, actuaries have been estimating life expectancies for groups/cohorts of people, not individuals. Surely? It's more like predicting cancers given X level of exposure to Y radiation in a population? You can pretty accurately predict a rate of cancer but you can't say which individuals within the population will succumb. "During the 17th century, a more scientific basis for risk management was being developed. In 1662, a London draper named John Graunt showed that there were predictable patterns of longevity and death in a defined group, or cohort, of people, despite the uncertainty about the future longevity or mortality of any one individual."

[@Andre-qo5ek]

@@CurtOntheRadio if you are part of the given group, it is a prediction of you as an individual in that group. the issue is only specificity of the data. there is NOTHIGN that can specifically predict YOUR specific death.. no... you would need a team of people that run around the world collecting specific data from everyone around YOU and people similar. im not sure i see your point?

[@CurtOntheRadio]

@@Andre-qo5ek "if you are part of the given group, it is a prediction of you as an individual in that group. the issue is only specificity of the data" No, it's a prediction for the group. There is no way to tell average outcomes for individuals - only the group.

[@CurtOntheRadio]

@@Andre-qo5ek Imagine throwing a 600 sided dice many times? You have no idea what number the dice will land on for any particular throw but you know the average will be 300. Human lives are like the single throw of a dice. One might live to be 110 whatever one's lifestyle, or one might die very young, regardless of lifestyle etc.

[@Andre-qo5ek]

@@CurtOntheRadio form the video .... "This is the mathematical theory called the law of large numbers. Basically, the larger your data sample is, the more likely it is that the average of that sample will reflect what actually happens.:"

[@Alex1891]

I had anxiety about death until a conversation with my father in which he had me consider how I felt before I was born.

[@erindickerman6138]

Same! I just recently "learned" this and it has helped so much!

[@odin6108]

I have been aware of this 'thought experiment' since I was in 3rd grade or so, and it honestly terrifies me way more.

[@Alex1891]

@@odin6108 The reason it doesn't terrify me is probably a combination of the fact that I was experiencing high anxiety about it already and that I truly didn't care at all about anything, say, 5 billion years ago. It's not easy to reject your natural aversion to your own demise but doing so earlier in life will save you stress later. Seriously, be well and if you ever want to chat, I'm generally open to.

[@SteveRowe]

Celebrating Germany today with your running gear?

[@besmart]

How do you know it's not Belgium

[@dannya8614]

@@besmart Because it is horizontal, not vertical :)

[@trevinbeattie4888]

@@besmart Germany’s flag has the red stripe in the middle; Belgium’s middle stripe is yellow.

[@mr.invisible5528]

​@@besmart Amateur. ;)

[@NotSoMuchFrankly]

Can a computer be an actuary? Well, yes.

[@ericjhubbell]

😂 "ahoy hades" on the boat, dying. Pun intended

[@KageSama19]

I choked when that little cartoon Joe shoved the grim reaper out of the speedboat.

[@hungryjackman]

I'm sure Achilles tried it, but Charon's smarter than that

[@kennycloudhead6232]

“You hear that guys. I don’t look old!”

[@georgeh6856]

Never tell the actuary about your cocaine usage or that you are learning how to juggle chainsaws.

[@Josf-xz3hw]

1:29 I Like The Vsauce reference 😂

[@GeneralJoey747]

I think a follow up video exploring which factors are most predictive of long life is in order. What are factors people tend to over value? What are factors that marketers over emphasize? How important is family medical history vs country of residence? I think there are many more interesting angles to approach the subject from.

[@x--.]

How rich are you and how long did your parents live... I'd bet those are the most predictive factors by an AU or two. Wealth plays into so many other important factors (health, schooling, neighborhood, food abundance, stress, country). Now what I'd be curious about -- what factors under your control have an impact greater than 5 years?

[@vtksolid9127]

Like your content entertaining informative and funny can’t wait to see 40 + more years of premium content ❤😂

[@lyledal]

Well, gosh! This is a feel good episode!

[@GeanAmiraku]

I wish you live a healthy live into your 100s! Thanks for your videos.

[@shaider1982]

2:21 yup, an algorithm used by Target mall guessed correctly that a girl was pregnant from the online choices. The girl's dad did not know and complained why she was receiving offers for baby products. The dad apologized when he learned the girl was pregnant.

