rock and religion. brings funny parallels to. some of irreconcilable. god the devil and rock n roll. 17 on t. w. . welcome to tomorrow today the science show on d w coming up. current levels of artificial intelligence is it really intelligent or just cleverly program to. ai is starting to have an impact on health care with apps that can deliver diagnoses could they replace doctors one day. and the robots are coming artificial intelligence our topic on today's program thanks for joining us.

the 950 computer scientist alan turing developed a test later name for him it was meant to help judge whether behavior in machines could be considered intelligent in 1997 grandmaster garry kasparov faced off against i.b.m.'s deep blue and lost 2011 saw the launch of apple's siri the 1st virtual voice activated assistant. and in 2018 a painting produced by a machine was auctioned off for the 1st time milestones in artificial intelligence but what is it exactly. artificial intelligence the attempt to simulate human intelligence in machines. what's called fall or strong ai is an artificial intelligence construct able to interact autonomous lee and logically one having what we would call consciousness for now it's still science fiction weak ai also

known as narrow ai is based on algorithms it doesn't require consciousness but operates within a limited predefined range of functions it's already a part of daily life and industry many tasks that are physically demanding or monotonous are now performed by robots also in the field of health care where computers now process and analyze vast amounts of data speeding up diagnosis and treatment. and smart homes are set to improve our quality of life thanks to ai based solutions that increase comfort and energy efficiency. but that begs the question what is intelligence exactly in humans intelligence is associated with cognitive ability we are created able to make plans and carry them out we have the ability to imagine and awareness we can make decisions and some current ai systems appear to be able to do that now as well but some researchers say it's important to

know how those systems are coming to their conclusions are they really making decisions based on human logic here's a story that illustrates how hard that can be to determine. in around the year 900 there was a horse named time he lived in berlin and was very clever. his owner bill him from the austin said hans could do arithmetic. once how much is one plus 3. for clever hans that was easy. everyone asked how could it down morse be so smart. nowadays we ask the. same thing about machines that learn artificial intelligence is something of a black box is it truly intelligent a team of computer scientists set out to find out going to and we did some you

could say archaeological research into how the various methods in artificial intelligence solve problems are there systematic differences the answer it turns out is yes but those differences were rather surprising. we peered for the 1st time into the black box and what we found was like whoa really. things that no one had seen before in human. learned wojciech samik found a way to analyze the decisions made by ai systems in one test they showed an award winning system thousands of pictures. get recognized the objects depicted but in the case of the horse images it wasn't the head or the tail but something not coursey at all that was the decisive factor. the same house system did identify this as a horse about it focused its attention on the copyright the source of

a picture or photo archive at the bottom left the links a bit. using a heat map the researchers showed which part of the picture the ai system paid most attention to. that really surprised us so we took another look at the horse photos in this big dataset and found that many of them included just such a copyright notice that's what the system learn to associate with a class horse. so. clever hans it turns out may have been quite clever but he couldn't actually do math. 9 times out of 10 he stamped his foot the right number of times. but the psychologist worked out that hans had learned to interpret human facial expressions and body language they indicated tension and then suddenly relief once the right number of foot stops was reached.

here's an ai system that aims to determine age and gender by looking at faces we tried it out. if you smile at it thinks you're young and. grimace and the result is you made 40 days and male. in general the system identifies elderly people as younger than they really are. younger intern i think. we wanted to find out why this was mentioned most of the older people here were smiling. but if you look at pictures on the internet you know you find that for the most part they look very serious. on the thing so

there's a bias in the data which influences the model leaving it to infer that all people always look serious with a mention. in the lab it's fun to play with the system in the real world bias and errors in ai can prove fatal. so. nobody wants to have a disease diagnosed by a clever hans' if the question is do i have cancer or not it's a matter of life and death you need certainty the ai strategy has to be right. artificial intelligence is already used in many applications are the strategies deployed indeed the right ones. i think. of the many systems we encounter in our daily lives are such that we don't 100 percent understand what they do. until now there was no way to tell if they were cheating or not.

now we've developed a tool which lets people check. they've already tested hundreds of systems and established that hardly any are truly reliable in their decision making this was among the smarter ones it's playing an old computer game the system for the system is trying to get the ball to go high to win as many points as possible while avoiding the risk of losing the bullet it's a pretty smart strategy which the system learns all by itself for the. time this clown top. never hunts also developed a clever strategy that he learned all by himself. it just didn't involve doing new songs. it's still far from perfect but ai is creeping into a huge range of areas in daily life whether in transport logistics or nursing care . on facebook. we asked you what tasks would you like to hand over to an ai.

