gary's bet times ten. we will be speaking to gary injust a minute. he hasjust he has just offered he hasjust offered elon musk he has just offered elon musk $1 million bet that al may not be superhuman he has in mind. i will introduce you to that technology tonight. the ai avatar who will be joining us as a panellist on this programme. we will also focus on openai. the new york times reports on open al's latest new tool voice engine, which can generate a convincing clone of anyone's voice using just 15 seconds of recorded audio. that might scare you a little bit. it has security implications. we will introduce you to someone who has resurrected his career using that kind of technology. and is al technology actually good enough to threaten the livelihoods of musicians? the guardian looks at suno, an artificial intelligence music generator that can create original songs using a simple text prompt. as usual, stephanie hare will be here to guide us through it all. agi

here to guide us through it all. a61 is our top story tonight. there is always a debate over whether or not it can exceed human intelligence, do you think it will get there? what kind of timeframe you have in mind? it stands for artificial general intelligence, you may have heard it called the singularity. it is the hypothetical moment, it is very speculated, even a bit science—fiction, perhaps, when machines become smarter than humans. now, they are already smarter than humans and some tasks like beating us in chess or at go, but can they do all of the things all humans can do? no. so, they might be domain specific better than us, but are they better than us and everything? are they better than you or me, all 8 million others? that is the

question. 8 million others? that is the question-— 8 million others? that is the cuestion. , ., ~ , 8 million others? that is the cuestion. , ., ~ . question. here is elon musk's recent tweets marking _ question. here is elon musk's recent tweets marking a — question. here is elon musk's recent tweets marking a flood _ question. here is elon musk's recent tweets marking a flood of— question. here is elon musk's recent tweets marking a flood of online - tweets marking a flood of online bets. i think $10 million has been waged so far. he said ai will be smarter than any single human next year, and by 2029, probably smarter than all humans combined. what does the public think of that? a listen. do you think ai will surpass human intelligence by the end of next year? no. why do you think that? artificial intelligence is obviously coming on leaps and bounds, but i don't think it will be in a year. i think by the end of this decade it probably will, but not next year, no. i guess it depends. i mean, in a holistic sense, i think it would have to be very similar to a human's emotive abilities, in addition to its... because it's already very computationally advanced. i think that's probably a pretty lofty goal. i i know elon musk is pretty prone to making some ambitious calls.

i wouldn't bet against him, - but i think by the end of next year is probably a little bit optimistic. very ambitious prediction. i'm and personally not very informed enough to say something about the timeline, but i think in terms of the trajectory, the direction, that sounds like a sensible thing to say. 0ne one of those who has waged $1 million on the spat with elon musk is a scientist professor gary marcus, a leading voice on artificial intelligence. there are others who have joined artificial intelligence. there are others who havejoined him in that bet. it is now up to $10 million. professor gary marcus joins us from vancouver. lovely to have your on the programme. when you talk to the public about agi the programme. when you talk to the public about a61 and what it will be capable of, there is a mixture of excitement, curiosity, ignorance, fear. you are prepared to bet $1 million that it will not surpass

human intelligence anytime soon. why are you so sure? it is human intelligence anytime soon. why are you so sure?— are you so sure? it is a long way to co. are you so sure? it is a long way to to. you are you so sure? it is a long way to go- you can _ are you so sure? it is a long way to go. you can think— are you so sure? it is a long way to go. you can think about _ are you so sure? it is a long way to go. you can think about the - are you so sure? it is a long way to go. you can think about the old - go. you can think about the old 80-20 — go. you can think about the old 80—20 rule. are 80 of the way there, you have _ 80—20 rule. are 80 of the way there, you have language models, chat bots which _ you have language models, chat bots which sound like people, but they make _ which sound like people, but they make a _ which sound like people, but they make a lot — which sound like people, but they make a lot of dumb mistakes, they hallucinate, they make things up, they are _ hallucinate, they make things up, they are bad at logic, reasoning in general, _ they are bad at logic, reasoning in general, they are not very good at planning _ general, they are not very good at planning things. it is like we saw with driverless cars, it was easy to -et with driverless cars, it was easy to get a _ with driverless cars, it was easy to get a demo— with driverless cars, it was easy to get a demo that was 80% of the way there. _ get a demo that was 80% of the way there. ilut— get a demo that was 80% of the way there, but getting the last 20% are proven— there, but getting the last 20% are proven really hard. elon musk once driveriess— proven really hard. elon musk once driverless cars that go across the united _ driverless cars that go across the united states in robo taxis. he promised — united states in robo taxis. he promised that every year from 2015 to the _ promised that every year from 2015 to the present. he is often making those _ to the present. he is often making those pronouncements. the reason now for the _ those pronouncements. the reason now for the better, which he has not yet taken, _ for the better, which he has not yet taken, is _ for the better, which he has not yet taken, is i — for the better, which he has not yet taken, is i think we need accountability. a lot of hype is driven — accountability. a lot of hype is driven around people making empty promises. _ driven around people making empty promises, and i wanted to put some on eton _ promises, and i wanted to put some on elon. ., ,,

