but the industry insists that vapes are recyclable and help smokers give up cigarettes. and tennis star ons jabeur hopes to make history for tunisia and africa by winning a grand slam title later today. she's facing czech marketa vondrousova, the first unseeded player to reach the women's singles final in 60 years. now on bbc news, it's time for talking business. hello, everybody. a very warm welcome to talking business. let's have a look at what is on the show. the risk and rewards of artificial intelligence. in business, it is already being talked about as a generational game changer across—the—board, from farming to finance and from construction to customer service. finance and from construction to customerservice. some finance and from construction to customer service. some satyrs going

to destroyjobs, kill industries and could even blow up the world. others say it could help cure disease, create newjobs and even help solve global warning. where should we be opening our arms to artificial intelligence and where should we be keeping it at arm's—length and who should be making those decisions? i am going to find out from these two. one of the founders of ai for goat and one of the world apostle vs leading computer who has helped shape the official uk strategy. also a legend from the advertising world. he built the world are possibly s biggest advertising company for soppy talks about how this technology is helping change his industry already.

wherever you adjoining me from around the world once again a big hello and a warm welcome to the show. have you heard about a little thing called ai? unless you have been living under a rock, you cannot have failed to have noticed the phenomenon that is being called the new tech revolution. artificial intelligence itself has been around for years but since the launch of the latest large language model processes, your chat gpt, for example, we have all begun to take notice and the reaction has been healed. it has sparked a gold rush in the tech world and a new global —— for a piece of the cake. the ability for computer to create humanlike responses to complex problems offers huge changes to the way that we live and work. you will also be made aware of the potential pixels. it is being sounded by front

line workers sounded, worried about their work. line workers sounded, worried about theirwork. democracies line workers sounded, worried about their work. democracies fear the power of misinformation and deep lakes and some of the most distinguished people building the tech voicing their fears. the people behind jack gpt and others all signing a letter stating that mitigating the risk of extinction from al should be a global priority... that certainly makes you think. all part of the reason why people and countries are approaching ai with a mixture of fear and excitement. optimism and concern put into numbers in one recent survey of workers, with over a third saying that they thought their jobs were likely to be eliminated by artificial intelligence. front line

workers and surprisingly more negative about the impact of machine learning than managers. and international differences also showed up. the most optimistic, brazil, india and the middle east. and the least optimistic, the united states, the netherlands and japan. my states, the netherlands and japan. my first guest is certainly optimistic about the possibilities of the technology. in fact, she is one of the founders of the organisation ai for good. a real pleasure having you on the programme. we have looked at how different countries are approaching ai. i am wondering, are using different regions embracing the tech and others that are not? we different regions embracing the tech and others that are not?— and others that are not? we are seeinu and others that are not? we are seeing pretty — and others that are not? we are seeing pretty much _ and others that are not? we are seeing pretty much a _ and others that are not? we are seeing pretty much a race - and others that are not? we are seeing pretty much a race to - and others that are not? we are i seeing pretty much a race to being the first in al, the leader in al around the world. the approach however is different. some countries are taking more thoughtful approaches to building it safely, taking a cleaner... clearer view on

the risk this technology poses was up the risk this technology poses was up regulation is emerging in a similar way. up regulation is emerging in a similarway. european up regulation is emerging in a similar way. european commission is putting this forward. there are others who are more out to be faster in adopting and implementing ai across the border. the approach is different. but the application is pretty global at this point. as is the talent that is involved in developing it.— the talent that is involved in developing it. the talent that is involved in develoin: it. �* ., ., , developing it. and how have you seen the reaction — developing it. and how have you seen the reaction from _ developing it. and how have you seen the reaction from developing - the reaction from developing economies? i the reaction from developing economies?— the reaction from developing economies? ., ., ., ., ., ., ~' economies? i had done a lot of work in buildin: economies? i had done a lot of work in building ai — economies? i had done a lot of work in building ai solutions _ economies? i had done a lot of work in building ai solutions for— in building ai solutions for communities around the world in supporting them, using these tools to get access to services, such as helping them understand the law, or issues around mental health, in many communities around the world. i had in a lot of work in south africa, india, and so on. and what we are seeing is the impact of this technology is so much more scalable in communities where there are

access issues from the get go. it is not a replacement of human skills, it is an augmentation where we just don't have critical infrastructure at a fast enough pace. to give you an example, in communities where front line workers are overworked, exhausted, do not have the right tools, this technology can help people take that first step a little bit easier and connect to the right services. this is about augmentation in those communities. what we are not seeing being solved with al though is the lack of digital infrastructure, where conductivity is already an issue, it does not... it limits the use of ai in many ways. i would say there is a real need to improve the underlying infrastructure, so we can start using and reaping the benefits of ai for everyone. using and reaping the benefits of ai for everyone-— using and reaping the benefits of ai for everyone. where do you see the most potential. _ for everyone. where do you see the most potential, from _ for everyone. where do you see the most potential, from a _ for everyone. where do you see the | most potential, from a development angle, for al to be used? i see most potential, from a development angle, for al to be used?— angle, for al to be used? i see the most potential— angle, for al to be used? i see the

