I, V, M. Yeah. And I want to start with the Profundity of the Week, our short-lived segment. You know how Facebook now has these past memories, because you're, because they finally realize you've run out of things to say so they keep showing you stuff that spoke about in the last few years.
So I like browsing through this. It's good to see the, you know, when I, is a revolution as I go backwards, a few days back, I scroll all the way to the bottom and I came across a profundity which I myself posted back in 2009. Wow. Okay.
And it goes like this. Not sure if I'll get by, even with a little help from my friends. Wow. And it happened to me that day.
Well, a lot of these things are like, they're like snapshots of your mind. Okay. So when you take pictures, it's a snapshot of how you're looking, but very rarely did, I mean, the only way to remember what you were thinking in the past was to write down something. That's why people used to have diaries and these two.
And if we does that, I've had an experience and, you know, you see something you posted a few years ago and then a lot of memories keep start, start crushing in and you remember what you were thinking at that time. And what was I thinking of that? Yeah. You know, I don't know, but in this particular case, probably not, but in my case, a lot of associated stuff will come back.
I remember if I was feeling sad, happy and we had something at someone friends who you're very close to that point of time. Yeah. And sometimes people, you were very infuriated when people say something to you and looking back, you sometimes realize how silly it is. Right.
So yeah, and the other way around. And there are times when you really should have slapped the sumbitch and you didn't exchange some very little. Oh my God, he got it. Yeah.
Anyway, moving on. Big news last week in the tech world and not in Elon Musk again. Okay. You might have to do something about your man, Kresch Bader.
Okay. Now get this. And allow you to pay to unlock a feature of their car. Wow.
What does that mean? Okay. So the model S has two batteries. I mean, not the same car.
The two types of batteries are 70 kilowatt hour one and there's the 75 kilowatt hour one. The latter offers you 19 miles more of range. Now even though it's only 5 kilowatt hours more, it costs 3,250 dollars more. So there's a lot of money.
There's a lot of money, right? So when you want to buy the car, you might choose to buy the smaller one. You might choose to buy the 70, but later you decide that, hey, I want to upgrade to the 75. So what Musk has done, what Tesla has done is they're actually shipping cars, all the cars with the 75.
But it has a capacity only of 70 and you have to pay this money to unlock the rest of the 5 kilowatt hours. It's like software. It's like an app. You have to pay to unlock a feature and that's mind blowing.
The first time something has been applied to something like hardware and hardware like a car. I mean, absolutely mind blowing idea that and the logic that he has provided that Tesla has provided is that it's actually easier and cheaper from a logistics perspective than they don't have to manufacture two different types of batteries. They don't have to make sure the right batteries go into the right cars, have different sales, challenge for each and stuff like that. They've done this before.
They've done this before. So I think that's the first thing I've done is that the auto pilot feature, you could pay to unlock that as well. So it's not the first one that has done this, but that again was more software. This is hardcore hardware.
Just imagine you are utilizing less than what your car is capable of. I think that's a fantastic idea. Which is mind blowing. I mean, is this the future?
You know, fully loaded devices, but you access only what you pay for? It's mind blowing stuff. Yeah. So the repercussions this could have amazing.
Like imagine the iPhone for instance, you get only, I mean, everybody shipped out the top level iPhone. I mean, the top specs and all that, but you pay to unlock from 16 to 32 to 64 GB and things. It would be pretty interesting if stuff like that does happen. Yeah.
So I think people want to come like this. Right? What is that? I mean, what if people are like, I have paid one long feature.
So okay, we are on grim guy and all you can do is start packing. I clean out the gears. I can't stop packing jokes. Yeah.
We are like the ad where people pay for the premium version. We'll stop existing. So it would be an interesting thing. So when you go for a job and you know, if you're over qualified, and you have to dumb it down, the company say, okay, you know, you're very qualified and we know you want to know X, but we can give you only 0.7 X.
Okay. I perform only 0.7 X or I'll come only for three days or something. So we have to pay to unlock the full, that'll be pretty awesome if that happens. In the future, everyone is an app.