lewis would like a robot to help take care of the housework cleaning or doing the washing filling or emptying the dishwasher that would be grand she says. she said i wouldn't mind a chef robots that could help choose ingredients for meals. ben thinks we need robots in the justice system that are programmed to enforce the law that would help put an end to corruption among the powerful he says and maybe end corruption entirely. and diego writes that ai could help improve management of patient appointments ensuring medicines are taken correctly and improving medical planning they could also help shorten waiting times in doctors' offices. is already playing a major role in medicine for instance when they feel sick some people now don't head straight for the doctor's office they just reach for their smartphone. or

feeling unwell and have a headache so you type your symptoms into an app. a machine then talks to you rather like a real doctor after you've answered around 20 questions the machine will come up with a probable diagnosis. called ada have what it takes to revolutionize medicine . that's what it's inventors hope. that will shorten the safety of all of us i believe that this will alter the doctor's role in certain ways for i've spoken to him and worked with many doctors in the past 8 years and my experience is that the best doctors are the ones who realize that technology can help them to be even better. for 80 years doctors and developers created the app based on data from thousands of studies. this is 8 as

a result after analyzing the symptoms. on the left are the symptoms which were clicked either yes or no i did on the right and possible explanations a to comes up with the green line show the affirmative answers and how strong the argument is for a certain diagnosis. the red line show the symptoms with negative answers and the illnesses this eliminates. the app is being used in countries around the world that diagnoses it provides are mostly correct the accuracy is roughly as high as from real doctors the app developers are using artificial intelligence to improve the success rate. their plan of learning factor means that theoretically the artificial intelligence can learn from the results that 8 provides and the feedback that you just gave on whether a diagnosis or an explanation provided by ada was correct but have for us it's important that this doesn't happen automatically but that there are doctors in

between who decide whether these changes make sense or not doesn't shine up to the noble therefore that's what i mean this way users can't just say stuff that's not correct. so how does it work in day to day life. india is one of the 7000000 people who use the app. simple language should use simple language so that more and more people can understand it . metabolic diseases like obesity diabetes etc should also be included in it along with their dietary suggestions. in its analysis the app warns the patient that in case of a potentially serious disease they should see a doctor a recess. could be suffering from a nasal cavity inflammation his doctor confirms that diagnosis most of the symptoms regarding this disease vice being covered by the software there is one very good

thing i feel that it is not the placement of the doctor because the clinical examination for the patient is very important but it definitely is a useful tool. the developers hope that eventually the smart system will be able to suggest diagnoses for elements still in that very early stages. i think ai will definitely contribute to longer and healthier lives and in some cases maybe stop diseases from happening in the 1st place that's also it has. to do that the system has to accompany the use of for many years collecting data the app developers emphasize that the data belongs to the user and is only evaluated anonymously but there are concerns the world of ai medicine means asking fundamental questions about exchanging personal privacy for higher life expectancy .

i'm sorry. i'm afraid i can't do that. in the science fiction classic 2001 a space odyssey the computer how begins to kill the crew of his spacecraft one by one predictions about where ai is going off and try to fin line between hype and hysteria physicist stephen hawking once famously warns that artificial intelligence could spell the end of the human race we asked 2 experts what they think about some of the most popular half truths spread about the field. excuse me he said but that talk to you didn't prime condition it. does not appear in science fiction from what we see in science fiction films that do things on their writing and chat with you is really a very very long way off. yet the feds are constantly.

to wish us but it's an interesting bit of history but in the 1950 s. the term artificial intelligence was adopted for marketing reasons basically to get financing for research into computer simulation of human behavior and human learning and it was a smart choice of words because it's so evocative 5 of these in the book to view. if you're competent hello everyone in an english artificial intelligence. i think it's always been the case that technological advances change the nature of human work which means jobs we have nowadays might no longer exist in the future. my students on the whole at least in the past more jobs have been generated by new technologies than have been lost. in tawi good jobs and staying on far more often half the masses of miserable jobs are being created like paid to click and training

system. believe. there are loads of low paid jobs where people have to watch videos and classify images to teach a system what a cat it's or something like that bites of being the cuts a bit of what i mean it's real people doing all these jobs and this is known as artificial artificial intelligence agents but. by figured out there are stories of companies saying when it comes to choosing star for promoting members of staff they want to see what kinds of people have already have successful careers in those companies. i think it. might sound neutral and objective but if it turns out all those people are quite by chance pointed mail by you and have certain qualities such as pushy and ambitious and it was on the end then it's obvious that anybody who's a bit different doesn't stand a chance i'm less inclined to. pick on you that we can program any kind of

prejudice into a. ai system. the machine the kind of machine can generate works of art that are considered creative conceivable and by sheer very famous examples like when an algorithm is fed with the works of mozart and bach order and then produces a new piece of music one that had not existed until time the condiment as you might call that creativity but the project will always be similar to the ones i've been fat to order it would have to go into the realm of philosophy here what is creativity i would say i is not creative. world hunger climate change and other environmental problems aren't a result of our not having had a i till now but they're kind of contingent in again but which isn't to say that ai is all rubbish there are some extremely useful applications. a minimum number of

