Automation, AI & Work
What will automation and artificial intelligence mean for the future world of work?
For more information on Aleph Insights visit our website https://alephinsights.com or to get in touch about our podcast email podcast@alephinsights.com
Transcript
Hello and welcome to the Cognitive Engineering Podcast produced by Tell Me Studios for Aleph Insights. In this series of podcasts we take a look at interesting topics and discuss what we think they tell us about analysis and decision making. I'm Fraser McGruer and I'm here with Nick Hare and Peter Coghill of Aleph Insights and this week we're discussing what automation and artificial intelligence are doing to the world of work. So Peter can you start us off and tell us, answer this question for us. What are automation and artificial
Speaker B:intelligence doing for the world of work? So I'll preface with saying that I'm a technology optimist so I may have a rosy view on things. The automation intelligent agents of the form of machine learning algorithms up to more complicated general artificial intelligences are likely to become prevalent in the coming decades and as we already see we are devolving to narrow artificial intelligences various things that we used to do ourselves. So a good example which you may not think of but the DLR, every time you get on the DLR you're putting your life in the trust of an autonomous system which is making decisions about when and where to stop trains so they don't crash into each other. Same with Google, every time you search query you're relying on an artificial intelligence to prioritize the results you get to give you the best results that fit your term. So we already started doing this, we're putting in trust in autonomous agents and I think this trend will continue. More and more agency will be given to these things to do things on our behalf and possibly to the point where more complicated life-death decisions are being made on your behalf. So a very near example are autonomous vehicles, so you'll soon be careering down the motorway at 70 miles an hour in a car that's completely controlling itself without any input from the driver. I think this will creep in and we'll become more and more comfortable with this and these autonomous agents become more and more capable of doing more and more things for us. Okay so a paradise of technology awaits us. Yes, as I say I'm very optimistic. Okay no downsides. Are we going to go on to that later? I think we'll probably touch on that later. I think many of the objections to this are a modern form of ludditism, a fear of technology from taking jobs and a fear of moral and ethical reasons and a hankering for a previous skewed view of what the world used to be like back in the day, a sort of past oralism, a view of a world that never actually
Speaker A:existed. Okay we can euthanize all those people who think that. Well no I think they'll just
Speaker B:sort of wake up. They'll die off. As many of the luddites, they just took on different jobs, they didn't expire, they and their prodigy just went on to do
Speaker A:different things, took different jobs. Okay I want to hear from Nick but just before I do, we've all have, actually I'm going to do, I'm sure Peter definitely does, but do you have a calendar, a Google calendar or similar to run the family calendar? Do you have that Peter? No we don't
Speaker B:actually. My partner is resistant to, she still has a paper diary, she likes it, mainly because of her love of stationery rather than because it's a practical way of doing things but yeah
Speaker C:she's not yet fully digitized her calendar. Yeah a constant source of annoyance to me that we don't have such a thing but she's got a file of facts, would you believe. Right okay. Can you still get
Speaker A:file of facts paper? Yeah you can, I see it all the time in XX Minute. No well I'm actually very very surprised, especially with you Peter, but I'm sure you're sort of, you're flying the flag. I'm chipping away, I'm chipping away. Because even in my household for seven years now I've been saying to my wife look we should get a Google calendar to manage all this but she insists on using this paper calendar in the kitchen which just drives me mad because we just, you know. It doesn't
Speaker B:automatically sync with your Google calendar. I think this is an interesting thing, maybe a subject for another podcast but for them it's convenient because it's in their pocket or their handbag all the time. They don't experience the inconvenience that it has for everybody else that needs to needs access to that information. Yeah, god wives and partners
Speaker A:they're so selfish. Okay so Nick, do you chip in here? Yeah well I suppose I want to present
Speaker C: always adapted so you know by: Speaker B:value better than some machine. But to continue the optimism vein I think that... That wasn't optimism I was talking about turning people into goo. My optimism vein... Because if you think that's optimism I'm terrified of the world you... No if you, Robin Hanson and other writers that have made predictions about what a sort of artificial intelligence driven economy could look like they generally on the positive side if it's an entirely automated economy that the rate of growth is huge potentially doubling every year rather than growing incrementally by a few percent every year doubling every year. So the amount of production far will rapidly far out the human races need to survive. So there's so and this if managed well there's just huge amounts of surplus of everything so that you know which will which will lead to a could lead to a huge increase in the sort of social state kind of arrangement where actually we are just we just sit around not doing very much and live off this machine this global machine that produces stuff for us and every desire that we ever can ever imagine is fulfilled by this machine. All our only job is coming up with new desires and everything is taken care of. So that's an opportunity. I think that there we start buffing up against like natural resource restrictions and an amount of energy we can produce but assuming that the world population is allowed to continue to grow at its current rate that there may be a problem but if because we're not because we're everyone standing and living increases sort of European US level and everyone starts having 2.4 children in a family then the population growth will tail off and we can easily sustain that.
