putting them they used ai technology to create a new. diagnosis. you know you literally saved my life when this get. the diagnosis took all of 10 minutes a human expert would have needed 2 weeks to produce a similar analysis. a i could process massive amounts of scientific data a stack of documents taller than mount fuji. this is the research hospitals supercomputer. we've come here to talk to such tolu miano an expert on bioinformatics we asked me on whether ai could one day replace doctors no i don't think so. simply for. clinicians and our not only nice. clean efficient one full of good and out of the pot. we're told pelley don't let's just go. on points down if you can move they're going to be. good then we need a car. and this is the same. bullet followed through we're not only telling friends something where to but. at the nearby rican institute researchers are developing an ai diagnostic program that could be used to test for stomach cancer. but one expert here disagrees with the toral me on his opinion that ai will never replace doctors. so

by an ai you get alerted to the fact that this is not a human. otherwise it's just going to be a nightmare of phishing scams and so on because suddenly cost nothing you know to waste 10000000 people's time and trick the most gullible thing people into thinking things. we return to san francisco. the city and the region around it are home to countless high tech startup companies many of them use artificial intelligence technology to develop their products and services. eugenio arrived here 4 years ago from moscow. she co-founded her own company called replica and is now the c.e.o. . replica is best known for creating a chat bot an artificial intelligence system that can interact with people. and. the concept began as a tribute to one of her best friends who was killed in a traffic accident. long gone by. my friend from moscow. in the last year or so will live together here and some since. he was working and i was working on mine so it was like. trying to figure out some cisco and you know this kind of chapter of our lives. here's a visionary and i

then they used ai technology to create a new diagnosis. yet for us on a i literally saved my life that we just get all. the diagnosis took all of 10 minutes a human expert would have needed 2 weeks to produce a similar analysis. ai could process massive amounts of scientific data a stack of documents taller than mount fuji. this is the research hospitals supercomputer. we've come here to talk to such tolu miano an expert on bioinformatics we asked me on whether ai could one day replace doctors no i don't think the same place stuff for clinicians and our. lineage. clinicians unfollowed you and i was not. there i told pelley don't let's just go. someplace known for if you can move they're going to be. good then we need to work. and this is the same. bullet followed through we're trying to live and something where is the but. at the nearby rican institute researchers are developing an ai diagnostic program that could be used to test for stomach cancer. but one expert here disagrees with the toral meon those opinion that ai will never replace

music the ai system or eventually stop it going on and endless metal live stream on you tube the ai plays around the clock. humanisation is can't do that but what really is the point. of quick reaction bethlehem dessie to what you just saw there what scares me at least. is the data especially the creativity that comes from the machines i feel like it's especially one sided right from global point of view when you see it specially when you come from from africa there's not a lot of. creative data or painting so our music that's being represented in creating these creative art or music ok that's a very interesting point and we will get to that thanks very much to bethlehem does it microscopically what's your immediate response upon seeing some of that stuff well i think you find that painting on your on your wall well i don't think it's a painting as a matter of fact it doesn't qualify or not at all i mean you might think it's a painting about it's not there something that humans have produced namely a machine that produces something else so it's not really a i doing anything i think so the

in which the ai happens so when i start a new ai project.they want to say we have the data and women clinching the data so we are ready for ai and i say put the data aside and tell me what the decision is that ai will be used for because unless we understand the con plex context of the decision that will lead to outcome i may build a wrong ai system and they said you can do this without data and i said yeah data might help and we will build ai later, but-- this is back to software engineering. if you don't understand the requirements-- there is research lately that says nine out of 10 ai projects fail and this is part of why. >> i like your point about the car. i think that's an interesting point. one topic that came up that i'm aware of is a strong correlation to making women having access to birth control said they can do family planning, a simple link. humans are intelligent all over the world. of this is like a missing link people are not aware of and yet they talk around it. what are we going to do about overpopulation? we need to educa

will maintain its leadership in ai. [laughter] >> the new york times been running a series of interesting reporting on how the chinese have applied their surveillance technologies, which use ai in particular to improve the accuracy of facial recognition and so forth. they are using their involvement totelecommunications systems -- appis malevolent tap that is intended to their involvement in the communication system in a way that is enabling them to propagate their surveillance system to elsewhere, not just in china, but in africa and other parts of the world. i think we have got a compelling warning that the technology is seems it'sdly and it most malevolent applications are the ones that are selected first, rather than later. stage i woulds like to see if there are any questions from the floor that did not make it into electronic form. do we have a mic? >> grace. i go, thanks for your great work. we are great at cutting edge research. what role does it play to lift the standards of basic education in the nation? mr. k

