tv Talking Business BBC News January 7, 2017 8:30pm-9:01pm GMT
one of the guard surprised him by posing for a photograph alongside the youngster. the video were shared online by one of the guards involved. with milder temperatures reaching the uk this weekend, below—freezing temperatures continue to sweep across europe. in italy, sub—freezing temperatures were blamed for the deaths of a half—dozen homeless people. heavy snow and high winds resulted in re—routed flights, delayed ferries, cancelled trains and closed roads. at least ten people have died in the cold that has gripped poland in recent days and several greek islands have been blanketed in snow. in what let's cross over to the weather. not as cold as we are seeing in other parts of europe. certainly not. there is a change to something at the weekend. and the moment, it is mild, a murky night
tonight, with extensive fog the hills. clearer skies are times. eastern scotland and north eastern england, touch of frost here. but temperatures mostly above freezing. another fairly temperatures mostly above freezing. anotherfairly dull temperatures mostly above freezing. another fairly dull day. extensive low cloud and craig first thing in the morning, patchy rain and drizzle. a few cloud breaks east of wales, north—east england, east of scotland, but for most, stick with the cloud. wetter in western scotla nd the cloud. wetter in western scotland later on. but once again, temperatures above where they should be for the time of year. as for the weekend, prepare forsomething be for the time of year. as for the weekend, prepare for something more winds to re—. all in all, it turns colder. not quite as cold and snowy as eastern europe but some of you may see some snow by thursday. see you in half an hour. hello, this is bbc news with nicholas owen. the headlines at 8.30pm.
labour has accused the government of plunging the nhs into crisis through lack of funding. it comes after the british red cross warned some hospitals were suffering a humanitarian crisis — a claim rejected by nhs england. a us army veteran is in custody after five people were shot dead at fort lauderdale airport in florida — a british—born great—grandmother was among those killed. more than a0 people have been killed in a bomb blast in northern syria. it's thought the islamic state group could be to blame — they're not part of the recent ceasefire. and wayne rooney has equalled sir bobby charlton's record as manchester united's highest ever goalscorer, scoring his 249th goal in today's fa cup match against reading. now on bbc news it's time for talking business. imagine a future where machines permeate every aspect of our lives.
that reality is frankly already upon us. artificial intelligence is in our homes, in our cars, even in our pocket. but, how will artificial intelligence or ai changed the way we do business in asia bastion mark and what challenges will it bring? all of that and more on this edition of talking business. welcome to talking business. there is so much talk about artificial intelligence these days but what does it actually mean? and deep learning, another phrase
that is often thrown about. we'll be speaking with a panel of experts in just a moment but here is a quick refresher. artificial intelligence is a blanket term for any kind of technology or software that enables machines to perform tasks on their own. it is very much a part of our everyday lives. from something as simple as automated factories to intelligent personal assistance like google assistant, and basic customer service robots. but ai is slowly evolving into something much more complex. deep learning is one branch of machine learning that teaches computers to process information on several different layers at the same time, much like neural networks in human brains do. part of that is for machines to learn for themselves by the accumulating and extracting
patterns from huge amounts of data over time. but the eventual goal is for machines to incorporate aspects of human intuition and emotional intelligence in their decision—making process. as we teach machines to be more independent of human beings, we are also asking questions about what could potentially go wrong, and more importantly, what all who the smart machines of the future will be held accountable to. artificial intelligence or ai is already being used in many industry and business processing today. deep learning within ai is really the next frontier. but what kinds of possibilities does this open up for businesses? with us we have modar alaoui the founder and ceo of deep learning software developer eyeris.
it makes, amongst other things, eyeris is a deep emotion recognition software that reads facial micro—expressions in real—time. vasilios vonikakis, a research scientist with the advances digital sciences center. and the chief technology officer of coin tech, dealing with deep learning and automation in the transport and logistics industry. thank you, gentlemen, forjoining us today, to talk about artificial intelligence and deep learning. i'd like to start with you, modar. what is deep learning? how would you describe it to our audience? in image recognition, deep learning enables algorithms to recognise the same face from different angles, and in text recognition, it would try to identify the text in different handwritings.