Yeah, I think. On that, uh, my lead profile note, we are going to take a little break and we will be back with artificial intelligence in just a while. Are you looking for finance products and services that can enhance your personal finance experience? Are you looking for a space to talk about your financial product or service?
And are you looking for a Chris Proxio where the conversation is all about money? Well, your search ends here. Hi, my name is Arubak Gupta and I post all the best of us at the podcast and I invite you for the conversation about your personal products for each Monday on the IBM podcast app or the website or at any podcast streaming peppers. See you folks.
It feels like everyone's talking about two things right now. Pickleball and podcast. So what happens when someone who's lived her whole life on the court decides to bring those two worlds together? I'm Sanya Mirzam and I'm bringing you a brand new podcast, Serving It Up With Sanya.
I've had some incredible conversations with friends, fellow athletes, artists and entrepreneurs. Each episode starts with the game of Pickleball and once the rallies are done, the real job begins. Serving it up with Sanya, watch and listen on YouTube and wherever you get your favorite podcasts. And we are back with Simplified.
Now, we're going to talk about something that's being hailed both as a savior and the end of civilization itself as we know it. Didn't we promise we're not going to talk about IPL again? No, no, no. I'm actually talking about artificial intelligence.
Okay. Yeah. You know, you don't know what artificial intelligence is doing right now. Of course, what engineering college students display during Viva?
Fair point. Actually, all in a blonde hair dies a hair black. That's all good deal. Anyway, artificial intelligence is a technology by which machines are made to think more and more like human beings.
Given what the human race has been doing to itself over the decades, I don't know why anyone would want that. We should be doing the exact opposite. We should be trying to think like machine like machine like machine like machine like as human like as possible. We like Mr.
Spock. Yeah. Let's not get too philosophical right now. In any case, like, you know, coming back to the top.
Yeah. Yeah. Why don't you enlighten us about artificial intelligence? Yeah.
So I've been doing a bit of reading about AI. He's a term coined by a scientist called John McCarthy back in the 1950s. Just around the time computers were becoming big. Pretty literally big.
Literally big. Yeah. He's the whole room's back. We have one in my engineering and money part, right?
So it was there was this computer which had out, I don't 16 kilobytes of memory. I mean, it occupied an entire room. We have to remove your shoes. So actually, you know, inside.
All right. So anyway, the concept of artificial intelligence is actually quite simple. Like I said, it's getting machines to think like humans. Now that's a pretty broad definition and over the ages, the outcome of artificial intelligence has evolved.
I mean, right now when you say artificial intelligence, we think of high fund art things like self driving cars and voice assistants like Siri. Yeah. I mean, back in the day, I'm sure things were far more humble. Right?
Yeah. So back in the day, a calculator was considered artificial intelligence. Yeah. Okay.
So you know, the movie on Turing's life. Okay. This entire army of women who are actually paid to compute. Yeah.
Pay to compute. Right. Now, this is something like that something about today, but that was cutting edge technology back then. So artificial intelligence actually is pretty much everywhere.
I think a good thumb road to follow would be was it a job that was once done by a human or perhaps even an animal because you're thinking of like grazing and stuff like that? If you got a machine to replace it, then chances are you've got yourself artificial intelligence in some shape of form. Wow. So a computer type?
Yeah. Artificial intelligence used to be done by dictation and the stenographer before. Right. And now you can do it.
Yeah. Yeah. So that would qualify as artificial intelligence. Yeah.
So artificial intelligence in a way as well. And yeah, I mean speaking of ATMs, the new avatar is things like mobile payment apps, you know, networking and stuff like that. That's again, another evolution of artificial intelligence you could say. The question here is, is that really intelligence?
Is that, you know, is that, is that it's difficult to define, but is it something a human would be able to do easily thinking on the fly and a machine wouldn't? Yeah. So like I said, a good thumb rule to follow. I mean, let's, okay, one thing about AI is everybody disagrees on it.