[@Artyomi]

Okay that case was not really even that much due to the algorithm - rather the incompetence and audacity of the dad. It would be like saying “a forensics program predicted you’re gonna make amphetamines because you have receipts that you bought solvents, acids, phosphorus and ammonia from home depot”

[@jillcrowe2626]

I read that book too. "The Power of Habit" by Charles Duhig

[@shaider1982]

​​@@jillcrowe2626 yup, I first read about it in that book.

[@boblangill6209]

As people age, a common assumption about them is they sense they have less time remaining. At 74, I don't sense that the time I have left is getting shorter, just increasingly indeterminate.

[@oliviafeltis3604]

If you're old then man, I'm ancient 😂

[@aaron101889]

What was Mesopotamia like Grandpa/ma?

[@bloodycomedy1927]

do you have pet dinosaur?

[@lakshyasingh1527]

Hey Joe! Really love the videos. A request, please make a video about astroinformatics. How AI and Data Science is helping in astronomy.

[@bnthern]

again well presented

[@helmann9265]

Fantastic 💯 👑 thanks

[@DynaCatlovesme]

The original Lloyd's of London wasn't a company that sold insurance. It was a place in a building where people who sold insurance congregated.

[@Gigaheart]

Fact check needed.

[@rdapigleo]

You look fine Joe, don’t worry about it, until you’re 50. Great episode, thanks. Episode on living longer? David Sinclair?

[@markusseppala6547]

Have a high VO2max and don't go caving and you're good.

[@aaron101889]

I'd rather die young than train in a way that increases my VO2 max

[@jeremyscungio16]

I do remember showing my 7th grade science teacher your 12 days of evolution series when it was new

[@dremichael2335]

1 min into the video and I’m rolling.

[@victoriaeads6126]

8:16 NOTHING is free from human bias. The computer can only use data we provide, therefore it also is biased at some level. That's why so many early ChatGPT type programs that were trained on social media quickly became racist, misogynistic, and homophobic. ChatGPT uses much larger data samples and algorithms that are tweaked to try and avoid that pitfall.

[@NicitoStaAna]

Never forget Tay ❤

[@deadliestt]

Working for a life insurance company, I have heard a lot about predicting when someone will die. I don’t work in the department that deals with that sort of thing, but it absolutely influences my job. It’s almost surreal to hear words used in my industry and know that in the next several years, my company will more than likely be using that AI software.

[@royalfelineandtracygrant]

Question: does the computer also take in possible events? Having a kid, adopting a pet you’ve never had before (like a snake or something) etc., or stressful events like your job gets really bad, someone you love dies, etc. if it can not consider possible events, how is it accurate?

[@xomiachuna]

It is accurate in a sense that for a large population the predictions on average will be fairly close to the true results. On a case-by-case notion it is not guaranteed to be precise (many random things affect the mortality), but in general it will be close (in the order of 1-10 years in many cases) for many people. So think of it as of a vaguely correct, but not exactly precise.

[@richardbeck8945]

For the most part, yes. Many common life events such as if and when people will have kids or what types of pets people will have and at what ages those things will happen are recorded in all that data they amass. So they have a fairly good idea how many people will own poisonous snakes and at what age a person will most likely be when they have children and how many they have. But also keep in mind, the younger you are, the less accurate the predictive model will be, because of all those predictive variables haven’t happened yet. Which is why the guest mentioned that at birth the prediction for Joe would be to die in his 70’s. But now it’s his 80’s. i.e. the older you are, the more data points you have resulting in more accurate predictions.

[@vinniepeterss]

love this

[@margarettaylor2057]

One thing that I wished this episode had made explicit is that life expectancy means you are 50% likely to make it to that age NOT that you are likely (unquantified) to make it to that age

[@diane_princess]

I sort of learned about this when I had to learn how life insurance works and determines how much you have to pay them. I had to do it by hand with given percentages. It was extremely difficult. But the principles are the same.

[@gibberishname]

I honestly JUST finished re-reading "Machine of Death: A Collection of Stories about People who Know how They Will Die" and "This Is How You Die: Stories of the Inscrutable, Infallible, Inescapable Machine of Death" about a week ago.

[@AlexiHelligar]

In the spirit of staying curious, I would like to to see how close the fortune teller's prediction of your likely death is to the computer's prediction.