Speaker C:Okay so look let's try and simplify and that's an optimistic vision apparently which sounds slightly worrying to me for a number of reasons. Not least there's the implication there of a really substantial change in our approach to government policy and particularly we've generally over the last hundred years moved towards this default assumption that the best way of allocating resources is through markets where resources are effectively infinite markets no longer do that but someone has to distribute those resources there has to be a mechanism for distributing resources. So I mean I think there's a suggestion here that actually we will effectively be all living on a kind of automated welfare state where machines produce stuff and we get given a portion of that. I mean that would be quite a substantial change and so I think we will be and there are also going to be huge transitional costs entailed in this vision and then there's also the question of whether or not whether or not humans who as we know are all sort of quite flawed and irrational would actually be happier living in this kind of paradise when constraints are removed because you know the evidence is that humans can't don't behave very well in the absence of material constraints you know we tend to find it quite hard to manage ourselves. We have evolved in a environment of scarcity and when you remove that scarcity we don't really know what might happen.
Speaker A:Okay look there's so many this is such a broad subject there's so many different avenues we can go down here. I want to narrow it down a bit but before I do that what you were saying there reminds me of the matrix where they said look we created this world but it just didn't work and everyone knew there was something up it would have didn't it didn't people didn't believe in
Speaker B:it because everything was perfect. Humans can't just can't live in a utopia. Well it's more like
Speaker C:Wall-E. Have you seen Wall-E? No I've not. So it's about a where earth has been abandoned because it's an industrial wasteland and it's about a little robot who's kind of still fulfilling his role of going around sort of tidying things up and and looking wherever he can for evidence of new plant life for the humans to return. A spoiler alert here right and then they find a plant it takes it up to the humans who are all living in orbit and they don't want to return because they're all they're basically all they do is sit on these you know motorized chairs and watch telly all day and they're not remotely interested in coming back to earth and living in that kind
Speaker A:of you know natural world anymore. Sounds good. Well look there are so many different directions we could go with this but there's one direction I want to go in. So one of the things we talked about is this kind of narrowing of the gap where humans are useful and let's for want of a better word are better than something artificial intelligence can do. So one area that we've talked about before is let's call it in the area of creativity. What was that awful tv program script written? Do you mean Sunspring? I mean the daring experiment Sunspring that produced the most awful uh script I've ever seen and I think one of the things that we talked about and we talked about in other podcasts as well so I think one of the things we kind of all agreed on was that actually there are some things that humans will always be able to do better than than AI okay and and it seemed to be used in this in the area of creativity. Now first of all that may be right or wrong but let's assume for a moment that that's correct but one of the things you said earlier Nick is well even if that is the case what could be let's say the economic value of something like that but is that let's pursue that is there an area where humans will always be better at something and that that can add value and that is vital to the system playing out?
Speaker B:Yeah well I mean I think I think you're right I think it's going to be a long time before we have artificial intelligence which approximates the level of creativity and the uniquely human things about creating new things having new ideas um but I think where you know where where humans will be uniquely uh good at is being human and and doing things that machines just won't be interested in doing um even if they have complete agency over themselves so you know we may we may it may it may be that we we have some sort of cooperative alliance with these uh with the entirely autonomous agencies where where they devolve to us creative tasks or tasks that we are only good at that they get value from and vice versa so that they will suffer a scarcity of humans uh that we can provide.