And nobody disagrees with it disagrees on it more than the people who actually work in artificial intelligence. Because there's a big thing. You can say that in 2016, we are thinking, we are thinking self driving cars. Very, very high funder.
It's very hard for us to say that calculators were artificial intelligence because or, you know, something which could convert your speech to text. Or even a regular car for that matter. It may seem simple to us now because of being off for so long, but heck it was cutting edge at one point of time. In fact, there's a term for this and I love this.
It's called the AI effect. Wow. And that applies even to all things. Yeah.
Yeah. So what the AI effect actually is, it's a sort of semi-clinical and a self-ironic term coined by the AI industry themselves. And basically means that if something that is AI goes mainstream, it's no longer AI. Wow.
Now we think about self-driving cars because wow they're so out in the future but say 3040 years later actually becomes mainstream Nobody thinks of self-driving cars as AI So it's too mainstream, it's almost like a sips to it's too mainstream, it's no longer a hipster AI is actually the most hipster thing there is So it's like the magic has won off right? Yeah So young married men listening when your wife tells you you're not like how you work you can tell them that the AI effect is just the AI effect has been I think it's time to take a little break Absolutely and be right, be right back It feels like everyone's talking about two things right now Pickleball and podcast So what happens when someone who's lived her whole life on the court decides to bring those two worlds together? I'm Sanya Mirza and I'm bringing you a brand new podcast Serving it up with Sanya I've had some incredible conversations with friends, fellow athletes, artists and entrepreneurs Each episode starts with a game of pickleball and once the rallies are done the real job begins Serving it up with Sanya, watch and listen on YouTube and wherever you get your favorite podcasts Are you looking for finance products and services that can enhance your personal finance experience? Are you looking for a space to talk about your financial product or service?
And are you looking for a crisp workshop where the conversation is all about money? Well, your search ends here Hi, my name is Arupak Gupta and I post other best of us at podcast and I invite you for the conversation about your personal finance for each Monday on the IVM podcast app or the website or on any podcast streaming platforms See you folks And we are back So we were basically talking about the AI effect and how mainstream things are no longer concerning AI So just imagine for example, you could transfer cash by thinking You know, like, okay, we were for doing said I owe you 500 bucks So instead of taking up my phone and sending it to you Which was AI which is actually AI right now And just close my eyes, I don't know, I touch your index finger or something like that To authenticate and just get automatic transfers Or it doesn't even need to do that My phone is listening to the fact that okay, I own Rn500 Rs and it automatically debits money from my bank account It may seem futuristic to us right now, but you know, that's the future You know, so the AI keeps evolving The outcome of AI keeps evolving is what I like to say So like a well told story Yeah, or badly told lie, it keeps evolving So I'm going to get a little deeper into what AI is Pay attention, lean listeners So I'm going to simplify AI for you AI is actually very very complex It takes into account, if you think about it, it takes into account almost every aspect Every useful aspect of human knowledge Computer science, mathematics, psychology, linguistics, philosophy, neuroscience And specialized fields within those And then not to mention depending on what the specific AI is used for You need expertise in specific fields Like for example, finance, trading, medicine, engineering and stuff like that But not not MBAs Yeah, that's why I said useful aspects of human knowledge Well, please get it Okay, so it's coming together of all these many many diverse fields That artificial intelligence is composed of There are many legs to AI After all, if you want a machine to behave like a human You need to figure out reasoning, knowledge, planning, learning, language processing, perception Ability to move and manipulate objects Mind boggling the amount of stuff that you actually have to do in order to just make something like a robot Amazing So so a PhD in artificial intelligence will pretty much make you a jack of all trade Pretty much but I'm guessing that you will specialize in one aspect of this I don't think you'll have a PhD in artificial intelligence as a whole That's probably impossible You know, I guess you specialize in one part of it Let's rewind a little bit I mean