[@DuluthTW]

It must have been weird for you to see your face as a "dead guy" on the ABC TV show Will Trent. I got a kick out of that. I hope you live a long time to continue sharing interesting science. Thanks for sharing!

[@alchang1515]

Noting new, period!

[@dmfouge]

Hey Joe, smart people here! please make videos more often, they are always so interesting

[@roger7341]

My iPhone has a builtin accelerometer, and if it measures a 100g impulse it will probably figure I fell out of an airplane and dial 911, assuming it survived. I'm 81 years old and have outlived three iPhones, and over the years, not one of those iPhones has accurately predicted its own demise, so I'm not holding my breath that my current iPhone will accurately predict its or my demise. I believe in the inevitability of probability: wait long enough and the probable is inevitable. There is about a 50-50 chance that the computer I'm typing this comment on will die before I do, but it won't admit that, so who can you trust these days?

[@Questerer]

I always lean so much from Joe.

[@secretagent86]

I correctly matched 86 before the actuary.. but is used to sell life insurance

[@IsaiahRStudios]

Reminds me of the death clock from Futurama

[@DavidCaveperson]

I like to think the guy who wrote that joke studied a lot of math and kept talking about becoming an Actuary and passed several exams and never got hired, so he was poking fun at becoming a bitter old man (i.e. not a recent college graduate) and now he feels like a joke, but at least he can write comedy. But I don't know... what do you mean my impression is based on personal experience???

[@jennysrp]

I think it’s important to clarify that your individual tracking apps like health trackers are not submitting your data to mega-databases on a super computer somewhere to be harvested for global data - it does explain in this video it takes a collection of data to create a ‘Frankenstein’ of data and that’s important to remember. Your ‘personal’ evaluations and recommendations are based on Frankenstein data, not the same as individualised personalised recommendations. So yes, more and more companies are collecting your data every day, but it isn’t kept connected to your individual identity throughout the data line, and overall estimates and analytics are what are being looked at. Anyone who works in analytics knows how misunderstood it is that all data lives somewhere on an infinite computer database accessible somewhere :D

[@shorts_and_video_from_mini]

i love ur videos man you're the best :))

[@SachidaNand-ft9qb]

if all the visible Ray passes through glass then why we see glass white in colour

[@Petch85]

Well "When a measure becomes a target, it ceases to be a good measure", thus people will change there behavior. You might not seek a psychologist when you need it cause you don't want it on you record, you might install software on your computer that dilutes your data, or something else.

[@valuethug]

This is sinister. From it's inception: who does the pirate data help: the sailor or whoever funded the voyage? In this era: the data does little for the individual but is used by powerful groups that don't consider you but rather a dehumanized 'average you'. Creepy really.

[@Rantir]

Heroes may die, but legends live forever... or something like that... if you are unlucky you might die earlier than expected, yet since you have built something up, others will remember you for quite a bit. some famous guys are dead for centuries now, but history still preaches their deeds or misdeeds. hopefully you can enjoy your few good years in peace and may haps spoil us with your ever-growing wisdom in the future, but keep in mind some good folks still went down the river of Styx way to early.

[@shantanusapru]

Q: Can a Computer Predict Your Death? A: Sure! Why not? Heck, an octopus can predict your death! As can a tarot card, or a parrot... The real question is how accurate will that/those be...?

[@fortierma64]

Yup! Not sure all Canadians exercise regularly and eat healthy foods but we do have access to good healthcare when we need it AND it’s “free”. 😊

[@lajya01]

Not the Canada I live in. I have to pay private care or wait until death.

[@fortierma64]

@@lajya01, I am sorry to read that, I am in Quebec and it’s a different reality at least for me and my family.

[@lajya01]

​@@fortierma64 That's exactly what's happening in Qc. Wait until your GP retires...

[@fortierma64]

@@lajya01, I hear you. Mine is retiring in a couple of years so I’m good for now but you’re right, could prove to be a challenge then. Fingers crossed. I still think that when everything works we have a good system.