in, for example, speech recognition, it would try to identify the same keyword or phrases from different accents. so, it needs a lot of data to be able to do this. typically, a lot of data but a lot of quality data as well so notjust quantity, but also the quality of the data. yuris, in your business, can you explain how you use deep learning in a real—life application? yes. so, we do a couple of things, basically. one of the things we do is we do recoup addition of passing trains and different kinds of things about it, what is normal, what is not, starting from very simple things like numbers, which was initially how we started it. and, then, moving on to identifying different kinds of issues. lately, we also use this to identify things which, from video feeds, what is happening in the city, like how many pedestrians, cyclists, cyclists with helmets,
cyclists without helmets, with child seats, without trial seats in the back, transport all the information you can get from any video stream, transferring this into actually valuable real—life data. vasilios, how do you see deep learning being useful in asia? because you are working here at the advanced science centre in singapore. how far along has the research and the technology become advanced? basically, there is a lot of research in this field going on in the local universities and in many different research centres around the area. i have seen a lot of start—ups, actually, utilising this type of technology, start—ups that have to do with observer recognition, in order to help shoppers find similar looking items from a big database and going directly and buying this item online. it has also been used in autonomous vehicles, a lot. this is a very ongoing research in this direction. and in many other cases.
modar, where do you see the most advanced usage of deep learning in asia? image recognition still holds a number of challenges, simply because we cannot map the entire world in a very short period of time, nor can we do it that efficiently. so, i do believe that is an area that is also ripe for research. not in asia, but in the world in general. what i'm interested in is the deep learning that is trying to break. because i understand having developed from machine learning, as a result of the limitations of machine learning, would it be fair to say that there are limitations in deep learning as well?
i would say that deep learning isjust the next step, not even close to full artificial intelligence. the limitations set there by humans. the limitations are set there are regarding the ideas about what are the tasks that we want deep learning science to achieve. do we want to recognise images? do we want to recognise speech? what kind of images are we searching for? so, it is not about making decisions. it is actually about doing specific tasks at a certain level, so precision is also important, to understand what is relevant and what is an acceptable level of precision, for example. so, to follow up on that, how is it being used in real life? if you can talk specifically to some industries, what kinds of functions or what kind of people is deep learning replacing? yes, some examples of the things we are doing. like, when you talk about on the railroad, passing over the train or the wagon between two companies or two motocross and points or something like that, we actually have a bunch of people running around doing different kinds of checks,
visually checking it, doing this stuff 2a hours a day, seven days a week at all the time. so, what we do, we actually managed to replace them, to take the necessity of way, and this is a very simple and but there are still thousands of people who are actually doing these things. so, vassilios, currently, deep learning isn't taking over the entire supply chain of the labour force employed in these industries. do you see that changing in the few? do you see that changing in the future? do you see more businesses employing deep learning and replacing people with this technology? yes, i most certainly do. i think that i can see in the future a lot ofjobs being displaced by ai. so, many people are talking about new jobs will be created.
yes, for sure, but i think the rate at which this happens will not be as great as the rate with which the jobs become obsolete. so, for example, if you take autonomous vehicles, actually went this bull happen, you can say that a lot ofjobs will be displaced there, like truck drivers, taxi drivers, for example. and this isjust the beginning. yes, i can add on top of this as well, a lot of jobs like security. you talk about security guards in the malls and everything else, that is another industry where there are a lot of people working for it, they are considerably not the highest level of jobs, they are not that technology intensive, and these are very likely to be either replaced or optimised as part of applying more and more technology solutions. modar, do you see this becoming a sort of long—running industry trend for the future or will deep learning, as we have been discussing, hit a barrier and another breakthrough has
to come out? as in everything, technology will definitely be a barrier, and a new technology will come out. just like what we have seen with machine learning and other technologies before. but i do have to say that deep learning is still in its infancy, there is still a long way for it to mature, and to disrupt a number of industries and lives before we can see it being potentially update did buy a new technology. exactly. so, as it is now, deep learning requires a lot of data and a lot of good quality data, as you mentioned. exactly. so, as it is now, deep learning requires a lot of data and a lot of good quality data, as you mentioned. so, many people are working towards making this technology able to generalise better with less number of data, a smaller number of data. so, that is one thing. how can you train in the neural
network with fewer examples in order to generalise as well as it will be with a large area of data assets? another direction is how to make these deep learning networks learn by themselves unsupervised without having to tell them exactly. but on that point of unsupervised learning, we have seen incidents already just recently of when unsupervised learning or an attempt to make this kind of technology more independent, if you will, go wrong. one example that comes to mind, for example, the hate tweets. that is an example of trying to educate a chat bot, if you will, to do a certain thing, has ended up having the adverse and opposite reaction. supervised is here to stay they because at least we know that, you know, ground truth data is being labelled basically by human beings, therefore we are stirring up
the level of accuracy there. unsupervised is still going to have a certain number of question marks around it at least for the foreseeable future. there is a lot of effort being conducted there. one thing i wanted to add to this also is that it is notjust that deep learning technology which has to advance, we also have to add arms the way we connect with the world. the ability to interact with people, the ability to see the world, so we talk about all sorts of vision development which include also sound, image, video, whatever you are processing that you can actually see. so, it has to be hand—in—hand to become better, so it is notjust one field which we talk about. hold that thought, gentlemen, we are going to be talking about deep learning in just a few minutes again but before that, let's take a tongue in cheek look at deep learning with our comedy consultant.