promises, and i wanted to put some on elon. ., i. ., ,, promises, and i wanted to put some on elon. ., ., ,, ., , ., on elon. can you talk to us about how ou on elon. can you talk to us about how you would — on elon. can you talk to us about how you would use _ on elon. can you talk to us about how you would use a _ on elon. can you talk to us about how you would use a scientific i how you would use a scientific method to test this bet? how will we know if artificial intelligence has surpassed humans? he know if artificial intelligence has surpassed humans?— know if artificial intelligence has surpassed humans? know if artificial intelligence has surassed humans? ~ ., ., ., surpassed humans? he kind of made a crazy claim. — surpassed humans? he kind of made a crazy claim. which _ surpassed humans? he kind of made a crazy claim, which is _ surpassed humans? he kind of made a crazy claim, which is that _ surpassed humans? he kind of made a crazy claim, which is that machines - crazy claim, which is that machines would _ crazy claim, which is that machines would be _ crazy claim, which is that machines would be smarter than humans in every— would be smarter than humans in every way, — would be smarter than humans in every way, and so, ijust have to find _ every way, and so, ijust have to find one — every way, and so, ijust have to find one or— every way, and so, ijust have to find one or two ways in which machines— find one or two ways in which machines are not as smart as people, and i_ machines are not as smart as people, and i think— machines are not as smart as people, and i think i_ machines are not as smart as people, and i think i win the bet. i laid out how— and i think i win the bet. i laid out how i_ and i think i win the bet. i laid out how i think about that in a sub-stack— out how i think about that in a sub—stack essay. i gave 12 different criteria _ sub—stack essay. i gave 12 different criteria for— sub—stack essay. i gave 12 different criteria. for example, a human can -et criteria. for example, a human can get into— criteria. for example, a human can get into a _ criteria. for example, a human can get into a vehicle that they have never _ get into a vehicle that they have never been in before where there is no man— never been in before where there is no man and — never been in before where there is no map and figure out how to get around _ no map and figure out how to get around and — no map and figure out how to get around and avoid obstacles, and regent— around and avoid obstacles, and regent street signs, and even hand lettered _ regent street signs, and even hand lettered three times. people can go into a _ lettered three times. people can go into a home they have never been in and clean _ into a home they have never been in and clean it — into a home they have never been in and clean it. there are all kinds of things— and clean it. there are all kinds of things people can do. a leading scientist — things people can do. a leading scientist who comes up with nobel prize discoveries or people who write _ prize discoveries or people who write oscar—winning screenplay is, and so _ write oscar—winning screenplay is, and so forth. there are many things ordinary— and so forth. there are many things ordinary humans can do and many

things— ordinary humans can do and many things expert humans can do, and, if he is _ things expert humans can do, and, if he is really— things expert humans can do, and, if he is really trying to claim that al is going _ he is really trying to claim that al is going to — he is really trying to claim that al is going to be ahead of all humans, then i_ is going to be ahead of all humans, then i can _ is going to be ahead of all humans, then i canjust go out and find some humans _ then i canjust go out and find some humans do — then i canjust go out and find some humans do things machines still cannot — humans do things machines still cannot. there are tonnes and tonnes of things— cannot. there are tonnes and tonnes of things now in 2024 that machines cannot— of things now in 2024 that machines cannot really do. i think it is a wildly— cannot really do. i think it is a wildly optimistic claim. i cannot really do. i think it is a wildly optimistic claim.- wildly optimistic claim. i think what we are — wildly optimistic claim. i think what we are driving _ wildly optimistic claim. i think what we are driving at - wildly optimistic claim. i think what we are driving at is - wildly optimistic claim. i think. what we are driving at is whether the machines can become emotionally intelligent, whether they can empathise. when you go the extra yard with that, i suppose the ultimate question is whether the benefits of it all outweigh the risk? at benefits of it all outweigh the risk? �* , , ., ., risk? a very good questionnaire about if the _ risk? a very good questionnaire about if the benefits _ risk? a very good questionnaire about if the benefits outweigh l risk? a very good questionnaire l about if the benefits outweigh the risk. right now, with what we call large _ risk. right now, with what we call large language models, the dominant technology firm generative ai. i am not sure _ technology firm generative ai. i am not sure that they do outweigh the risks _ not sure that they do outweigh the risks. people using them in their businesses, they are good for brainstorming, but people are creating — brainstorming, but people are creating misinformation and disinformation with them, the uk