after I did bother reading the entire Wikipedia article for AI That's the extent of research The concept of no idea Now I'm expon I'm not quite expert What's the term used? Ultra-creeped area is the official word of simplified The concept of making machines more human-like is nothing that's very silly Convolution Okay, old Greeks used to worship idols they say believed had intelligence And it's always been a part of pop culture and literature Like the most famous example would be Mary Shelley's Frankenstein Yeah, side note many people think that's the monster that's Frankenstein But it's actually the creator that's Frankenstein I remember this fantastic cartoon that was created a comic that was created Where a postman comes for a delivery on Frankenstein's monster opens the door And he says no no no no I'm the monster of Frankenstein lives next door It's a common mistake Okay, interestingly enough The these pop culture reference like Frankenstein and other examples of artificial intelligence The same concerns as people are expressing today Which is hope, fears, ethical All those sort of models of the story as it were from Very interesting So some things don't change Some things actually don't change With the introduction of formal reasoning Logic was formed Computers were born Alan Turing The famous name which you mentioned a little while back Was one of the only thinkers he was a famous mathematician And he believed that just with zero and one A computer could simulate any mathematical deduction And then of course there is a famous world famous Turing test Which means a computer does something Or a machine with artificial intelligence does something that is In discernible from a human From what a human would do So there are various types So you often say that if a machine passes the Turing test Then another human being can't make out whether a human has done that Or whether a machine has done that So that's sort of like a thumb rule I suppose People start wondering something like an electronic brain possible And let's be honest That would be useful for many people Yeah, I can think of dozens more Hundreds of thousands Insulin me There's a college in the US called Dartmouth College in Hanover And that was sort of like a hotbed for artificial intelligence I'm sure your college was the hotbed for many things My college was the hotbed for many things which I can't mention My college was the hotbed for hotbeds Alright, moving on from bad jokes So in this college Dartmouth College A lot of they spent a lot of time on artificial intelligence And this is back in the late 50s early 60s types Simple programs that were developed by students Would actually stun audiences Checkers solving algebra problems Word problems in algebra Proving theorems etc etc Soon enough the Department of Defense They realized okay this is probably something interesting And we should invest some money into this It was a bold time for the artificial intelligence And it was a very optimistic time Yeah, you know Yeah, you're constantly looking at new horizons And every time you look at it the horizons shifted Yeah, yeah Like solving checkers today bombing enemy territory tomorrow It's just a short step So people thought that it was going to be huge and yeah, yeah That kind of thing Like like internet of things Right now, nobody knows how big it can be How big it can actually be And of course, slums obviously happen But in the late 90s is when it really really took off Because of a lot of things The requirement for logistics data mining Medical diagnosis And not to mention Change in computing power also helped You know, by that because artificial intelligence The weight was going, it needed so much computing power That it needed Moore's law to kick in It till the 90s So that you could actually have computers That are powerful enough to process these requirements Sure, back in the 60s, you had computing power That is enough to play a game of checkers I suppose But anything more than that wasn't happening So hardware also progressed And hence AI progressed parallely And also what happened was It stopped being a niche thing in some lab Okay, consumer effects of this could be seen in real life And you started entering pop culture And for part breaking things For example, I'm a big fan of chess And one of the big things in the 90s was deep blue The computer beating Gary Kasparov So Kasparov was bad ass just there And a computer could beat him that comprehensively I remember the video, I remember seeing the video He just shook his head and walked off He would make all those bug-eyed expressions What the hell is that? So there are plenty more moments like this Very recently, you would know There was a computer called AlphaGo So there's a very complex game called Go Which a lot of Chinese people play And which is very resistant to brute force calculation And a machine