[@male20yearsold]

Humans : predict my death. AI : prediction will be endless, but i can end it for you anytime. Humans : that's sounds right, but a bit wrong,

[@therealmagicmonti]

The logic about insurance, is that you spread the risk from an individual to a whole cohort. Such that an occurrence of an event is financial, too expansive for the individual. When predicting an individual's behavior causes an insurance ad absurdum. Because you don't spread the cost evenly to a cohort anymore. Obviously, this is the desire of the insurance company to lower risk and gain more profit, but the idea of spreading risk is lost. But luckily, health insurance companies are not allowed to process such very private data, otherwise the healthcare system would crash, this is in most European countries embedded by law, not the idiotic-false-believed-freedom state. *Addedum; that's why you should only take out insurances, for events which could bring you in a financial ruin. And not for an event like accidentally dropping a smartphone. *Nota Bene; When the prediction of events are "too" good, people will stop paying for the insurance, so the business model of insurances will vanish.

[@NiTeLightYears]

Hmm the computer in a more advanced version existing on Death Note will make an interesting plot

[@onetruetroy]

It would be interesting to have an AI call/text you on your birthday: “Happy birthday, ! I wish you many more, but you’ll only get to enjoy more of them. Bye!”

[@traxstaromega3467]

I'm not old but it still scares me out of my mind

[@MildStallion1]

"Ahoy Hades" will be the name of my next album.

[@rustycherkas8229]

"Suggest some good videos to watch while we're waiting..." Scene from "Meet Joe Black" when Hopkins knows his time is short and there's his "daughter" worries... He's playing Solitaire (with real cards)... Kinda makes ya think, no?

[@etownshawn]

Super funny intro! You're a runner like me Joe, we'll get to 90 easily. Do we want to be 90 though? ehh

[@-Subtle-]

90 with a 30 year old body and brain? Yeah. 9O with a 90 year old body? No

[@dillonrose3428]

I am still baffled by my inability to distinguish between Hank and yourself. And I apologize. 😅 🤘🏽

[@astronics]

they do look quite a like, i just think hank is the more chaotic one!

[@dillonrose3428]

@@astronics nahhhh, they are both chaotic lol putting their own spin on the way they choose to present themselves. It could maybe be from me watching them both daily, but this one had me lost in thought thinking this was Hank! Embarrassing as it is, I was confused and convinced Hank found the most beautiful and flawless wig! 😅😂 I was baffled

[@aussie405]

You may have Prosopagnosia.

[@dillonrose3428]

@@aussie405 well if that be the case, fingers crossed everyone gets their name in Braille on their faces lol 🤞🏽

[@richardbeck8945]

It’s easy now that Hank has curly hair. :)

[@ALAN4SV0GUE]

Hi, please could u make a video on the frontal lobe development?

[@samhill206]

Just curious....any reason why his sweat bands colors are the as the German flag?

[@rickseiden1]

They should start calling the "black swan" events, "Dean Winters" events. "Mayhem, like me!"

[@jackovoltraids5937]

"Dennis, our lives are in your hands and you've got Butterfingers!"

[@raphaelgarcia9576]

Thanatos phobia, the fear of missing a Marvel movie.

[@verifiedcartophiliac]

Got my daily quota in, thanks to Joe.

[@sylak2112]

Me reaching 45 years old this years hurted. I don't feel old, but.. that is not a lot of life remaining... totally feel that mid life dread.

[@Joshua52391]

its ok you still have the potential to live another 30 years maybe even 40 years if you're health isn't to shitty

[@JVDetmer]

Damn! This reminded me to Rehoboem from Westworld 😱

[@whycantiremainanonymous8091]

Two huge logical leaps in the script: 1. You can predict how long somebody like me will live on average, not how long I will live. 2. I think of death several times a day. I don't feel anxious about it.

[@morbidiablack5321]

I agree with both

[@whycantiremainanonymous8091]

@@morbidiablack5321 Nice username 😉

[@drieschnoronha]

8:17 Is it really 'free' from human bias? Given that humans are selecting/collecting/etc.. the data which is fed into the ML algorithm, wouldn't thete be some bias making it's way into the process?

[@KuruGDI]

The name of the boat should have been _Slice of Life_ Just saying...

[@Tharkon]

3:30 Wrong, I actually am very predictable, I have already determined what to eat four weeks from now for example. 3:40 Wrong, because my sleep schedule is terrible. 3:50 Wrong again, because I'm unemployed.

[@lc7ineo]

"All we have to decide is what to do with the time that is given us." GtMFG

[@19lamborghini92]

0:26 not me thinking he didn't have shorts on 😂

[@SojournerDidimus]

4:30 Is he really called Dall-E HAL 2001?!!