it's an ai system.if they had built a decision model listen here's the variable this thing is using and this is how the prediction will be used in this larger context, i think they would surface that unintended consequence much faster. it's not perfect but at least it makes it visible and subject to critique and continuously improve it. around the context of the ai and also technology solutions that we use. thank you for the question. [inaudible]. i'm wondering if you could talk a little bit about conflict. and how that sort of or the role that might play in outcomes. there is so much on the above thinking. i was making at the table making the decisions to define the outcome. and if in your experience, in developing this methodology if you thought through, what you do when outcomes are in opposition. >> i believe that in many situations, outcomes that appeared to be in conflict, actually part. and that if we do these kind of maps, and resting this is, in fact this is preliminary work. the whole purpose

doesn't really have a reputation for being a leader in ai we've seen the u.s. with its tech giants leading the charge china because the government there is pressing for the technological advancement there is this the e.u. trying to catch up i see that chinese and u.s. firms can account for 86 percent of ai related passes globally so can the e.u. catch up. the way i understand the vision here is not exactly to the. economy catching up with with business the bill is behind a high but rather and regulator a key governing body on the on the on the able to respond to risks and threats generated by ai so you want it simply to speed up growth related to this technology and in that sense compete we've chinese or u.s. based companies i would say that would be a failure what we expected and what is visible in this white paper published yesterday is it different vision in being you have a new leading role in training and development of this technology so that it benefits societies and not only businesses and it also includes problems such as an environmental impact of thi

the machines are getting smarter of course but experts are divided as to when we'll move from basic ai to the scary stuff of science fiction if at all but there are real and relatively immediate threats almost 50 percent of american jobs are under threat over the coming 15 years and health care will probably get more expensive as ai increasingly gets involved in keeping you alive. well banks are using it to detect fraud and predict changes in the stock markets is used by air traffic controllers to help ensure flight safety in the air and on the ground it's helping the police forces around the world identify suspects from c.c.t.v. images ai is also helping self drive cars to navigate road system one example of course tesla with its predictive capabilities machines that can identify images of being used by doctors to spot disease jobs done by humans are under threat to and they are increasingly being used to decide whether you will even get an interview in the 1st place so is the future excel or it's towards us may be worth reminding ourselves the great design and great art well that's a

ai technology is their bread and butter. this engineer has come from jordan to market his invention a pocket spectrometer. we open the application on the fire. proof song to me. he's real. yeah you should be for the journey to feel. just like one get to the device like this and here are your kids nice some information about tomato so you can see how much garbage brought in fiber inside this tomato and the most important what is the freshness for this so that's why it's very useful for people. you know it's true it's like the same growth and used by china to identify faces by getting one face to face is i own 1000000 or so that's how we use the parenting and i just picked on he says. in china it's all about economies of scale. with 1400000000 inhabitants over half of which have smartphones china is the home of big data. the whole country is investing in artificial intelligence china is no longer satisfied with imitating its competitors. it is looking to take on the u.s. for 1st place in the race for global technological supre

we need to lean on ai.ed to be working strategically to make sure it is american, that is leading in this wedding the future and working with our allies around the world to come and follow our leadership rather than china and their aggression right now to be the one that is leading the world. and it's going to take the leadership of donald trump in a very committed republican majority to ensure that its american leadership that is winning this future for us. important to our economy, to individual rights to national security, national defense, it's all related. [applause]. gordon: jim, you and your mom have 58 seconds, to close us out. james: first of stability of tiktok on your phone, delete it right now. one of the hundred 20 million americans who are spying from the chinese. wave to mom. thank you. love, this is the most important thing. were all in this together. we have a great president. america can win this fight. but you are a part of that. you can win this. [applause]. you are a part of this and th

though, she is ai. so ted is trying to understand why there's peace negative unintended consequences. with great power comes great responsibility. you have the ability to take actions as an individual and as an organization that had a giant impact. i am an optimist but believe that they are at the beginning of the solutions renaissance was not just a guy that all of these technologies will come together under a common blueprint. the cultural revolution, all of these fields many of them traditionally viewed a soft field. we are the experts that focus on the outside. what we start to know using some version o d. i is to crystallize solutions to understand the solution to water impacts property. the impacts hit all of us, government, democracy, compani companies. the only way we consult them is to have a new approach to understanding how actions don't surround the whole world to lead the best outcomes. thank you. [applause] do you have a case study or story to tell us i think the example i showed with farm