Speaker C:Yeah I mean I'm not I think the the we've got to be careful of not confusing um just contingent things things that happen to be the case with things that are sort of fundamental here you know there's nothing fundamental about necessarily adding value to stuff like creating stuff I know you know doing a job actually doing a job where you're adding value um is something that we do because things are scarce right and if we take that scarcity away you know there's no why should we sit around creating stuff if you know if machines are good at creating stuff themselves but I think to take a slightly different angle to Peter there being human you know is is actually in itself something that people value when they're in relationships and I don't mean you know just sort of romantic relationships but people just like interacting with other humans for its own sake we've evolved to do that we like it for the same reason we like going on roller coasters or listening to music it's a it's just a fundamental need that people have and um and I think you know to take that a little bit further you know people will like things because they're made by a human you know uh that's intrinsic it will have intrinsic value um you know people will like watching a play the piano uh because it's nice to see a human do that uh you know it's kind of easy boring to watch a machine play the piano um the first you know after after the first few hundred times
Speaker A:okay so look we've gone a little bit abstract there um so let's sort of bring it back down to earth a little bit and go back to the initial question so um what automation and artificial intelligence are doing to the world of work so we've sort of gone quite far in the future but let's sort of bring it back to the here and now um and thinking about our own careers let's say the next you know one to five years um how how is it going to impact on us um does anyone got any
Speaker B:information any thoughts on that so in the world the world that we live in the world of analysis and processing data and and helping formulate rational decisions based on available information i think that that's right for um an intelligent assistant there's a lot a lot of this work involves very mad very uh tedious manual uh reproduction of data uh sanitizing data cleaning up data which if which can could be automated many much and much of it is beginning is being automated now so i can imagine a a desk analyst who's looking at a part of the world or analyzing a market or keeping an eye on stocks uh will have an intelligent assistant that they can task to do things for them to help them to speed up to augment their ability to explore new hypotheses generate new and and generate data to test those hypotheses okay nick uh yeah i mean i think that's
Speaker C:that's right in in in one sense everyone's job is is transitory um you know everything is vulnerable to automation in the sense that you know you're doing your job you're using information you're making decisions everyone's job is like that and you know in principle all of that can be automated um as peter says though part of a lot of what we do involves taking a mess which is the real world and helping people understand it um and uh at the moment there's the real world is not remotely quantified enough to be subjected to a kind of repeatable process um so you know where you find uh that machines currently are doing a lot of the work is in things that are highly quantifiable so things like um uh you know face recognition or um uh automated share trading you know things where i think where it's quite easy to measure and to quantify and to model um in in nearly all of the real world the real world that people have to make decisions in in business and in government is just nowhere near that you know we we have a tight usually tiny little fragments of data uh about things and actually most of it is is is still reliant on humans under looking at a system and understanding it in a way that is not true it's not a sort of data-driven process it's a systemic sort of approach so so i think you know we we are you know inevitably are part of the process of automating the world um but there's a hell of a lot that needs doing before we get
Speaker A:there okay um i feel we're coming close to a conclusion here um i'm just trying to think with my own sort of uh field of work of it can be quite difficult i feel looking at one's own work and working out how that could be altered by automation or ai um it's difficult to be objective but just if we just break it if we just you know i do many things if we just say you know i'm a filmmaker um just thinking off the top of my head there might be um in editing it could be very useful um and but the first the main stuff that springs to mind to be honest is is stuff that's right for automation which is things like the marketing of my services things like that and sales but the actual creative process is i find it a bit more difficult maybe with choosing shots there could be some stuff maybe with script writing there could be um so i mean in but in
Speaker B:i mean i've witnessed this when we set up for these podcasts there's quite a lot of faffing around with machinery the mic stands and things to get it ready you had a if you had a little robot that just came in with a mic and just put it in exactly the right place got to set the levels for you would be very easy to do um that would be half an hour to an hour worth of setup time that you would get back to do whatever you wanted check your facebook profile write a write a small play
Speaker C:do something useful something something and just i mean you know to illustrate how far away we are that we are nowhere near having robots that can set up mic stands the complexity of getting that mic attached to the chair in the way that it is is huge and we don't there are no machines that are going to be generally capable of doing those kinds of things um for for you know a few years
Speaker B:at least um i think i just well you're constraining it by giving it the same mic stand i'm imagining a little robot that is the mic stand it just comes in oh yeah no that sounds quite nice
Speaker C:with a chirpy little voice as well yeah ready to record now sir yeah taking you're taking your levels and i think um i but i think i think what we you know that the issue here is the sort of transition um is going to be pretty painful for quite a lot of people i think you've only got to look at you know the an analogy of what happens with the sort of mining towns you know that actually um you know that let's say it's economic progress you know miners were no longer effectively destroying value by you know mining things that were worth less than the cost of mining them um and you know but what was the result with having uh you know a lot of people living on welfare it wasn't it wasn't good and it wasn't not good simply because there was sort of material disbenefits it wasn't good for social reasons that i think we don't fully understand but which we will probably have to face questions about in the near future but that those problems were
Speaker B:caused by the fact that their value the value center that they were attached to was physical you know the the the coal was in the ground in wales and the people had to live next to the coal to mine it whereas in where more people now are in the information space where their where their job is to process one kind of information into another kind of information that there's no reason why most people in the city can't work from home yeah i know but it's just more the
Speaker C:fact that people derive value from being feeling like they were doing something useful i mean people divide derived psychological value from that quite apart from the whatever economic value
Speaker A:was produced well that paints sort of two different visions of the future because i can imagine you know riots and the battle of hoxton square as we've got thousands of disillusioned um hoxton um shoreditch hipsters who no longer got their jobs as web developers and designers so making cereal breakfast automated craft ale production yeah yeah so we've got that estate agent rights
Speaker C: of london yeah yeah: Speaker A:there um suffice to say that in this imperfect world although i don't yet have a robot to help me set up my mic stands um in the meantime i do have um a graduate of engineering namely peter coghill to help me with my massive exploitation and he he too also has a chirpy voice um so gentlemen thank you as always um you've been listening to the cognitive engineering podcast i'm fraser mcgruer i've been here with nick hare and peter coghill thank you as always for listening until next time bye
