okay so what I'm going to talk about ties in nicely with what Fabian just said before I do that let me just say that talking about the singularity is actually really hard or close to impossible without you know just waving your arms around and getting all excited about the big idea that you're that you're talking about there and for me the reason why that is the case is similar or you know it's analogous to the reason why it's hard to talk about black holes and the people who first introduced the term technological singularity which is us as far as i know is vernor vinge and write code for they really borrowed this term from relativity were you know the singularity is the black hole and has two interesting properties first of all if you're at a safe distance and observing the system you cannot really exchange information with it you don't see what's going on there you don't see beyond the event horizon you don't really really have no way of knowing what's going on
in there so that applies to the technological singularity I think and the other one is as you're speeding towards it you don't notice the event horizon there's no flash of light or an event or nothing you just go towards this big unknown at increasing speed and I think with the technological singularity we are pretty much in the same situation so the question when is it going to happen lets you know give it a date is really pretty much meaningless bye-bye principal reasons we can't know so we've also mentioned five and has also mentioned at the top notion or the popular culture notion of the singularity of technological singularity which goes like this there is a team of people they are very bright they do some research and development and at some point they make an invention or you know do something specific and then you get this real quick exponential take off and this intelligence explosion and usually it's also you know visually spectacular because the people who are making the pop culture here are
out of Hollywood and you know need spectacular visuals and I tend to think that this notion or of you know the creek heart take off singularity is sort of misleading but that's fine because it's really hard to talk about and what I'm going to show you now is just an alternative way of looking at the same set of ideas and doesn't necessarily claim to be more true than that so at the heart of all this at the heart of the idea of the technological singularity there is really the idea of positive feedback loops from cybernetics and this works like this you have a system and there's dynamics in there and you can observe two variables and there's a positive feedback relationship between the two variables so you did not chop the one thing the other goes up and vice versa and you know if you put some activation in there like you know you increase the number of cattle running the panic increases which causes more cattle to run and eventually the whole herd is
panicking and if you plot either of the variables you get this nice exponential graph or if you're in an actual system and you know in an ecosystem or you run out of cattle so you don't actually get an exponential graph but more something like I said like a sigma light so it saturates and the idea when we talk about the technological singularity is really we have the same sort of relationship of mutual reinforcement between technology and intelligence so there's some intelligence that creates more technology you know improves on itself and you know this feedback loop gets kicked off and now let's go from high school cybernetics to furniture this is the no veto cabinet done by a guy called David Rankin in 1777 it's on display at the Queen scribble Museum in Berlin and it is obviously elegantly beautiful but it also does this and now let's hope to be actually place does it oh yes it does move so there is lots of mechanism this is just one of them in there and things
happen and then you get this its transform of furniture bang okay so why is this in a museum apart from it being very beautiful before David when caen furniture didn't really look like this he was the first guy who was able to build something like this and for one of the reasons is he was very talented and the other reason is he was allowed by special permission to employ 80 people not only three in his workshop so weary he could really could put together of the brains of Engineers and artisans to make something like that why did he need special permission because there was in Rococo times something called Solvang so in order to manufacture anything you needed to be part of it soon and this meant you had to follow a set of rules one of which was you couldn't employ more than four people and that leads to the question why did I do that why do people put this regulation in place you know you have not solved lang cool transform of
furniture we have solved 100 guests desks so why were you doing this where they're stupid and I would say that probably not whenever somebody tells you that people in the password stupider than us they are probably selling an idea to you that you shouldn't buy this was a deliberate choice by then and my interpretation was they were doing exactly this they were preventing the feedback loop between intelligence and capital from being kicked off this is a deliberate feedback damper that prevented the reinvestment of money that you got from making something really beautiful into more technology and something interesting happened since they did that if you look at the picture on the right you see something that looks like a pyramid and on the bottom you have people who create food and then there is an there's a there's a layer of people who provide for security and shelter and then there's a couple of clerics for the more emotionally needs and then you have a couple of nobles and kings and they essentially paid yeah
aditi artists and if you look at this and you've studied economics then this looks like the famous pyramid of needs so in Rococo society pretty much everyone makes food in our society does anybody know the number of the percentage of people who are still in producing food [ __ ] okay so you know the number i only know the number for the primary sector which is 2.1 so absolutely nobody is making food what's happened in that time my take on this is techno capital happens the combination of you know capitalism and increasing amounts of technology in the system now in philosophy they have a concept called deterritorialization by yielded arisen phillips qatari and that is the process of detaching people from all their old context getting them off their land that has provided sustenance so far putting people into the city's putting them in contact with ideas putting them under competitive pressure and make them do amazing things and another word for that you know it's a positive process in a way is liberate them from the
work they had to do and that's important to know that you know this puts a lot of stress on people and if you notice this there is there's nostalgia for times before for merry ol England and the grass was greener and the hay smell sweeter and all that and nobody is really doing deterritorialization nobody is responsible for you know it's not the Jews like the Nazis would have you that are doing all this capitalism thing or it's not the bourgeoisie order cable to the capitalists like the Soviets without you this is something that happens like traffic nobody does this this just happens to us it's a thing on its own it's an abstract dynamic there's a couple of names for that even you can call it if you want to be very fancy d territory ization and you can also just call it modernization or rationalization I like to call it capital or techno capital is really the same thing and why does it happen Fagin says because we're lazy I think it goes deeper