[@jillcrowe2626]

My husband has almost no digital footprint. He uses my daughter's Amazon account. He barely answers his cell phone.

[@mrbfros454]

As soon as I saw how happy you were that he thought you were about 40 I was like wow he is old😂

[@crystalclear6864]

The insurance companies will love it!!!

[@raphaelgarcia9576]

So oooooold. Wait, what’s the alternative?

[@eschwarz1003]

to those trolling about someone looking "Old" I hope you're lucky enough not to have to suffer the same fate

[@xtieburn]

I mean... the video literally just explained where this is actually going to be used, and it isnt to help us make the best of things. Its for insurance companies that will be able to better tell who they can make money from and who should be left to die.

[@CurtOntheRadio]

But arguably what is better for the insurance industry is better for everyone - it's a more efficient use of resource. How wealth etc is distributed is a political matter, not one for the insurance industry.

[@jayplayzlol8701]

While we have free healthcare in Canada, we sometimes just walk away from it because it's been 5 hours waiting at the hospital and you had the time to convince yourself you'll be fine 😂

[@leehamilton4459]

And here in the US it is common for people to refuse ambulance service because they don't want a $1,200 bill coming to their mailbox. My husband fell off a roof, he was unconscious for a moment, an ambulance was called, he woke up and flat refused the ambulance ride. I ended up taking him in my car. Crushed R hand, ankle damage and thankfully just a bump on the head. His hand took most the impact.

[@secretagent86]

Robert Heinlein had a short story on this topic but when Lazarus Long was test the scientist said his machine was broken.. forgot the story name but entertaining

[@dennismccormick9204]

The 4am prediction missed me, oooops. Guess I'm not so predictable.

[@datonz]

"Thanks to denial, I'm immortal."

[@ElicBehexan]

Oh, I know what you mean... I am a 70 yo female. My grandmother died about 61 or so from a stroke. My mother had her first stroke in her late 60s. I am neither as healthy or even as active as either my mother (healthy) or grandmother (active.) I walked away from totaling a car shortly before my 19th birthday. I took 6 or 8 falls off horses, including one I landed on my face. That was probably at 15 or 16. I believe I should've died in both cases. I kind of believe in the multiverse and in other universes I did die. Am I going to make it to 84 like my mother - well, if I do, I hope it isn't in the same shape as my mother because she was suffering from stroke related dementia. Do I think about death... come on, I'm 70! Do I worry about death? Not so much. There are worse things that death out there, like being a vegetable because of strokes.

[@L33t5uPaH4x0r]

Yes i went to the website

[@kamyafred3114]

We're heading towards Westworld Season 3 apparently

[@sorosch6960]

I wanted some kind of website where we can fill in our information and get an estimate bruh

[@sarahreavis6085]

Where do we go or who do we contact if we want to hire AI to predict our own personal death?

[@NWDestroy]

man, as peaceful as death was, I sure do prefer the times when I'm alive

[@toughenupfluffy7294]

Don't worry-we're all going to die.

[@biomatrix8154]

My impression from this video is that no matter how unique we think we are, we're all just number vectors in the end.

[@fidogyrenuzanon2335]

I have a few relevant paradoxes. Firstly: For those who are not immortal and will die, merely living is a temporary act of cheating death, and it will eventually fail for the individual. Next: Without the existence of life, the universe is merely one big collection of rocks. Yet, if it has life, then efforts must be given in order to sustain that life, and life will end up being very costly for the universe. Next: The only way for a mortal to effectively cheat death is to live long enough to have offspring and thus mar death’s mark on the world. Finally: Due to the deaths of all mortals, those who are immortal will never truly know that they are immortal, since in their minds death could still take them. In other words, mortals exist for a numerical amount of years, but the immortal exist for an unnumbered amount of years, with no set start. The mortal and the immortal cannot mix.

[@heybro345]

3:24 😂

[@dianasofia5075]

Hey, don't have a heart attack, you're not 92 yet.

[@crocshock911]

Sadly, predictive analytics are used far more for marketing and money-making than they are for things that actually matter like life expectancy.

[@CurtOntheRadio]

Doesn't "money-making" provide a means to realise "things that matter"?