series— disinformation with them, the uk series getting polluted, people are using _ series getting polluted, people are using them to make nonconsensual deeufake _ using them to make nonconsensual deepfake pawn, people are using them to rip off— deepfake pawn, people are using them to rip off banks, many things are going _ to rip off banks, many things are going wrong. nobody really knows the outer limits of generative ai, the uses _ outer limits of generative ai, the uses people will find. we know they are not— uses people will find. we know they are not reliable. that limits the upside — are not reliable. that limits the upside. criminals don't care. when they send — upside. criminals don't care. when they send spam, if they have one victim _ they send spam, if they have one victim who — they send spam, if they have one victim who falls for it out of 10,000, _ victim who falls for it out of 10,000, that is fine. they do not —— need _ 10,000, that is fine. they do not —— need a _ 10,000, that is fine. they do not —— need a reliable performance. ithink what we _ need a reliable performance. ithink what we have at the moment is more useful— what we have at the moment is more useful for— what we have at the moment is more useful for criminals and ordinary people _ useful for criminals and ordinary people who need a reliable system. so, absolutely, we need to be asking about— so, absolutely, we need to be asking about gusts— so, absolutely, we need to be asking about gusts and benefits. in a long time, _ about gusts and benefits. in a long time, al, _ about gusts and benefits. in a long time, ai, which needs a lot more invention — time, ai, which needs a lot more invention than we have done yet, can really— invention than we have done yet, can really revolutionise science, medicine, and i think we should be working _ medicine, and i think we should be working on— medicine, and i think we should be working on it, but we are also evaporating the stuff we have right now _ evaporating the stuff we have right now. we _ evaporating the stuff we have right now. ~ ., evaporating the stuff we have right now. . . , , ., now. we have barely even begun to reuulate now. we have barely even begun to regulate al — now. we have barely even begun to regulate al now _ now. we have barely even begun to regulate ai now when _ now. we have barely even begun to regulate ai now when it _ now. we have barely even begun to regulate ai now when it has - now. we have barely even begun to regulate ai now when it has not - regulate ai now when it has not surpassed humans. what would

regulation look like in a world where artificial intelligence is smarter than us?— where artificial intelligence is smarter than us? that is a really aood smarter than us? that is a really good question- _ smarter than us? that is a really good question. i _ smarter than us? that is a really good question. i think— smarter than us? that is a really good question. i think we - smarter than us? that is a really good question. i think we have l smarter than us? that is a really - good question. i think we have some time at _ good question. i think we have some time at a _ good question. i think we have some time at a figure that out. i think we are — time at a figure that out. i think we are not— time at a figure that out. i think we are not doing that great a job, especially— we are not doing that great a job, especially in the usa, with actually having _ especially in the usa, with actually having regulation with teeth. i think— having regulation with teeth. i think the — having regulation with teeth. i think the eu is ahead, the uk has worked _ think the eu is ahead, the uk has worked a — think the eu is ahead, the uk has worked a day. the first thing you need _ worked a day. the first thing you need is _ worked a day. the first thing you need is an— worked a day. the first thing you need is an international ai agency so that— need is an international ai agency so that we — need is an international ai agency so that we can work together and, when _ so that we can work together and, when things get more complicated than they— when things get more complicated than they are now, we are prepared. ithink— than they are now, we are prepared. i think that _ than they are now, we are prepared. i think that what we have now is a rough _ i think that what we have now is a rough draft. — i think that what we have now is a rough draft, we are working things out technically and legally, the eu ai out technically and legally, the eu al act— out technically and legally, the eu al act will— out technically and legally, the eu ai act will not be perfect, we will learn _ ai act will not be perfect, we will learn from — ai act will not be perfect, we will learn from that. the biggest problem is that— learn from that. the biggest problem is that we _ learn from that. the biggest problem is that we do not have any reliability guarantees around current — reliability guarantees around current ai, and we do not know how to get— current ai, and we do not know how to get there — current ai, and we do not know how to get there. when somebody builds a bridge. _ to get there. when somebody builds a bridge, they can say: under the circumstances i know it can bear this load — circumstances i know it can bear this load. we need formal, mathematical guarantees based around mathematics like when we build an aeroplane — mathematics like when we build an aeroplane. we have a knowledge about

how to _ aeroplane. we have a knowledge about how to do _ aeroplane. we have a knowledge about how to do that but we don't have that for— how to do that but we don't have that for the current ai, which is trasically— that for the current ai, which is basically unpredictable. it depends on exactly what it is asked, and that is— on exactly what it is asked, and that is in— on exactly what it is asked, and that is in a _ on exactly what it is asked, and that is in a very idiosyncratic way, and it— that is in a very idiosyncratic way, and it makes— that is in a very idiosyncratic way, and it makes it hard to make the systems— and it makes it hard to make the systems work systematically. the most _ systems work systematically. the most dramatic regulation, i think, would _ most dramatic regulation, i think, would be — most dramatic regulation, i think, would be to say that if you cannot interpret — would be to say that if you cannot interpret what your system is doing, we do _ interpret what your system is doing, we do not _ interpret what your system is doing, we do not want it here, unless you can say— we do not want it here, unless you can say it — we do not want it here, unless you can say it is — we do not want it here, unless you can say it is safe and some other way, _ can say it is safe and some other way, we — can say it is safe and some other way, we are _ can say it is safe and some other way, we are only going to accept systems— way, we are only going to accept systems which are interpretable. nobody — systems which are interpretable. nobody can do that right now. my prediction— nobody can do that right now. my prediction is generative ai will flame — prediction is generative ai will flame out. it will do some things, but it— flame out. it will do some things, but it is— flame out. it will do some things, but it is not— flame out. it will do some things, but it is not going to be valued with— but it is not going to be valued with companies being $86 million. there _ with companies being $86 million. there will— with companies being $86 million. there will be a cooling off period, may be, _ there will be a cooling off period, may be, and we will think more about how important interpret ability is and how— how important interpret ability is and how important safety guarantees are. and how important safety guarantees are now. _ and how important safety guarantees are. now, we are not in that position— are. now, we are not in that position to _ are. now, we are not in that position to do it. we are ahead of our skis— position to do it. we are ahead of our skis now. we position to do it. we are ahead of our skis now-— position to do it. we are ahead of our skis now. we are a bit squeezed for time, our skis now. we are a bit squeezed for time. i — our skis now. we are a bit squeezed for time. i hope _ our skis now. we are a bit squeezed for time, i hope you _ our skis now. we are a bit squeezed for time, i hope you will— our skis now. we are a bit squeezed for time, i hope you will come - our skis now. we are a bit squeezed for time, i hope you will come back| for time, i hope you will come back in the programme and talked to a simmer? it in the programme and talked to a simmer? ., , , , ., , simmer? it would be my pleasure. thank ou simmer? it would be my pleasure. thank you for— simmer? it would be my pleasure. thank you for having _ simmer? it would be my pleasure.