actually managed to beat the best Go player in the world Now there are self-driving cars It's just like how deep blue beat Gary Kasparov IBM's computer Watson beat The two greatest ever jeopardy players in a game And there are other things like Microsoft's Kynakt, Bursil Assistant's X-Series So pretty much AI has been catapulted into public consciousness And people are no longer You know, it's no longer a Jotrap They expect these things to happen It also makes the outcome of AI seem a lot more tangible Otherwise AI just becomes somebody And some lab is doing something It will just otherwise seem like an airy-fairy lab experiment I want to get a little deeper into AI Okay, that sounds very naughty No, no, no, no, no, no It's actually not If you actually have to be surprised That becomes a legit use for artificial intelligence in the future AI used to be a bunch of algorithms Maybe today as well It's best to think of artificial intelligence It's just a very large algorithm A big as flowchart Like a really, really huge as flowchart If this that stuff like that Only these flowcharts have evolved from a simple one Giving a calculate output To something that could actually act as a pet Like I remember in the late 90s There used to be these things called Tamago cheese Which are virtual pets That itself was fairly advanced in AI Now you actually have real pets I mean, by the way real pets are physical virtual pets Who would have the range of behavior that you would expect from AI And without the poop cleanup as well AI got better instinctive intuitive He started filling in the gaps and self-learning All this of course came with the massive increase in computational power requirements Yeah, I mean, early computer power would never be able to handle it Absolutely not AI has two parts Which is knowledge representation Engineering, which is how it reads or uses data So it's going to actually get damn complicated here So if you're going to go any further, we'll need an actual AI expert But essentially what these two things do is AI needs to know about everything and their relations Yeah, such as for example Through learning Through learning for instance Artificial intelligence A robot can figure out that this is a boy And this is a ball Okay, so that's like data Now it needs to figure out the relation between the two Okay, the boy seems to be picking up the ball and he's throwing it So maybe he doesn't like it Or maybe he seems to be doing it over and over again So he's enjoying himself Which means he's actually enjoying it So he's doing this for past time So now he's learning the fact that the relation He's learning what the relation between these two things are And the thing is that the computer is or the program Is sufficiently empowered to pick up these two Absolutely, absolutely So that's the sort of El Dorado as it were Did I use the term El Dorado right?
I'm not quite sure That's the end part I'm not sure at all Is that right? I'm afraid now So it is more than just a massive excel file And that's the challenge for people developing AI The major goal of AI is self-learning through experience So that sounds awesome And scary And I think on that note We should take a little break before we get back into AI Yeah, so I just saw this and you'll be right back Yeah Now I say, sashir kal Good morning, evening afternoon My podcast is titled Says When no guys, no topic and no filter is safe from Father Roger That's me, Monday to Friday, Saturday Sunday time We love kids We chat with students who never meet Traffic jams on streets and people from all professions Every Monday I go live with the show Ask me anything, AMA You follow the questions I know it's better than a Mumbai Chaterer I even host a panel discussion called Calk and Bull Where we dive into current affairs and wonder why anyone even cares at all And Friday releases all the surprise episodes for you all So tune into CyberSayers on YouTube and all our new streaming platforms You know that feeling when you finish a movie and you just need to talk about it? That desperate need to discuss why Kalwina Ho hits different Enter IBM's brand new podcast, Cine5 It's streaming and current just yapping over Hindi cinema every Friday Romcom's director of UW's career defining moments Whatever they're vibing with, they're breaking it down Join the podcast on IBM Pub's YouTube or any audio streaming platform Caution, you'll want to re-watch everything after listening to us Come join us on Cine5 And we're back To waggle your imagination a little more With artificial intelligence You can mean that Yeah, I've seen that given human beings also do the same things And have done far worse I actually say we don't have too much to worry about But sure, artificial intelligence can be used to replicate Meanial tasks, right? Physical or mental But can they simulate creativity in the worthwhile visuality?