it caters to our needs and that's all of them it includes you know the real monkey brain stuff so if you ask the question why does it happen what drives it I think this question is actually equivalent to the question why do we buy a fancier cars or why do we buy a fancy or smartphones yeah they're useful what we're doing other things with them you know we we transformed with signal startles we attract mates we do all this this sort of thing so the monkey brain is really what fuels all this high-tech stuff that we're doing in the end okay so one point what was it two percent of people are in food what is everybody doing some of us have become a dressers most of us have not in 2014 forty percent of the american workforce was white collar jobs and white collar jobs what they what do they do they manage complexity that's really what it's about they manage complexity and the automation that we have been placed for have put in
place for producing things and managing complexity and managing the automation is actually a very good description of what intelligence itself is that's that's what intelligence is for so what techno capital does in this big sweeping process 300 years process of deterritorialization is it extracts intelligence from the land that's the rare resource you know make people use their brains that's what's going on capital drives people into the cities to use their brains which is where intelligence automation comes in so some of this can be automated and Fabian has you know named the topical examples of intelligence that has been automated like making calculations storing and retrieving information driving cars playing chess playing go winning a jeopardy you name it so all of this is happening and if you ask the question just this process is probably going to continue if you ask a question which jobs will be replaced my answer would be pretty much anything that you can automate will be automated a eyes are cheaper they're more reliable they scaled really well and you
are in a situation where they're also a mate they're just less winey you know people go and strike and this sort of thing and a eyes never do that ok so now that this is happening what's the singularity going to be like I might take would be I don't really know but I would be surprised if it would be the hard takeoff scenario that's not you know how anything in this continuing evolution that we have seen ever happened not what you get what's going on right now is that there are many teams across the planet they're all smart they're all making inventions and none of these inventions is you know producing the heart take off that's why having a system on google that you know answers emails is not kicking off the heartache of singularity so we're rather going to get that will be my take is an accelerating and accelerating evolution of intelligence is that compete with each other sounds like good news for the humans whoever that is I don't think it is
we will be less and less relevant in that scenario because accelerating means that you know they believe you know they've what they believe olve faster than we can even understand what's going on between the intelligence automation systems okay so do we conclude I don't think anybody has any reasonable crew as to how close we are to the singularity or when it will happen it tirmidhi the question is almost meaningless you know it's a process that is going on and it's been going on for centuries that that we're in and what we do know is the feedback dampers have been long removed you know that's not just sort of slang it's a lot of things like the ban on capital on credit lots of things that were in place to prevent exactly the world that we're now living in have been removed and it has been kicked off so it's happening and you know various attempts that have been made in history to stop it ended with us killing each other big time but nothing else
happened the reason we're talking about the singularity now is probably not that you know we feel it approaching or we have any informed guess that it will approach soon but just there were major breakthroughs in machine learning three years ago which triggered a lot of science fiction which triggers the interests that we now all have in the topic so it's a bit fashionable right now right the other thing that you can take away from this perspective is optimizing for intelligence whatever you do is probably the right thing to do so whatever you do ask yourselves are you optimizing for intelligence in that process because if you do that you're pretty much aligned with history and the real takeaway is intelligence is not artificial intelligence or intelligent automation is not a technology or an invention like Wi-Fi or rockets or computers it's more like the ultimate mega trend the thing that capital wants to do right it's the telus of techno capital in a way okay and I'm going to conclude by reading the first paragraph
of a piece by a philosopher called Lakeland who worked at the University of Warwick in the 1990s and he certainly wasn't to note the first to notice you know this dynamic and this reading of history but he was certainly the best writer amongst those who did the story goes like this earth is captured by a techno Cap'n singularity as Renaissance rationalization and oceanic navigation lock in tacoma dome monetization take off logistically accelerating techno-economic interactivity crumbles social order in Auto sophisticated machine run away as markets learn to manufacture intelligence politics modernizes upgrades paranoia and tries to get a grip the body count climbs through a series of globe wars emergent planetary commercial trashes the Holy Roman Empire the Napoleonic Continental System the second and third reich and the Soviet international cranking up World Disorder through compressing phases deregulation and the state arms race each other into cyberspace by the time soft engineering splitters out of his box into yours human security is lurching into crisis cloning lateral G no data transfer transferable replication and cyber erotics
flooding amongst the relapse on to bacterial sex neo China arrives from the future hyper synthetic drugs click into digital voodoo retro disease nano spasm beyond the judgment of God met down thank you yeah I mean I tend to think that the self-awareness is a different thing so that's a separate system you buy you build for very specific reasons you don't need a system that self-aware to rule you right that's the other thing is the transfer the you know the the DS realization if you if you if you so want we need ideas there I think there are a couple out there and there are people who aren't just doing machine learning but are trying to integrate machine learning into systems that can do real artificial general intelligence my group happens to be one of them I'm not claiming we got it right or we'll get it right but you know it's it's it's an active field of research and it's ongoing we're not quite there yet it's really hard so you know we do simpler
stuff for money and too or just knowing where we go aligning with the trends here my guess is five to ten years before you get machine learning that actual actually produces representations that can be used in different context because that's what you're what you're really asking about yeah yeah okay so the question was how can we get the data out of the data silos that we have created with the big process of centralization that we have let happen in the last ten years on the web so we would on our data and we would own you know what the AI is do with the data I don't have answers but I agree it's a very important problem you know there are technologies out there that these centralized that trying to read a centralized the web I happen to be a fan of orbit I don't know if you've seen that one so there I our ideas but you know these are nerd ideas and you need to get the people who are on facebook to
follow otherwise you will end up being in the asperger their famously lots of nerds or nobody else so whether that happens or not is you know up two different up to different people i don't know i don't know i haven't I agree it's superb