thank you for having me. - simmer? it would be my pleasure. thank you for having me. we - simmer? it would be my pleasure. thank you for having me. we are l thank you for having me. we are auoin to thank you for having me. we are going to have — thank you for having me. we are going to have a _ thank you for having me. we are going to have a break, _ thank you for having me. we are going to have a break, the - thank you for having me. we are going to have a break, the other| going to have a break, the other side, we will look at how it is being used to help somebody who has lost their voice.

this welcome back. through the course of this programme, i want us to see ai in action and to see it grow. i have invited somebody onto the programme who some of you may have seen it before and who is developing ai and is tonight going to introduce some of his works. his name is samir mallal. we're nowjoined by our ai handy man samir mallal, award—winning filmmaker and ai developer and ceo of one day. he has somebody with him. and over the weeks we do this, i want her to be a part of the show she is not a finished article, that is the point. i want to see what she learns. before we speak to her, i want you to tell us how you would use and how you think she is going to evolve. i

wanted to create an ai because i recognised the potential, and i found as a creative, i wasn't involved in the process. i built her to create and expand what i already do it. i wanted to see my creativity from different angles. she do it. i wanted to see my creativity from different angles.— do it. i wanted to see my creativity from different angles. she is on the left. you from different angles. she is on the left- you are _ from different angles. she is on the left. you are going _ from different angles. she is on the left. you are going to _ from different angles. she is on the left. you are going to type - from different angles. she is on the left. you are going to type in... - from different angles. she is on the left. you are going to type in... 0r| left. you are going to type in... or ami left. you are going to type in... or am i going to talk? you direct. she can hear me, can she? as i go along, presumably she is going to be a moving avatar? will she be animated? exactly. that is one of the things that we will be developing. i5 exactly. that is one of the things that we will be developing. is she listenin: ? that we will be developing. is she listening? one _ that we will be developing. is she listening? one second. _ that we will be developing. is she listening? one second. ok, - that we will be developing. is she listening? one second. ok, go i that we will be developing. is she listening? 0ne second. 0k, go ahead. what do you think about agi and

where it is going? fiifi what do you think about agi and where it is going?— what do you think about agi and where it is going? ok, give her a second. just a second. where is the sound? this is precisely the point, people will be using this technology and we need to see how responsive it is and how long the latency is. it is not very useful if you have to stand there at the atm and it doesn't work. ,, , , ., , ., , there at the atm and it doesn't work. , work. she seems to be a bit shy in terms of talking. _ work. she seems to be a bit shy in terms of talking. i _ work. she seems to be a bit shy in terms of talking. i do _ work. she seems to be a bit shy in terms of talking. i do not - work. she seems to be a bit shy in terms of talking. i do not speak. work. she seems to be a bit shy in terms of talking. i do not speak to j terms of talking. i do not speak to her over text. you can still talk to her. she is saying, you want a power 9000 nightmare or a skynet scoop? it makes me nervous, this technology, like it makes other people nervous. it makes you nervous as well. let us

see what she says.— see what she says. feeling apprehensive, _ see what she says. feeling apprehensive, huh? - see what she says. feeling apprehensive, huh? ai - see what she says. feeling i apprehensive, huh? ai taking see what she says. feeling - apprehensive, huh? ai taking a walk in the _ apprehensive, huh? ai taking a walk in the wild _ apprehensive, huh? ai taking a walk in the wild side dials up the uneasy style _ in the wild side dials up the uneasy style. what is on your mind? that if ou are style. what is on your mind? that if you are smarter _ style. what is on your mind? that if you are smarter than _ style. what is on your mind? that if you are smarter than me, _ style. what is on your mind? that if you are smarter than me, then - style. what is on your mind? that if. you are smarter than me, then maybe we don't need me. if: you are smarter than me, then maybe we don't need me.— we don't need me. c what she says. sure, we don't need me. c what she says. sure. because _ we don't need me. c what she says. sure. because a _ we don't need me. c what she says. sure, because a clever— we don't need me. c what she says. sure, because a clever algorithm i sure, because a clever algorithm totally— sure, because a clever algorithm totally replaces the human touch, right? _ totally replaces the human touch, right? seriously, no amount of tech can mimic— right? seriously, no amount of tech can mimic the real draw human creativity— can mimic the real draw human creativity or emotion, what has you feeling _ creativity or emotion, what has you feeling replaced? laughter a fundamental issue here is whether agi can replace human intelligence. all right... c what says. agi can replace human intelligence. all right. .. c what says.— all right... c what says. touching on some all right. .. c what says. touching on some existential— all right... c what says. touching on some existential stuff, - all right... c what says. touching on some existential stuff, there. | on some existential stuff, there. the question, can ai replicate the