most potential in _ angle, for al to be used? i see the most potential in areas _ angle, for al to be used? i see the most potential in areas where - angle, for al to be used? i see the | most potential in areas where there is a lot of knowledge and information already available but the applications of it have not really fully come to fruition, so one example is in the world of law. where we have huge amounts of information available around the law and how it works and historical knowledge that has been created by people who are professionals in this field. but accessing it often is difficult. we are all quite familiar with the law firm associate who has to work until 2am to get information passed the deadline. but we are now seeing applications such as creating a new draft, a first draft, using artificial intelligence will topic can help you do that within minutes, work that would previously take days. this information access of knowledge that already exist, information that exist, but is not available in a trusted way at your fingertips, is now being solved with artificial intelligence. if we can

use ai for social impact, social good, making knowledge available for everybody, for giving trusted information, solving real—world challenges that face us today, i will be very happy, but we can only do this if we use ai in a safe way, if we build ai in a trusted way and if we build ai in a trusted way and if we build ai in a trusted way and if we can get this regulation right globally. if we can get this regulation right aloball . �* , ., ,, if we can get this regulation right aloball . �* , ., i. ~' if we can get this regulation right aloball . �* , ., ,, ~' . globally. briefly, do you think that will happen? _ globally. briefly, do you think that will happen? do — globally. briefly, do you think that will happen? do you _ globally. briefly, do you think that will happen? do you think- globally. briefly, do you think that will happen? do you think we - globally. briefly, do you think that will happen? do you think we willl will happen? do you think we will get mankind... as humans, will we get mankind... as humans, will we get it right? a lot of time, we get things wrong. i get it right? a lot of time, we get things wrong-— get it right? a lot of time, we get things wrong. i think we have a lot of ulobal things wrong. i think we have a lot of global support _ things wrong. i think we have a lot of global support at _ things wrong. i think we have a lot of global support at this _ things wrong. i think we have a lot of global support at this point - things wrong. i think we have a lot of global support at this point and | of global support at this point and there is a lot of awareness around what this technology is about to do for humanity. i think we have to do this safely. safeguard against some of the challenges where we might not be able to get it right completely and this is where regulation comes in. i also think, ifundamentally believe, that we have the one area

where we need the most were, we have not done yet, is skills and newjob creation. there will be aimed a lot of new technology opportunities, dust —— just anyway the digital era did. i would encourage policymakers, global leaders, business leaders, to take an active well thought through proactive approach to skills development and newjob creation. let me ask you this, because ai uses plain language, whether it is english, french, spanish, italian, etc, not coding language, does that open the doorfor etc, not coding language, does that open the door for nontechnical, like myself? to be able to, i don't know, do anything with al? i myself? to be able to, i don't know, do anything with al?— do anything with al? i think that the potential— do anything with al? i think that the potential for _ do anything with al? i think that the potential for people - do anything with al? i think that the potential for people from i do anything with al? i think that i the potential for people from many different backgrounds, including language, linguistics, to come and be part of the ai creation is really, really exciting and interesting. for example, we have the model of instead of writing

code, we write prose, we talk to machines, i spend code, we write prose, we talk to machines, ispend more code, we write prose, we talk to machines, i spend more time talking to machines and humans. and this is an interesting time in how we can bring human curiosity to train the machines to do better things and it is now commonly said the newest programming language is playing human language. absolutely, this goes to the point of creating new skills and new roles that are going to be generated and may be one of those will be a ai conversation designer. fit, those will be a ai conversation desiuner. �* ., those will be a ai conversation desiuner. . , ., those will be a ai conversation desitner. �* . , . , . , ., designer. a real pleasure having you on the show- — designer. a real pleasure having you on the show. thank— designer. a real pleasure having you on the show. thank you _ designer. a real pleasure having you on the show. thank you for - designer. a real pleasure having you on the show. thank you for your - designer. a real pleasure having you| on the show. thank you for your time and we will talk to you soon. i on the show. thank you for your time and we will talk to you soon.- and we will talk to you soon. i look hoto and we will talk to you soon. i look photo talking _ and we will talk to you soon. i look photo talking to — and we will talk to you soon. i look photo talking to you _ and we will talk to you soon. i look photo talking to you again - and we will talk to you soon. i look photo talking to you again and - and we will talk to you soon. i look photo talking to you again and i . photo talking to you again and i hope you admissions have a happy friendship. the hope you admissions have a happy friendshi -. ., , hope you admissions have a happy friendshi. . , ., friendship. the fears about the possibilities _ friendship. the fears about the possibilities of _ friendship. the fears about the possibilities of this _ friendship. the fears about the possibilities of this technology| friendship. the fears about the i possibilities of this technology are massive as well, leaving governments around the world or scrubbing to catch and regulate it. i sat down with my next guest, who has been helping shape the uk ai strategy. she is one of the leading global academics in the field. a real pleasure having you on the