Yeah, okay, here's the thing And don't start to get worried about this There is actually something called computational creativity It's actually quite a heady mix of artificial intelligence Cognitive psychology, philosophy and the arts Again, I don't notice any MBA, right? It won't, right? It won't let it go Now computational creativity has three functions You can either simulate human creativity It could understand human creativity and formulate algorithms On its own, okay, I've noticed this piece of music And 1, 2, 3, 4, 5, 6 things that make it good So it can't actually create music on its own But it's able to pick up patterns and things like that And the third function of computational creativity Is designing programs to enhance human creativity So I'm guessing something like Photoshop for instance For example, a simple use of AI I would say is when you put in photos in Photoshop And just say autocorrect It just does it by itself And they're pretty good And they're pretty good In music, I've seen a lot of these electronic music Come out from its algorithm based I don't know how far that's true It's true It's just to suggest that patterns are also Because you know, it's very easy to program scales Once you get a hang of song patterns Which is say, verse, chorus, verse, chorus Even in an instrumental format And then you also figure out What are the points that make What are the sort of chord changes and tone changes Which trigger off certain things in your brain So bring them more not to me Yeah, yeah Introduce different phrases at different times Absolutely, like in the last episode with Mohan We were talking about the tritone, the devil's note So that sort of caused a dissonance in the mind Okay You know, a major chord changing to minor chord And things like that So yeah, AI could actually be used like that Yeah, so it's tantalizingly in that zone Where it's creative and not creative That's what it's talking about I guess we need to ask ourselves at a very philosophical level Can AI really be genuinely creative? Generally creative It would originate something Which would be indistinguishable from the creativity And now we are going to start getting into the really philosophical part of the discussion There's a question that's been asked by many In fact, for creativity, senseions itself Or what is called general intelligence Is this a big goal for artificial intelligence?
And whether we want to achieve it Is another matter altogether But we speak about it later By the way, do you know that the advertising agency in Japan has hired a robot with artificial intelligence as a creative director Wow, is your job on the threat now? Well, it created ripples, my job is on the threat For now, at least not from a robot From plenty of other things So what it does Okay, I remember while we spoke about computational creativity And formulating algorithms for creativity So what this sort of creative director does is It goes through ads And it figures out what is working in these ads And then is able to create a set of formulae or something That's so that new ads are created So it learns from existing ads So I don't think it creates ads Firstly, I think it's just direct So somebody, say a junior copioid or something Comes out with an idea Maybe the machine will be able to take it and run it By its set parameters And right in the market, yeah So I'm guessing Very likely to succeed or unlikely to succeed Absolutely But again, that's a huge shot in the doc Because even humans don't know what's going to work In the field of creativity, right? Okay, advertising In music, there are many times where a band has just created a piece of You know, just created a random song Because just because they wanted to fill in an album They didn't think it was so great But it's ended up becoming their biggest hit Like for example, the band Black Sabbath They were recording the second album And they had about three minutes left In their disc and the producer asked them Do you have three minutes left? Do you want to do anything?
So they just quickly Came out of the very quick song And they put it in, they just started a rubbish The song was called Paranoid And to today, it's probably the Balabax Sabbath's biggest hit So it was like the time, I mean, humans themselves don't know How do you get a machine to judge? So anyway, it's going to create a lot of ripples And I think at best it can inform creativity And not replace creativity At a philosophical level, whether artificial intelligence can be truly creative I love this one quote that a scientist gave That a scientist gave Sometimes one man's creativity is another machine's brute force analysis I thought that was fantastic I thought that was very profound I thought that was a brilliant line So there are many models and approaches The researchers often disagree with each other Should we be psychology first? Or should we be in the neurology first? Is human biology needed?
Like, understanding a bird wasn't needed to When we developed a plane, for instance So do we need to understand a human brain before we try to develop an artificial brain? Should we use complex or simple logic? Should the robot be blue colour or green colour? All the difficult profound questions How to be answered?