deep _ the question, can ai replicate the deep and — the question, can ai replicate the deep and often messy human intelligence? it is a bait for the ages _ intelligence? it is a bait for the ages do — intelligence? it is a bait for the ages. do you think it is up to the task? _ ages. do you think it is up to the task? is — ages. do you think it is up to the task? is something irreplaceable from _ task? is something irreplaceable from human cognition? | task? is something irreplaceable from human cognition?— task? is something irreplaceable from human cognition? i think i'm irreplaceable. _ from human cognition? i think i'm irreplaceable. we _ from human cognition? i think i'm irreplaceable. we are _ from human cognition? i think i'm irreplaceable. we are getting - from human cognition? i think i'm irreplaceable. we are getting a . irreplaceable. we are getting a feeling, not only is she responding, she is colloquial. she talks like me. this point professor marcus was talking about, stephanie, the empathy, that relationship, that is what people are looking for in al. go yes, but the woman on the show —— for the woman on the show, i am freaked out that you are referring her as a she. and that is quite a sexy picture that you have picked for shilo. she is quite hot, she has a babe, and she is a she. and i cannot help but notice that he built her and you are a man. i am thinking that if i'm your demographic is somebody who will be using the tool, i am not feeling a lot of empathy with shilo. i am like, why do we

keep calling it a she, why not in it? �* , , , ., keep calling it a she, why not in it? because she is a character. can i... can it? because she is a character. can l- -- can i — it? because she is a character. can l--- can i have— it? because she is a character. can i... can i have a— it? because she is a character. can i... can i have atom_ it? because she is a character. can i... can i have a tom hiddleston i i... can i have a tom hiddleston character? _ i... can i have a tom hiddleston character? there _ i... can i have a tom hiddleston character? there will— i... can i have a tom hiddleston character? there will be - i... can i have a tom hiddleston character? there will be hes, i i... can i have a tom hiddleston i character? there will be hes, shes, dess. character? there will be hes, shes, dess- you — character? there will be hes, shes, dess. you programme _ character? there will be hes, shes, dess. you programme to _ character? there will be hes, shes, dess. you programme to this. is i character? there will be hes, shes, l dess. you programme to this. is she a mirror of — dess. you programme to this. is she a mirror of you _ dess. you programme to this. is she a mirror of you are _ dess. you programme to this. is she a mirror of you are is _ dess. you programme to this. is she a mirror of you are is she _ dess. you programme to this. is she a mirror of you are is she making i a mirror of you are is she making her own decisions? 0r a mirror of you are is she making her own decisions? or it? this a her own decisions? or it? as a writer, i _ her own decisions? or it? as a writer. i see — her own decisions? or it? as a writer, i see her— her own decisions? or it? as a writer, i see her as _ her own decisions? or it? as a writer, i see her as a - her own decisions? or it? as a| writer, i see her as a character, her own decisions? or it? as a writer, i see her as a character, as a writer~ _ writer, i see her as a character, as a writer~ i— writer, i see her as a character, as a writer~ i took _ writer, i see her as a character, as a writer. i took the same approach. i a writer. i took the same approach. i can— a writer. i took the same approach. i can understand if, you know, of course, _ i can understand if, you know, of course, she — i can understand if, you know, of course, she has a certain luck and everything. — course, she has a certain luck and everything, because her character is that she _ everything, because her character is that she is _ everything, because her character is that she is from the 19905, the early— that she is from the 19905, the early 19905, that she is from the 19905, the early19905, she that she is from the 19905, the early 19905, she kind of lives in the i990s _ early 19905, she kind of lives in the 19905. that is the idea behind it. the 19905. that is the idea behind it there _ the 19905. that is the idea behind it. there will be other characters with different looks. i like the idea _ with different looks. i like the idea of— with different looks. i like the idea of giving them pet owners, and notjust_ idea of giving them pet owners, and notjust having it be an it. that is what _ notjust having it be an it. that is what we — notjust having it be an it. that is what we are _ notjust having it be an it. that is what we are trying not to do. why do ou want what we are trying not to do. why do you want to —

