Conservative wave in the country evangelical politicians are becoming increasingly powerful and the courted by the media and the government to like in rare there is now an evangelical man in the cabinet a minister the National Council of psychologists says it will appeal against the judge's decision the Brazilian army is being deployed in one of Rio de Janeiro's largest for Veliz for Senior off to 6 days of clashes between rival drug gangs a major highway running past the shanty town was closed because of a gun battle world news from the b.b.c. Officials in Wisconsin say it was one of the 21 u.s. States that the federal government says were targeted by Russian hackers in the presidential election the Wisconsin Elections Commission said it had been told by the Department of Homeland Security that Russian government cyberattacks actors had targeted it state voters registration systems a man who shot dead 8 people at a hairdressing salon in California has been jailed for life Scot to cry admitted opening fire at a salon in the town of Seal Beach south of Los Angeles in 2011 his victims included his ex-wife and her friend who had testified against him into child custody hearing the judge described a cry as the face of evil in the community. The people of New Zealand of voting in a general election with Prime Minister Bill English fighting to stay in office after a seesaw campaign Mr English of the conservative National Party has faced a determined challenge from just India done who only took charge of the center left Labor party last month neither party is likely to win an overall majority and each would probably have to build a coalition to govern the authorities in Moscow have suffered a major embarrassment days after in vain unveiling a monument to the inventor of one of Russia's best known exports the Kalashnikov a k $47.00 rifle it's emerged that part of the memorial actually shows a German made rifle want to walk or reports the monument unveiled amid much fanfare in Moscow on Tuesday consist mainly of a statue of Mikhail Kalashnikov but Shoprite Russians bombarded social media with complaints that a metal engraving at the base of the memorial actually showed not the a k $47.00 but the storm gave a $44.00 guy full used by Nazi troops at the end of World War 2 it closely resembles the Kalashnikov the sculptor admitted his mistake and workers have now cut away the offending part of the monument b.b.c. News. Hello I'm on wait outside across and the woman welcome to in the balance the program that delves behind the business and economic headlines in this edition rarely a week goes by that doesn't involve yet another major prediction about how artificial intelligence will change the way we work and live but the predictions are hugely divided either ai is going to create new jobs we can't even begin to imagine today and give economies a whole new boost or humans will simply be replaced by robots and chucked onto the work scrap heap so this week we're asking if ai is our friend or off 0 and who's in charge of where it's going anyway we'll get to those questions in just a minute with our panel of distinguished guests but 1st Nick Bostrom is the head of the future of humanity Institute at Oxford University because there's me a couple of things so one is if Ai one day actually succeeds at it's original ambition which has always been to produce the same general purpose powerful learning planning ability that makes us human smart not just solve specific domain specific tasks then that would be a very very powerful technology that would post this what I call alignment problem that is we would have to make sure that these things that ultimately become much smarter than we are are still aligned with our values and intentions so that's a technical research problem if we solve that with one face what you might call the governance problem like with this new technology in the world how should it be used for what purposes who should control it so just as with any other technology in principle it could be misused or used for good purposes do you think enough is being done to ensure that everyone gets a say in that the issues that I present us with today I think quite distinct from loss that would arise in the longer term if it really becomes one day fully human level intelligence or superintelligent So this longer term prospect is still a basic research for the time being I think that should be. Investigations into this technical alignment problem I said and there are now are various research groups springing up in different places starting to do research on that which I think is positive and then asks we move closer then I think it becomes appropriate to have a wider social conversation as well about how we would like this future to look when you know in the limit human labor ultimately becomes obsolete right if you have machines that can not just mechanical to do everything that our bodies can do but can do mental There are things that our brains can do done then I think that's a much more fundamental change in society that would need many different perspectives and voices to weigh in on what kind of timeframe a you're looking at when you make these kind of predictions well so I think there's a lot of uncertainty as to timeline here we did a survey off some of the world's leading machine learning experts and one question we asked was by which year do you think there is a 50 percent chance that we will have full human level general intelligence and the median answer to that was 2040 or 2045 depending on exactly how you phrased the question quite soon within a life to be if you're John certainly on both sides of that it could take a lot longer and it could also happen sooner truth is we just don't know yet it's a very gloomy future you're painting that basically for people in employment I mean I mean actually you're saying that everyone is replaceable eventually by a machine if we achieve that kind of level of general level of intelligence in machines and in a sense the goal is full on employment the goal is full and even the reason why we're doing technology and innovation in the 1st place is ultimately so that we can achieve more with less if that comes to pass one day then you would have I think really wonderful challenge but big challenge of our having to rethink human culture and the human condition from growing what I think you mean to Russ Right exactly but how do you find meaning in life if you don't have to get to work every morning turn a paycheck like how can you configure a life so that you get fulfillment. Meaning even if you don't have to do