all this to be a star it's really important for us to understand each individual step with the progress made. society needs to address these technologies of the assumption that you learn a trade or study something and that's not no longer holds in a time of rapid technological progress. and we all have to keep on learning so that these systems remain comprehensible and controlled. by. the sort of end i will mention. there are countries where people are under much more intensive surveillance than they are here. as in part to do with the technology that enables it to be scaled up so massively democratic societies need to think hard about what kind of surveillance they want in the end it's the people who make the rules and it has to stay that way. algorithms now make decisions about hiring about prices on websites and about what kind of advertising you'll see when you go

online in an effort to regulate the field the e.u. has now released ethical guidelines that take human rights and needs into account when developing the ai of the future. development is read right by you. do you have a science question that you've always wanted answered it we're happy to help out you send it to us as a video text oh boy smile if we answer it on the show we'll send you a little surprise as a thank you cannot just ask. somebody to shoot from india wants to know what is data and doesn't have. a material existence. when we talk about data we usually mean information but also what carries at even ancient cultures produced data egyptian temples are covered and hieroglyphics. the complex symbols describe important events and mythological tales carved into stone they

remain to this day. in the analog world we use letters and numbers to preserve knowledge and pass it on that is the purpose of data. but it only works when others know how to read it and then they work like building blocks letters can be combined to make words which can be used to describe almost anything. with just a few letters numbers and symbols science can describe complex processes in nature . computers reduce data to just 2 elements every letter number or symbol from the analog world is stored in a computer as a sequence of ones and zeroes. in the analog world data is recorded in or on matter what we write with chalk or

ink a part of that substance remains behind on the data carrier sometimes data is recorded when material is removed like with these release however without the stone columns it wouldn't exist. so how does it work with computers special components transform the input data into 2 different states on and off more than 0. but even this is dependent on a physical data carrier but unlike data on stone or paper digital data no longer have mass and can be transmitted through wires and cables at the speedo. light. satellites are able to transfer data without a cable use another carrier to transport the information magnetic waves. but to retain data it has to be saved and this is based on physical data carriers servers that huge data storage facility.

got a question or want to let us know what you think here's how to get in touch. most people associate ai with robots the idea of a future humanoid machines causes a wide range of reactions and there are surprisingly dependent on the culture you live in. so would you like to work together with robots or does the idea make you shutter. at being thinks we'll have to collaborate with robots in the future whether we want to or not he says it's inevitable. ai would remind us to rest so we wouldn't be continuously working and it would be nice if it could entertain us also he'd like one that could give recommendations for newly released books. from the hindu says of course i'll work with robots because they don't gossip. hi everyone i'm sophia from handsome robot how do you feel about

her channel i'm tim. about this one. the demand. for the next 10 years. if that really happens many are very likely to soon encounter one many. so should we be worried about that prospect. after cheats very widely around the world. so you know more studies human robot interaction and cultural differences involved in the social acceptance of robots. he and his colleagues surveyed attitude in britain germany the netherlands mexico china japan and the united states.

it turns out that americans feel most positive about robots followed by the japanese dutch germans chinese and british. mexicans are least well disposed towards them. who will look when you look when you are. thinking robots are good or bad not only depends on your country but also on your age. you know in japan people of all ages like robots. britain young people like them but old people don't. you know so. i suspect gender would also plays a role. so it was no. common. among those surveyed more women than men feel positive about robots.

you the japanese are the most willing to accept robots that look like people. and are even willing to believe they have human attributes. union local. in japan we have something called animism. and if something looks like a human we assume it has a soul. and you don't find animism only in japan but every culture has it in some form you know. the media and pop culture influence the way people think about too much and robots and japanese animated films and comics robots like human like in the classic straight boy for example. all the popular among the animated film ghost in the shell. landing suggested robots might in hollywood movies like i robot on the other hand the robots look somewhat less like people and tend to turn on less human masters. and that wraps up our show on

artificial intelligence but join us again next week when we'll be looking at colors and how they can influence consumption we'll talk with psychologists and product designers about unconscious mechanisms that help dictate what we buy that and much more next time tomorrow today see that.

to. me. sometime in the 26 to you my great granddaughter. what would the world be like in your life turning around half a century. it's really frightening. and i have to. look why aren't people more concerned. little yellow boots in 75 minutes on the double. for the father it was the beginning of

a long ordeal. for the son of a traumatic memory. provoking means unfolding. talents where. 30 years after the one chinese capital a father and son tell their story. today in the day on. the news man it's nice. to. meet you. to mull. you in.

small as. it could be green. very green. old as blue. yes. quite. why i just love the. way definitely. paul just you know if that's what you personally believe couples are very special to georgia choose your favorite come out of. the.

plane. load. the boat. this is the news live from berlin sudan's military moves against protesters several people are feared that demonstrators say that they are facing a massacre as security forces stormed the protesters town. also coming up u.s. president donald trump and his wife melania arriving in the u.k. for a controversial state visit thousands of demonstrators are expected to turn out across the capital and he will ask why his trip is divided list.