what we are trying not to do. why do you want to make _ what we are trying not to do. why do you want to make it _ what we are trying not to do. why do you want to make it more _ what we are trying not to do. why do you want to make it more human i what we are trying not to do. why do | you want to make it more human and -ive you want to make it more human and give it— you want to make it more human and give it a _ you want to make it more human and give it a persona? _ you want to make it more human and give it a persona? will— you want to make it more human and give it a persona? will that— you want to make it more human and give it a persona? will that do- you want to make it more human and give it a persona? will that do not. give it a persona? will that do not consume — give it a persona? will that do not consume to — give it a persona? will that do not consume to confuse _ give it a persona? will that do not consume to confuse people? i give it a persona? will that do not. consume to confuse people? young people _ consume to confuse people? young people and — consume to confuse people? young people and children? _ consume to confuse people? young people and children? anyone - consume to confuse people? young people and children? anyone with. consume to confuse people? young. people and children? anyone with an anthropomorphised _ people and children? anyone with an anthropomorphised ink— people and children? anyone with an anthropomorphised ink relationship i anthropomorphised ink relationship with technology _ anthropomorphised ink relationship with technology. is _ anthropomorphised ink relationship with technology. is it _ anthropomorphised ink relationship with technology. is it dangerous? l anthropomorphised ink relationship with technology. is it dangerous? i| with technology. is it dangerous? i do with technology. is it dangerous? do not think so. this is built for creatives. each of our characters has certain characteristics. and so, they have certain things that they are interested in, and certain tastes, and that is what i think is currently missing if you look at chatgpt, for instance. as a creative, i don't want to interact with on it. i want to interact with... b, with on it. i want to interact with... �* ,,, with on it. i want to interact with...- not - with on it. i want to interact i with. . .- not necessarily. it with... a babe? not necessarily. it is not about _ with... a babe? not necessarily. it is not about that. _ with... a babe? not necessarily. it is not about that. it _ with... a babe? not necessarily. it is not about that. it is _ with... a babe? not necessarily. it is not about that. it is about i is not about that. it is about connecting, right? you know, iam a child of the 90s. so, yes, that was my first starting point.— my first starting point. shilo is aoian my first starting point. shilo is aoain to my first starting point. shilo is going to return _ my first starting point. shilo is going to return to _ my first starting point. shilo is going to return to the _ my first starting point. shilo is i going to return to the programme, we will develop her. we are also going

to talk to you about suno, the music app to talk to you about suno, the music app which creates music on the basis of taste. we are squeezed for time so we will do that at the end of the programme. perhaps you can play it out. losing your voice to cancer is a terrible thing for everyone, but especially if it is what you rely on for your living. this is a man who found himself in the situation until he found a new technology to help him rebuild his acting career. haste him rebuild his acting career. have ou heard him rebuild his acting career. have you heard of _ him rebuild his acting career. have you heard of the _ him rebuild his acting career. have you heard of the recent _ him rebuild his acting career. the you heard of the recent success of flanders? you may not realise it, but the anonymous author who penned the tale was me. i've been a professional voice user all my working life. i injured my voice shouting unnecessarily in a theatre production in the west end and it never recovered properly from that.

i then discovered one day that i was no longer able to speak properly. my voice sounded like i had a dreadful hangover. but i knew immediately that i was in trouble, something was wrong. i decided to give it a few days to see if it would come back to normal, but it didn't. and ijust had to go and see a consultant, who diagnosed squamous cell carcinoma, stage one cancer. and the only thing i could do then was rescue my life by surgery, and have a total laryngectomy, which is the removal of the voice box. iam i am really pleased donal is with me along with his agent, the director of soho voices who has been helping donal recreate his voice. welcome to

you both. if i shut my eyes and listen to the adverts are know that you have done with this technology, i would think it was you speaking. it is truly remarkable. how did it come about?— it is truly remarkable. how did it come about? ~ ., , come about? well, peter was a very determined — come about? well, peter was a very determined that _ come about? well, peter was a very determined that i _ come about? well, peter was a very determined that i should _ come about? well, peter was a very determined that i should not - come about? well, peter was a very determined that i should not lose i come about? well, peter was a very| determined that i should not lose my career. we have worked together for over a dozen years, and aside from professional relationships, we are very, very good friends. and peter began investigating a company who create this type of cloning of the voice. and i independently was researching the same area. now, i sing with acquire called the shout out to cancer acquire. we are a group of people who have had lower

inject emmys. there is not a voice between us. i call it singing outside the box.— between us. i call it singing outside the box. �* ., ., ., ~' outside the box. and how do you work with donal, peter? _ outside the box. and how do you work with donal, peter? how— outside the box. and how do you work with donal, peter? how do _ outside the box. and how do you work with donal, peter? how do you - outside the box. and how do you work with donal, peter? how do you sell. with donal, peter? how do you sell it to clients? presumably you are honest about what it is? brute it to clients? presumably you are honest about what it is?- it to clients? presumably you are honest about what it is? we are at the beginning _ honest about what it is? we are at the beginning of— honest about what it is? we are at the beginning of this _ honest about what it is? we are at the beginning of this process i honest about what it is? we are at the beginning of this process of. the beginning of this process of presenting donal to the world. donal 1.0. presenting donal to the world. donal w i_ presenting donal to the world. donal w i am _ presenting donal to the world. donal w i am not— presenting donal to the world. donal 1.0. i am not unlike a blind presenting donal to the world. donal 1.0. lam not unlike a blind runner who is— 1.0. lam not unlike a blind runner who is running a marathon in the olympics. — who is running a marathon in the olympics, and they have somebody running _ olympics, and they have somebody running along with them holding their hand to guide them in the right— their hand to guide them in the right way _ their hand to guide them in the right way. with the use of donal pass— right way. with the use of donal pass voice. _ right way. with the use of donal pass voice, and knowing donal as well as— pass voice, and knowing donal as well as i— pass voice, and knowing donal as well as i do— pass voice, and knowing donal as well as i do as an actor and as a voicea _ well as i do as an actor and as a voice. i— well as i do as an actor and as a voice. iam— well as i do as an actor and as a voice, i am able to read on his behalf— voice, i am able to read on his behalf and _ voice, i am able to read on his