So now let's get to the users of artificial intelligence In of the gap There are several places, of course, computer science In finance, for example, unauthorized use of your debit card Is something that can be tracked by artificial intelligence Trading as well, I've heard of a few services in the US For instance, that do trading via artificial intelligence Without algorithmic trading Yeah, hospitals and medicine Where AI is starting to become very important Interpretation of, like for example, interpretation of an x-ray That's something that AI can do very easily Analysis and even companion ship, especially for older people and kids These things, robots and things like an actor's companions In heavy industry, of course, diffting and stuff like that Customer service And this, Warren's a particular mention right now Because Facebook of late has been using Have been focusing a lot on their bots for messenger So now many services can you can actually sign up for If you're a brand, a service brand like a bank or an airline or hotel or something That you can actually sign up for a bot which will identify keywords So think of it as an IVR type of service It might be annoying because many times you want that human interaction But it might also be quicker because you're able to get a response much faster One pretty interesting use of this I saw was 911, the service in the US, the emergency service They've actually gone on to Facebook And I saw an entire user flow of how it was happening Help, I'm stuck at home Then immediately a bot replies Please tell us your name and address and mobile number So you enter that Alright, please describe the nature of the activity House has been broken into, I'm on the top floor of the bot and floor So it picks up burglar and the thing and says Okay, we are detecting your location And clever enough to make sense of Absolutely Any jumble of words that might mind your time And so some people ask, okay, but when a burglar is in my home Why would I, if I wanted to use 911, go through this whole process Where I can just speak up the phone and dial But then the response to that was There are some cases where calling might not be possible Like suppose somebody is broken into your home And you want to be as quiet as possible You can just send it Yeah So it's actually quite interesting Maybe you're unable to speak Whatever reason What else? Transportation, toys and games, aviation Even news and writing, there's something on narrative science Which is able to pick up Yeah, this is really scary as you are It's going up It's able to pick out like the scan through an entire article And it's actually able to provide like a little summary of the headlines or something like that On top Data mining, spam filtering in your gn instance That's pretty strong actually And a lot of stuff that is not spam Also ends up going there Yeah, so it's not perfect But if you mark it not spam, then it's continuous learn Possum assistance, assistance And that's a pretty long list And I suspect things are just going to grow more and more sophisticated We're going to get more and more users And it's really interesting time Yeah, absolutely Now the important part of philosophy and ethics So many questions are there, is general intelligence possible Is it dangerous? What if machines become sentient? And the off-coated example is Terminator Although I remember, I saw this reading a lot about AI And I think on Reddit or something Somebody said, listen Terminator is the worst example of AI If machines were to eventually destroy us Because if you think about it sending robots And killing people is actually the most inefficient way of It's a very inefficient way of killing the human race If machines, sentient, machine really wanted to wipe out the human race E-efficient, he could have just done something like biological poisoning or something That's always much more efficient Or a nuclear destroyer They'll be immune, right?
Yeah, absolutely, absolutely The question is, do we want machines to be taking moral decisions for us? And while many of us might laugh at things like this Some huge names like Stephen Hawking, Bill Gates and Elon Musk Guys like these are worried about, or worried about I read one classic case, like you have a self-driving car And you're really going fast And an old lady just sort of steps in And if the car has to decide in about 0.1 seconds Should it slam the brake and risk killing all the occupants? Or should it knock the lady off? So whatever the decision, it is a heavy one And it is one that cannot be safely left to a machine Yeah, okay, so the simplified version of saying this Is AI just does what it's programmed to do?
So it cannot be mallevalent or benevolent Wait a minute, there is one thing that I really haven't thought this through But a critical part of AI as you were saying Is that it makes its own rules, right? So you see it and a lot of the rules that it picks up Is really not foreseeable with the level of seeding that you have done So how can you guarantee it won't be mallevalent? So I guess it's what it's so the self-learning is more for self-improvement In the sense, how do I improve doing this task? But what is the overall purpose of building the AI machine in the first place?
See, I think of AI as just another tool in the hands of humans It could be useful, good, it could be useful, bad Like for example Yeah, I mean a lot of things are stick I mean you could use it to walk And you could do it to the internet for that matter Could be used for spreading knowledge or it could be used for spreading terrorism Cards for instance could be used for either taking you from one place to the other Or for killing people A brewery for instance could be used to brew fantastic wheat We ought to be used to brew filth like fosters So you know, everything is in the hands of humans And they can... I personally think of it that way But then it doesn't stop the philosophical argument What happens when machines eventually become smarter? I guess that's a debate that is what it can be The other question is the effect on human labor Ah yes, the effect on human labor Yes, that's the last thing we're going to talk about Over here Yes, that's the biggest worry I think AI's effect on human labor is just polarizing people like crazy They say, eventually we're going to have a very wealthy elite That's going to control everything And push the majority into poverty Others say, and this is the common argument by pro-AI exponents Is that it'll create new jobs Just like how industrialization came It put a lot of jobs, people out of jobs But then it just led to the creation of new ones And that is a very logical way of thinking So if AI is going to become huge Then you're going to create jobs Which are to do with improving AI Or you just become more creative, right? So a lot of jobs are there today Were unthinkable of a few years ago So this...