percent of the terrorist propaganda from you know isis and al qaeda and folks that we take down our ai systems identify and remove before anyone on our network sees it so that's good but that's that's that's a good result and we need to make sure that we get there on all of the different categories of content. some are harder than others so for example hate speech is a particularly challenging one because we have to be able to train ai systems to detect really small nuances right is someone posting a video of a racist attack because they're condemning it which probably means they should be able to say that or are they subtly encouraging other people to copy that attack and that you know multiply that challenge of about subtlety linguistically by you know one $150.00 languages around the world where we operate and the ability to make mistakes we're taking down the wrong kind of thing but we're making progress 24 months ago on hate speech we were at 0 percent were taken down proactively and i think today we're at around 80 percent so it's so it's accelerating it is it's a hard problem i

in looking out where should ai be regulated.nking about again proportionality. let's make sure we capture the upside while addressing what might be unintended negative consequences. that is how we are looking to approach it. we went to work with regulators to make sure we get as a society the benefit of technology while minimizing these unintended consequences. amanda: that was alphabet cfo ruth porat. vonnie: time for the bloombergquint take. -- bloomberg quicktake. already a vaccine or drugs are treated? there has never been a need. the coronavirus that emerged in china had never before been seen in humans until two months ago. why are we hearing about treatments that work? that's because doctors often try unproven drugs they think have a chance of being successful. when they see positive results, they make those public so others can try the same approach. positive results don't always mean a treatment is effective. patients can recover on their own or see results from supportive care like rest and hydration. if health officia

. >> when training the ai system, we fed it tens of thousands of example mammograms just like this one. >> reporter: to cam bat misdiagnosis, they developed an artificial intelligence system to assist doctors and pathologists to actively detect cancer in mammograms. >> this is not something we trained the system on, but when we showed the example, here is a potential suspicious region, here is one. and it turns out in this case it was cancerous. >> reporter: she says the ai model reduced positives by almost 6% and false negatives by about 9%, and caught suspicious tissues on mammograms missed by the human eye. >> mammograms are the best studied. >> reporter: dr. elmore, professor of medicine at ucla is the brain behind an artificial intelligence program that helps correctly diagnosis breast biopsies. would it be a faster diagsis? >> yes. >> reporter: it would be much faster? >> yes. >> reporter: the computer was trained to recognize patterns of 240 breast biopsy images. and while still in the testing phase, the system is yielding promising results. in one test, the ai program outperfor

the 1st prominent victims their faces were super imposed on to pornographic videos with the help of ai the program enabling such fakes was 1st uploaded to the online forum reddit the anonymous user called fakes was quickly banned but other users took over his software code and kept expanding it today similar to software is still freely available online the name is a combination of deep as and deep learning technology and fake but how exactly do defense work how hard is it to make them look real the effect expert how lete explain it to me in los angeles. many deep fakes are based on so-called guns ugh. that are on supervised meaning they learn without human guidance instead they're trained with raw data real photos for example. you're not modeling anything you just collecting a bunch of pieces all you need just make sure there are lines like this and you crop the feast red and you're manually you would try to make sure that there's a balance of different lighting condition different expressions and it just learns that by itself this is what is so powerful about learning based approaches

out who say heat for things aren't necessarily getting smarter and saying hateful things so as the ai systems get better we generally are just catching more and more of the hate speech and they were able to take it down and it's not like like that like speech is getting more sophisticated so they're it's i think that as the systems get better we will get closer and closer to having a lower prevalence of that in the system is worse than something that is adversarial like elections. or election interference we just need to stay on top of it and i think we can take for granted this isn't a problem that you ever fully solve and we will keep on needing to to work on the the defenses but at this point i do feel like we're improving faster than the adversaries and there have been a track record since 2016 of a number of very important high profile major elections which i think there have been relatively clean results in an online discourse and that i think can give us some confidence going forward so would you saying is. if i can rephrase this i mean i read that in the summer of 2018 you wer

but i think ai is is coming i think it is part of the solution in terms of manifest in the galaxy in the government's right to focus on what is a legal route and getting into a complex debate about harmful culture at least not to go to children because that really would be a dangerous censorship but i think a i needs to be accountable and transparent and it's the role of inside information commissioner and data protection workers to ensure that a operates within a a framework which is in centric but because of the scale of the homes online and you know the need for rapid action i think the ai has to play a role where we're talking here about building a broader framework right so mira i want to ask you about the different approaches we've talked about different approaches different laws from different countries and it does seem rather ad hoc so is there the space then or should there be the space for a universal declaration of sorts given that the internet is a shared global resource should there be a consolidated global approach where all the countries come together and make a plan.