behalf and use the ai algorithm that we alone _ behalf and use the ai algorithm that we alone have the licence to use through— we alone have the licence to use through the person who created it, so that— through the person who created it, so that i_ through the person who created it, so that i can produce a piece using dohal's_ so that i can produce a piece using donal's voice but with my intonation and knowledge, and donal can be there _ and knowledge, and donal can be there and — and knowledge, and donal can be there and give me guidance as well as to _ there and give me guidance as well as to how— there and give me guidance as well as to how he thinks he might do it. i as to how he thinks he might do it. i have _ as to how he thinks he might do it. i have some — as to how he thinks he might do it. i have some real interest in this. my i have some real interest in this. my mother lost her voice due to mnd and we were told to record her voice, but we were not able to replicate it. i can see enormous value in this. i know that commercially there are actors using it in ourfour commercially there are actors using it in our four audiobooks, commercially there are actors using it in ourfour audiobooks, to take off some of the pressure for their careers. is that how you see it developing? iii careers. is that how you see it developing?— careers. is that how you see it develo-ina ? . ., ., developing? if i were a name actor who recorded _ developing? if i were a name actor who recorded a _ developing? if i were a name actor who recorded a lot _ developing? if i were a name actor who recorded a lot of— developing? if i were a name actor who recorded a lot of books, i developing? if i were a name actor who recorded a lot of books, then | developing? if i were a name actor| who recorded a lot of books, then it is quite _ who recorded a lot of books, then it is quite a _ who recorded a lot of books, then it is quite a long winded process. it takes _ is quite a long winded process. it takes you — is quite a long winded process. it takes you a —

is quite a long winded process. it takes you a good week or so, depending on the length of the book, to record _ depending on the length of the book, to record it _ depending on the length of the book, to record it. but, with this kind of technology — to record it. but, with this kind of technology, if the actor had an avatar— technology, if the actor had an avatar made of him, or her, or they, then— avatar made of him, or her, or they, then they— avatar made of him, or her, or they, then they can— avatar made of him, or her, or they, then they can use that to reach 100 books _ then they can use that to reach 100 books and — then they can use that to reach 100 books. and i believe the technology is now _ books. and i believe the technology is now there for that to be brought into other— is now there for that to be brought into other languages. it may not 'ust into other languages. it may not just be _ into other languages. it may not just be in— into other languages. it may not just be in the english language, but it could _ just be in the english language, but it could be — just be in the english language, but it could be heard in other languages as well _ it could be heard in other languages as well so. — it could be heard in other languages as well. so, it could be to augment that actor— as well. so, it could be to augment that actor and given life and enable them _ that actor and given life and enable them to— that actor and given life and enable them to do— that actor and given life and enable them to do things that they couldn't do along _ them to do things that they couldn't do along with the work they currently can do.— do along with the work they currently can do. donal, peter, i wish we could _ currently can do. donal, peter, i wish we could talk— currently can do. donal, peter, i wish we could talk more, - currently can do. donal, peter, i wish we could talk more, we i currently can do. donal, peter, i wish we could talk more, we are currently can do. donal, peter, i. wish we could talk more, we are a little squeeze the time. it is a brilliant thing you have done. congratulations both. thank you for coming and talking to us about it.

thank you. coming and talking to us about it. thank you-— wa nt to want to focus on this other story thatis want to focus on this other story that is quickly coming the goatee for music in these assets. it is known as suno ai which was developed in cambridge and its creator sung in whatever genre you want from some text and audio. while samir mallal has been sat here, he has created some music. what are you going to do with it? i some music. what are you going to do with it? ., some music. what are you going to do with it? . ., ., . ., ., with it? i am going to create a sona. i with it? i am going to create a song- i hope _ with it? i am going to create a song. i hope that _ with it? i am going to create a song. i hope that we - with it? i am going to create a song. i hope that we have i with it? i am going to create a | song. i hope that we have luck with it? i am going to create a i song. i hope that we have luck with the idea this _ song. i hope that we have luck with the idea this time. _ song. i hope that we have luck with the idea this time. it _ song. i hope that we have luck with the idea this time. it is _ song. i hope that we have luck with the idea this time. it is currently i the idea this time. it is currently generating. _ the idea this time. it is currently generating, but... _ the idea this time. it is currently generating, but... you - the idea this time. it is currently generating, but... you only i the idea this time. it is currently| generating, but... you only need the idea this time. it is currently i generating, but... you only need a few words. — generating, but... you only need a few words, not _ generating, but... you only need a few words, not of _ generating, but... you only need a few words, not of an _ generating, but... you only need a few words, not of an interview? i generating, but... you only need a i few words, not of an interview? what genre will be get? 3�*05 few words, not of an interview? what genre will be get?— genre will be get? 70s rock. and we have aot genre will be get? 70s rock. and we have got some _ genre will be get? 70s rock. and we have got some pop. _ genre will be get? 70s rock. and we have got some pop. letters - genre will be get? 70s rock. and we have got some pop. letters have i genre will be get? 70s rock. and we have got some pop. letters have a i have got some pop. letters have a . o . _