You feel the very feel that you specialize in digital marketing Did not exist mainly maybe five years ago And if... So you'll have people who are building AI systems Overseeing AI systems Making the AI systems Do more creative things So the potential is immense But the worry that I also possibly have Is that the pace at which change is happening So industrialization happened over a long period of time Two decades of time in fact Yeah, some jury Yeah, a statistic that I read was During the industrial wage power and efficiency doubled in 70 years Over here you have a case of computing processing power doubles in 18 months So that's a huge worry Although some people say don't worry about the pace of change etc So like I said, there's a lot of debate Not just about how to build AI But also the outcomes of AI and its effect on labor This is a lovely graph though that I saw Sorry, this is an audio medium So you're going to have to try and visualize the graph So it's a US economic value Increased with the labor force Over a period of time So both were going parallel And then they sort of diverged Which showed that as technology increased More people being put out of jobs But the economic value was increasing There's also less people who are creating more value And it's kind of scary Like I gave you an example of the creative director in advertising agency Now there are also robots being hired by law firms Now to read through Parts are not to mean level And they're even able to validate hypotheses At a level Surprisingly enough When I read about jobs that are not at risk Some of them are actually very manual You would think that AI would replace the most manual of jobs There are many manual jobs that are not replaceable Like for example folding clothes I figure was immensely difficult for a robot to do Cooking, gardening, terrain based Like if it's a job that involves navigation over terrain It's very difficult for a Third action increases completely Yeah, absolutely Many fold A lot of creative and I think interpersonal is very important I remember reading somewhere that We will reach a stage where Suppose you're a bank You may choose to offer your armjanta Artificial intelligence But for your premium customers You will still have one to one personalized interaction This is pretty interesting One thing that you definitely have to do is Education will need to change Because we will need to teach up kids to be more creative Rather than focus on facts and things like that Which a machine can take care of A lot of scholarship is written because of Google We need to know I know a poem that poems by her Absolutely I'm going to end this section We're just talking about my favorite author One of my favorite authors after he's here His name is Michael Kriton Who's a genius who wrote books like Jurassic Park And one thing about him was he always used to write about technology He always used to talk about And all his books featured some Great, funder, high funder technology And somehow the technology ended up being a villain in many ways For humans But the underlying message of almost all his books were Technology by itself is harmless Or is also useless It's how its humans who screw things up So humans can decide to use a machine A tiny machine for books That's almost like a theme of all of his books running And I think that's a great message to leave AI with Before we leave you guys We have a couple of etymologies As always The first I thought we could actually look at the word artificial itself Artificial comes from the Latin word artificialium Which means made by man And hence not natural What is not related to what is artificial? I didn't know artificial is related to artificial That seems square And to deviate absolutely The place that I've selected is a place in Mumbai That actually sounds quite posh but isn't It's called cotton dreent Cotton dreent You can imagine rolling fields of Beautiful And in monocle and so forth I know it's nothing like that It's so it's named after an art deco building Of the cotton exchange And a series of warehouses And the original cotton exchange was in Colaba In a village green A village green is a fancy term for a very big garden So that is how cotton green is My etymology is intelligence From the Latin word Intelliger Which means to understand So it was extended to mean information too Especially in the Spanish business But in common lingo to be intelligent Still means to understand quickly So that's an interesting Artificial and intelligence So very nicely And I have a profundity of the week of my own And it is this little verse by Robert Frost That I found on the internet Forgive a lot, my little jokes on the And I'll forgive that great big one on me Wow So with that lovely thought We'll leave you for the week And have a great week Have a great time Go build a little more to something Yeah, go build a little more something Yeah, bye bye