anything you have to find out from inside yourself and for most people that's an extremely scary prospect right but if you look longer added I think it's ultimately possibly a really exciting prospect at the same time that was Nick Bostrom of the future of humanity Institute of Oxford University So let me introduce our guests experts in the field of Ai here in the studio with me Luis Perez Brett of the mit innovations program also sure know Haggerty executive director of Cambridge University Center for the Study of existential risk we're also joined by Kathleen Richardson professor of ethics and culture of robots and at the moment for university in Leicester here in the u.k. And by Professor Kevin Warrick He's Deputy Vice Chancellor research at the University of Coventry and Louis I want to start with you on the back of what we heard from the boss drum all our jobs are at risk I think there is no historic evidence of that they were having happened actually the history I disagree with the perception that innovation is about doing less for ourselves The argue that ever since we started doing technologies what has actually happened is we've actually learned how to do new things afterwards there are more kinds of jobs and works today than there ever been before on the other hand I agree with him and one perspective which is we now have less manual labor which means we get to actually entertain ourselves more and that might still happen these radio show movies the whole intertainment industry did not exist 100 years ago the way we see it today so it's easy to make these predictions what I would like to say on that regard is that what we do is we actually build futures we are horrible at protecting them Sean do you agree I think I agree with our specs both of what Louis says and what Nick says so we have in the past managed to automate her place a lot of what we do so there was a time of which the majority of people worked in our culture in some way right now it's something like 2 percent of the. Labor markets but we are producing more than ever before now I certainly agree with Louis that they historical precedent as we create new jobs and this concern that automation will eliminate jobs has been characterized as the Luddite fallacy time and time again we see the creation of new jobs however I'd caution that past performance is no guarantee of future performance and it's not necessary guarantee that this will always be the case particularly if we start to automate more and more cognitive tasks I would also say that I find it strange to think that we should base all our thinking around the idea that our Have life's worth and value should be wrapped up in what we do between the hours of one to 5 most the time when I hear people talk about the sort of sense of value and self-worth and their jobs it's people who have very interesting jobs like you and me it's not necessary the people who are stacking shelves 100 m. Check out I think there is enough to make a living well in principle but if we are using technologies to make our societies wealthier and to as Nic said produce more with less then there are other ways that we can support people Professor Kevin Warwick I mean I suppose the thing is the argument a lot of people make is yes jobs will be lost but new jobs will be created but what kind of jobs will be created Do we have any idea well I mean we saw I think agree a look with what Lewis said. Since computers came before computers we didn't have computer jobs but now we have a lot of jobs that are writing software or providing security for network a whole range of jobs to do with computers or in a sense it's servicing the machines and I think that's true even in the city of London used to be people themselves would borrow stocks and shares no a lot of people are involved with dealing servicing the artificial intelligence systems. That are buying stocks and shares but if I can really throw one thing what Nick was referring to that I see is a complete red herring and that is this desire or seem to get the general purpose intelligence or is as Nick was saying a human level intelligence so maybe a machine that completely replicates what you know I think what's this all about I don't see those general research going in that area at all machines have a full insuring set machines have a form of intelligence that I think in a way that is distinct to humans and we use them for the purpose they have lots of advantages and this concept that human intelligence is somehow absolutely wonderful the what is meant by human intelligence I don't know because there are also lots of different humans some with dementia some the Devonian just been born have different types of intelligence and so on so what the general purpose intelligence actually means but there's no it's some wonderful goal and then until we get to the goal everything's all right we can drink our cocoa at night we don't need to worry I don't see that because it's distinctly different that's where there is a potential threat the day long must was referring to the difference is important and forget about the human level intelligence that's a complete red herring Still I agree with you that the concept of human level intelligence is quite difficult because humans are very very and we don't quite know what we're talking about there is a concept of general intelligence and general problem solving ability that's distinct from the idea of a narrow intelligence a tool that's designed specifically to carry out a particular task but can't do anything else so you know the program that's able to beat any human chess but count Ti'iti'i lace or make a cup of coffee there are reasons to think that developing systems that would. The broader level of generality might be useful to us in many settings that might drive research towards us so if you think of let's say we had a rover on Mars if all that could do was carry out a very specific subset of tasks it's not that useful to us but if we had a system that was able to look into its environment learn about its environment and then look for interesting things figure out what sorts of experiments might be useful to do that seems like it would be very useful to us and you can imagine similar things in a lot of different concepts I feel like narrow intelligence narrow tools are going to be very very useful to us but at some point an ability to be more creative in how the systems think about problems and find solutions to them might be very complimentary to what we as humans do and I think there is some early work being done to wards more general systems and it's very primitive