music plays # it was hard when i lost my voice # a long time there is a whole concern about this for music people. it sounds like some music. you can imagine it cramming spotify. the original artists will not get a look in. .. ., ., ., look in. one cannot imagine what date it was _ look in. one cannot imagine what date it was trained _ look in. one cannot imagine what date it was trained on _ look in. one cannot imagine what date it was trained on and - look in. one cannot imagine what date it was trained on and if i look in. one cannot imagine what date it was trained on and if the l date it was trained on and if the artists— date it was trained on and if the artists had _ date it was trained on and if the artists had consent _ date it was trained on and if the artists had consent and - date it was trained on and if the. artists had consent and copyright issues _ artists had consent and copyright issues etc — artists had consent and copyright issues etc l _ artists had consent and copyright issues etc. ., artists had consent and copyright issues etc-— issues etc. i agree with you. clearly. _ issues etc. i agree with you. clearly, this _ issues etc. i agree with you. clearly, this was _ issues etc. i agree with you. clearly, this was trained i issues etc. i agree with you. clearly, this was trained on | clearly, this was trained on copyrighted data. obviously, a song is much more thanjust... this feels

one—dimensional. would anybody want to listen to this? this is not a tiny dancer. clearly. the question is, how is this going to be used? i think it goes beyond the ethical, it is also, is this good? stand think it goes beyond the ethical, it is also, is this good?— is also, is this good? and if you can connect _ is also, is this good? and if you can connect to _ is also, is this good? and if you can connect to it. _ is also, is this good? and if you can connect to it. listen, i is also, is this good? and if you can connect to it. listen, we're| is also, is this good? and if you i can connect to it. listen, we're out of time. stephanie hare and samir mallal, thank you. you are going to players out. mallal, thank you. you are going to players out-— players out. yes, this is a reggae version. music plays yes. perfectly authentic. _ music plays yes. perfectly authentic, right? _ good evening. bit of a weather cliche, i know, but a north—south divide with the weather story today. we had some beautiful sunshine, and with lighter winds across central and southern england, it felt a little warmer as well. this was londonjust a few hours ago. different story further north. it was cloudy and wet at times, a rather drizzly, overcast picture, as you can see in argyll and bute.

and that's because of this weather front. it's toppling around the high pressure that's pushing in for the weekend, that's bringing some rain. now, the progress of the rain has been quite slow. it's been moving its way out of scotland, pushing into northern england and north wales. for the rest of the day, it will gradually sink its way steadily south. so, we will see some outbreaks of rain through the night tonight across central and southern england. clearer skies further north, a brisk wind and a few scattered showers. and with temperatures to the tops of the mountains, perhaps into low single figures, maybe a little bit of a wintry flavour here as well, but a milder start to the morning for england and wales. a dreary, drab morning here, slowly brightening up into the afternoon. sunny spells, a few scattered showers, but, again, the winds picking up. that'lljust take the edge off the feel of things, particularly in eastern scotland — we could see gusts in excess of 40 miles an hour.

so here, only a high of 8 celsius in aberdeenshire, with a little more shelter further west, we could — with more sunshine — see 14—15 celsius. here's the high that's going to arrive through the weekend. a lot of dry weather around at the weekend. this weather front will just introduce a little bit of patchy drizzle. and as we go into sunday in particular, more of a northeasterly flow, making it feel cool on exposed east coasts and certainly producing a lot more in the way of cloud. so we will be chasing cloud amounts around this weekend. but on the whole, saturday starts off sunny. there will be a brisk east wind, north east wind coming through, making it feel cool here. further west, with a little more sunshine, we could see highs of 14 celsius. so this west—east divide continues, particularly for the second half of the weekend. in the east, it will be a gray, drab affair and the temperatures struggling, i'm afraid, from time to time. with a little more shelter and more sunshine in the west, it's not out of the question, we may see 17 celsius in parts of northern ireland.

hello, i'm christian fraser. you're watching the context on bbc news. if ukraine will fall, so the global, so the global system of secuirty wil be destroyed and all the world will need to find and will need to look for a new system of secuity i think providing lethal aid to ukraine right now is critically important i do. i really do believe the intel and the briefings that we've gotten that, i believe xi and vladimir putin and iran really

are an axis of evil. there are frozen russian assets in the uk and europe and elsewhere we should be finding ways of using those assets to help ukraine in its defence of this appaling, illegal russian invasion. the political delays, as kyiv points out, have caused ukraine to lose territory and soldiers. lives have been lost they argue because they have soldiers with dwindling ammunition, artillary shells and long range missiles. with the vote pending in washington a special programme tonight on ukraine and middle east. our panel... karolina hird — russia analyst at the washington based institute for the study of war, the former communications director for the republican national convention—doug heye. ukrainian politician, inna sovsun and the former head of the russia desk at the british mi6 intelligence service, chris steele