deep learning, it is the latest phrase that i'm going to impress people at dinner parties. but it will have applications that might actually surprise you. i've come to the seat of learning here at trinity college dublin, to find out about one of the surprising ways in which deep learning can be applied. deep learning, it is a great phrase, isn't it? it combines two wonderful words, learning, what a wonderful word it is. you take away key learnings from the meeting, use it at the foot of the master, learning lots of wisdom and deep, deep thought was the famous computer in the hitchhiker‘s guide to the galaxy. deep blue was the first computer to beta human at chess, so combine the two of them, it has got to be good, right? it certainly sounds
better than careful, now. let's meet dr ahmed salim at the connect research institute here at trinity college. he's developed, along with dr mohammed elgary, a portrait painting technique which is based on deep learning. so we have this van gough? this is how frank gough would paint me. yes, exactly. what we have here enables you to photograph and the van gogh painting, the model by itself, produces these results. look at me. a masterpiece. we are using these models to do feature extraction. then we upload these features into our mathematical model. it will generate the painting that we are looking for. professor linda doyle has a view on the bigger picture. actually, it is true, the eyes do follow you around the room. artificial intelligence is already actually embedded in the world around us, and people don't realise that. so, when they think of artificial intelligence, they tend to think
about robots and science fiction and things they have seen on tv, whereas in fact in everything that you do, in google, in facebook, and in all the applications that you use, there is some kind of artificial intelligence there. in a self driving car. if you have an algorithm that shows you how to avoid having an accident, there has to be some decision—making process that says a is more important than b so i'll bump into b rather than a. that have to be put in by humans somewhere. yes, there is a lot of to think about when it comes to artificial intelligence and this iterative machine learning. in other areas of human life. so, when you think about artificial intelligence, what you are saying is that the algorithms running in the background that are learning behaviour and making decisions, but they don't do that in a vacuum, so they are not completely neutral. if you believe that human beings are a finite number of cells or whatever it is, trillions, eventually, we are going to make artificial intelligence that will be smarter than us, and what happens then? we have to grapple with the ethical issues now because artificial intelligence isn't something
of the future, it is something that exists today. mm, where to immortalise myself on the wall? of course, when you talk about anything like ai, there are still so many questions that remain because so much of it is in the few sure. one thing i can say, now, though, is a rebuttal to anyone who ever said i was no oil painting. if you like that, you can catch more of those videos at bbc.com/talkingbusiness. now, back to some serious business with our panel of experts. today we have with us modar alaoui from eyeris. vassilios vonikakis with the advanced digital sciences center here in singapore. and juris puce from coin tech. before the break, we were discussing the possibilities of deep learning. now i want to talk about some of the fears that have cropped up,
as a result of this new technology. modar, let's start with you. there is always this discussion about how machines will one day replace humans. who is to be held responsible with the current technology that we have in place today for example, autonomous cars, if there is an accident when an autonomous car is on the road? maybe, modar, you could pick that up. yeah, well, there has to be a certain, you know, a structure in place. i do potentially think that there will have to be some sort of code of ethics for al. i also leave that companies and governments and authorities will have at some point to decide how the ai should respond in critical situations. for example, the autonomous vehicle accidents, recently an announcement was made by one of the automotive companies, they made the choice
for the ai that is going to be inside the vehicle to protect the driver at any given time, whether that protects the pedestrian or the person that is in front of the vehicle. the driver at any given time, rather that protects the pedestrian or the person that is in front of the vehicle. that is just one of the examples. who to blame and how to think about how ai would respond and who would be responsible. but doesn't this bring up a whole host of questions about how this kind of technology is going to be applied in daily life? given the fact that currently it is on track to replace so many jobs, and we were talking about it before the break, it is going to be something that we are going to see more of in the future, shouldn't we be trying to answer these questions sooner rather than later? from my perspective,