says no you don't get to write the rules for ai when you're the one who's deciding what ai is and therefore how to use it and quickly how much of this is just about are these big tech companies actually more likely to to adhere to these these demands when it comes to ai or or hate speech in just an effort to you know avoid digital taxes that are going on in the e.u. as well. yeah well i mean that's part of it as well and that's part of this whole discussion if you don't play it nice with us one thing that the eve can do is try to impose these digital taxes we know a lot of countries want to do with either way and so ultimately what the u.s. to figure out is you know what is going to be the costumer i mentioned data collection a minute ago what is going to be the cost to you as a tech company for having did it. action on our citizens and then those companies say hey if we do what we're supposed to do we play nice with politicians will you lean off the taxes so that's probably going to be somewhere in the middle that they find a kind of a balance between those 2 things what it means to play n

it wants to attract 20 billion euros of investment every year over the next decade in just ai.he reason these are being so closely watched is once adopted, they will apply to companies out of the eu but else where like facebook, amazon and apple so facebook ceo mark zuckerberg has been trying to get out in front of these issues, trying to focus the content moderation but there's a lot of scepticism the eu platforms will not step up on that issue or any other, as the commissioner told our colleagues at cnbc europe, the eu believes big tech has a special responsibility guys ylan, thank you for bringing us the latest out of europe which has been aggressive towards the u.s. moving forwards with these legislations. that being said, the workshop looking at the communications 230 decency act that protects online platforms for liability of users and the like. what is going on with that >> this is known as big techs liability shield the attorney general bill barr actually had very strong criticism of this law. he said basically the law is not only shielding big tech, it's also shielding

in fact, the ai may end up being more creative than humans because it can process so many more imagesseeing. the course is also preparing students for job losses in their industry to artificial intelligence, especially in areas like pattern cutting and other repetitive work. this really is something we can't deny. it is coming and we need to be prepared for it. maybe these people can be educated in some sort of position where they will be using this al to do this job, but if it is something repetitive, if it can be automatised, it will be automised. the next experiment will be feeding this algorithm photos from architecture as well as the clothing, to see what fashion designs it comes up with. an exciting time for those able to adapt, but with implications for workers around the world, perhaps a need to think quickly about the direction of this important industry. people love playing video games, but making them is a different matter. the art of creation used to require quite a few skills, not least the ability to code. but over the past few years, content creation tools have become a

and having a background in ai will be essential to their success.is important to have a knowledge of everything else coming up and how we can use that in fashion. with ai we will have much more ways and options, instead of sketching down ideas we can have thousands of them and can select quickly what we find beneficial for our project. but others worry about losing originality in this process.|j others worry about losing originality in this process. i think there is a danger of it stripping out creativity because there is a lwa ys out creativity because there is always that argument that designers create something from their brain, something that hasn't been done before whereas ai looks at something done before and then create something new. there is this paradox, it doesn't mean we will end up paradox, it doesn't mean we will end up repeating the same things over and over or will it mean we will be given new ways to create. in fact the ai may end up being more creative than humorists because it processes so many more images than we are capable of s

the ai that is fair transparent and democratic it also calls for more debate about whether to use the technology for facial recognition of proponents say it would increase security while critics argue it is too intrusive. ok let's get you more on this i'm now joined idea because one again is in brussels for you gail good to see you what exactly is the e.u. calling for our. well 1st of all of the you hopes to start a discourse among either member states or that the 27 member states with all the different regulations they currently have or sometimes have not on artificial intelligence and on the digital markets bring them together to find common guidelines and the hope is that those common guidelines will then unlock the huge potential that there is economically potential that there is also for the environment because the argument of the commission is that a greener europe can only be achieved with a more digital europe seeing that a i driven. culture for instance could help a lot by making europe more green so all of these things together they hope to put on the table find common guide

ai cannot understand whether a script is good.th warner brothers but the studio insists the machine learning software will be used in the marketing and distribution side of things. as far as hollywood studios are concerned, humans remain in the picture... ..for now. we're going to do this, you and me. really interesting stuff isn't it. that was mark and of course, this weekend is oscar weekend so we will know very soon which films did well this year. we've been looking at some of the best in the vfx category and here is ourfinal nominee. long have i waited. confront your fear, it is the destiny of a jedi. bringing carrie fisher back was an incredible tiered problem and i thought it was going to be difficult but i actually didn't appreciate how complicated it was going to be. you can't do it for me. what we did was build this enormous matrix of all of the performances of carrie fisher from seven and eight and at that, essentially figured out all the lines she said and thenjj and chris terrio wrote the script around those lines. and