compared to what Nick talks about it's how can one system learn to play a you know 20 or 30 or 40 different probably game the point the point is there are still very primitive these systems I mean are we jumping the gun Kathleen are we worrying about something that actually might not ever happen it's quite extraordinary to listen to some of this conversation because I think I actually think we need to rename artificial intelligence into advertising intelligence because the reality of what most people are talking about all these amazing breakthroughs that we're talking about and intelligence have come from corporations who are better at collecting data on consumers and sending products back to consumers all these innovations that people are talking about and I can remember like when I when I started my field work at mit artificial intelligence was almost at a dead end you know there were people in lab rooms this was in the early 2000 it was kind of a kind of subdiscipline its heyday was in the sixty's and then what happened there was Internet revolution which allowed all this new kind of data to be. To know about ordinary people and then the corporations and businesses have to think oh well what do we do with all this data how can we use it to sell products back to consumers so all these innovations that people are talking about here are really driven by these businesses and corporations and it's not actually human like intelligence tool it's the outlook of the corporation who has a particular kind of interest in selling a product to a consumer that's really what we're saying but you much of the research is concentrated in the hands of private companies most of it is concentrated in the hands of these private companies and at the end of the day you know what we're actually being fed back our if you like sophisticated algorithms for us to buy certain kinds of products so I think we have to have a reality check here that what's going on in the kind of Nick Bostrom world isn't actually happening in the world of artificial intelligence and even though these corporations are talking about the ethics I think that that actual fear is coming from somewhere else it's coming from a kind of desire to control these technologies and a projection of fear into what's actually going on in society through and I have to get you to respond to that there's a lot and I agree with Kathleen and I think Nick would also agree that what he's talking about is not artificial intelligence as it exists in the world yes on the issue of this being driven by corporations and certainly the case that a lot of the research and occasions are being driven by big companies at the moment but low the people doing the fundamental research in these companies or people of academic backgrounds who wrote believe are very much motivated all of all as you know in the discussion today or from an academic background how they can use their to make the world better in various ways Tell me a good example now of how it makes the world better that isn't beneficial to a corporation or a business it can be possible to monetize nearly anything but I. What does that mean like better Florence to answer your 1st parse for example research into automation of ways of diagnosing tumors it seems like if you get that right then you can provide that sort of service to people who would never be able to afford the services of a consultant twitt 30 years of experience and perhaps that leads to health care being available to a lot more people I think there are lots of examples given or lots of examples in the medical world I'm involved with Parkinson's disease and the use of artificial intelligence the dog knows the particular type of Parkinson's disease even predict when the disease is tremors are going to start and muscle locking is going to Louie's it wants to give an example yeah there are endless examples today of what you're doing thanks to our deficient even addiction but we're talking specifically about ones that aren't commercial Yup I use Google Maps to come here today I was late I had to walk but it was a protocol you know I didn't pay any money to Google I was signed out from it so Google did not drive in a particular benefit from me I did derive a benefit from artificial intelligence when driverless cars come around if they do and they are really that great and 100 percent efficient and safe and safe then you will be making a lot of money off your car lift will manage it for you your car will be driverless You can also think about that future the point here is really we get to build that future it's not built for us but the fact remains that there are people already in the process of losing their jobs as a result of advances in robotics in artificial intelligence but I don't think that that's actually I think it's an intelligence that's just robotics Intel Tomasz we're mixing a lot of things together data mining analytics big data machine learning and I'm sure if we can call all of those things as if they already existed today as artificial intelligence you know I guess we can't the point really is that the jobs that are being lost to the world are not lost to the machines someone. At the level of management is making sure that this issue and typically cost cutting decisions and they've decided to go for Sheldon route of automation it could also have used the same resources to think about new jobs to create but that's a management business decision has nothing to do with the robots the robots don't go in the end of the strike the not take the jobs from anybody rather it's just automation we've been doing that for ages the whole middle class of today is the byproduct of automation that started in the early 20th century so we have automation to think for the world we live in today I think back to this idea about unemployment you know the it the dream is fallen unemployment you know Nick made it sound as though he'd just come up with the idea or it's current today but in the early 1920 is there was an armed group called the dark Doris who basically wrote a manifesto of the future and in that manifesto they said We campaign for the systematic deduction of labor through mechanization and it was a dream in the 1920 s. That you could create this mechanized society so you could reduce the necessary labor time that people would spend producing at the goods that they needed and then the spend all the other time you know reading poetry and painting pictures that all men of caves dream of this you could be a fisherman in the morning and a philosopher in the afternoon and I guess what we're doing is we're comi