i would like to use a very simple "p" perspective. one of the ps would be for precision. that will differ from case to case, from application to application. another one would be personal data protection. that is another thing. the third thing would be predictability which would be actually to do with the ability to predict something. that will not always be there, 100%. and the fourth one would be the productivity issue. where we say that machines can be more productive than humans and replace them. if you look at this perspective, then the question of ethics and trust is in combination. because it is notjust about autonomous vehicles, it is also about health care, it is also about diagnostics, and so, the way people perceive trust is actually changing all the time. and i think that is also one of the changes which isn't going to be about technology, isn't going to be about science, it is going to be one of the changes which will be about the perception of people, the way they see the world, the way they see the tech ologies they trust, and then of course we can talk
about the basic principles built into the deep learning, principles about decision—making. into the deep learning, but the more complex systems get, the more demands there will be. how do we make the technology fair for everyone? click for example google along with amazon and facebook and microsoft, if i were to mention these as the four or five horsemen companies, they have the ability to do so much with aland at the same time they recognise that using it for the greater good of advancing humanity is of paramount role than to use it otherwise. what happens if or when an entity like this comes an
independent freethinking entity? i know we are delving into the realm of science fiction here but, you know, it was far—fetched about five or ten years ago to talk about deep learning, who is to say it is deep —— far—fetched to talk about this? if you look at where we are now, we are still miles, not even miles, millions of miles away from having an entity like this. unless there is some hidden research which nobody knows about. most of the research which is open is still focusing on specific tasks, it is still about being better in a very specific, narrow task. it is about being good about decision—making, being cognitive, making decisions on a larger scale, we are still very far away from this. do you think we are
getting closer to the stage where machines will be able to make decisions that currently humans do? gradually, yes, but it will take some time. i expect that the things that we already have will become better and better over time. usually, i think that weekend to overestimate what we can do in the short term. we tend to overestimate what we can do in the long run. i think, generally, ai will become better and better but it will always better and better but it will always be applications specific, at least for the near future. thank you gentlemen forjoining us for this discussion on deep learning. that is it for this edition of talking business in singapore. after a great and misty day, a murky
night on the way tonight. be wary and the woa kes, night on the way tonight. be wary and the woakes, fog would be a goblin. generally quite a murky night across the board. some clearer skies in the east and eastern scotla nd skies in the east and eastern scotland with some breaks, there is a typed that charge of frost. for most, mild enough start to sunday but much like this morning, it is going to be a fairly grey one as well. patchy rain and drizzle here and there, that damp feeling. the mist will lift, the grey will turn a slightly lighter shade of grey. northern ireland will see a brighter speu northern ireland will see a brighter spell pushed through. western scotla nd spell pushed through. western scotland will get wetter. more widespread rain compared with the morning. sunshine through the days in eastern parts of scotland. cloud
overin in eastern parts of scotland. cloud over in the afternoon. east of the pennines, few breaks in the cloud but on the pennines itself, fairly misty. the misty day install, one or two drier to breaks possible. eastern parts of wales, you might see some sunshine, in the east of devon as well, but even here there will still be a spot of rain or drizzle possible. let's hope the football is more exciting. if you're journeying home later on, there will be the problem of fog over the higher ground but through the nights, the winds will pick up further north and that will lift the fog but it will introduce the rain. in northern ireland, this weather front rings gale force winds. the heaviest rain not quite reaching the south—east corner, here temperatures in double figures. cold elsewhere. strong to gale forced winds on monday. a brief mild spell mid—week
but later in the week come colder air mattress plunges all the way south across france and into northern spots of spain as well. so, the week ahead, it will be dominated i colder, windy weather. we will see a brief mild spell midweek but colder later on an increasingly windy with a little bit of snow at times, too. but no where as cold as eastern europe. but high in poland was —16 in krakow. widespread snowfall brilliant disruption to airports and to travel. that by phone. —— despite her now. —— goodbye for now. this is bbc world news today, broadcasting in the uk and around the world. the headlines... donald trump has defended his position on russia, saying only fools would think having a good relationship
with moscow was a bad thing. us investigators are trying to determine the motive for friday's gun attack at a florida airport in which five people died. at least 43 people have been killed by a huge truck bomb which went off in a market in the northern syrian town of azaz. also coming up... a deal to end the mutiny in the ivory coast — the president says he'll accept the soldiers' demands. and how a wayne rooney goal seven minutes into an fa cup match against reading marks a moment of history at manchester united.