show. let's start with this, because there seems to be a great deal... the fear of missing out. everybody rushing in and it does feel like a rash. is the technology that is there now, is it ready for all of this? ., ~ there now, is it ready for all of this? ., . ., ., this? no. well, i mean, if we are talkin: this? no. well, i mean, if we are talking about _ this? no. well, i mean, if we are talking about al, _ this? no. well, i mean, if we are talking about al, then _ this? no. well, i mean, if we are talking about al, then ai - this? no. well, i mean, if we are talking about al, then ai has - this? no. well, i mean, if we are| talking about al, then ai has been around a long time and we all use in our every day lives, facial recognition, predictive text, all the things we use on our phones, it is all driven by ai, different types of technologies that have been involved in research labs for a long time. i am involved in research labs for a long time. iam beginning involved in research labs for a long time. i am beginning to think that if you remember back in the 19905 when the world wide web first came along, and built on top of the internet, and everyone piled in to suddenly u5e internet, and everyone piled in to suddenly use this web, which was free to use, ring a bell? like chat gpt, free to use? busine55 models

were not there and the technology was not ready. the underlying infrastructure wa5 was not ready. the underlying infrastructure was not ready for the mass use of the world wide web. if you remember, we used a quality worldwide weight because do you render how long it is to take to download a picture? and then it will time out because we had modems. i am just beginning to think that we might be in that place with generative ai today because i think the investors will pile in now. it looks like they are already doing it. . ., ., y looks like they are already doing it. . ., ., , ., �* looks like they are already doing it. the technology, i don't think, is read . it. the technology, i don't think, is ready- and _ it. the technology, i don't think, is ready. and so _ it. the technology, i don't think, is ready. and so for— it. the technology, i don't think, is ready. and so for mass - it. the technology, i don't think, is ready. and so for mass use, i it. the technology, i don't think, | is ready. and so for mass use, in it. the technology, i don't think, l is ready. and so for mass use, in a way that companies can make money out of it was not if you remember, with the internet, we had to learn how to make money out of it. that is where the advertising came along. the inve5tor5 where the advertising came along. the investors are there that the technology is just not ready. it technology isjust not ready. it might not be ready. ijust that we might not be ready. iju5t that we are going to see some attempt to make money in this area, to create

commercial companies, which is what we want to see because this is fabulous technology. it is going to help us a huge amount to do things we could not do before. but i am just not quite sure that... i think the investment might be again ahead of the curve in terms of the technology. of the curve in terms of the technology-— of the curve in terms of the technolo: . ., . ., ., ., ., technology. how much of a threat do ou seeinu technology. how much of a threat do you seeing things — technology. how much of a threat do you seeing things like _ technology. how much of a threat do you seeing things like deep - technology. how much of a threat do you seeing things like deep fakes - you seeing things like deep fakes for example? ai having the ability to mimic human content, how do we 5afeguard to mimic human content, how do we safeguard against that? it to mimic human content, how do we safeguard against that?— safeguard against that? it easily difficult and _ safeguard against that? it easily difficult and i _ safeguard against that? it easily difficult and i think _ safeguard against that? it easily difficult and i think at _ safeguard against that? it easily difficult and i think at the - safeguard against that? it easily l difficult and i think at the moment what i am saying is to people that there might be an existential threat down the line, it is a long way away, we need to learn to regulate the use of ai and how people... companies u5e the use of ai and how people... companies use it. but the immediate threat5 companies use it. but the immediate threats are the deep fake5, this information we have a number of major democratic elections coming up next week, the us president, the uk, spain, the eu, they all have

democratic elections and i think the idea of how we are going to control the fakes comedy disinformation, i5 the fakes comedy disinformation, is much more a clear and present danger than any accidental threat. what than any accidental threat. what about the opportunities? - than any accidental threat. what about the opportunities? what has got you excited about this? the top answer is always _ got you excited about this? the top answer is always health _ got you excited about this? the top answer is always health care. - got you excited about this? the top answer is always health care. we i got you excited about this? the top. answer is always health care. we are beginning to see ai, the image processing piece of ai can analyse scans of tumours, potential tumours, and check them fa5ter scans of tumours, potential tumours, and check them faster and more accurately than a human being. but the other when i am excited about is education. when chat gpt came out, everyone said, we cannot have the kids u5ing everyone said, we cannot have the kids using this. it will help them kid5 using this. it will help them write e55ay5. it is like when you calc later came out. i write essays. it is like when you calc later came out. i remember that at school. calc later came out. i remember that at school- you _ calc later came out. i remember that at school. you might _ calc later came out. i remember that at school. you might want _ calc later came out. i remember that at school. you might want to - calc later came out. i remember that at school. you might want to learn i at school. you might want to learn how to do long _ at school. you might want to learn how to do long division _ at school. you might want to learn how to do long division but - at school. you might want to learn how to do long division but you - how to do long division but you don't have to do it in your daily life. my father was an accountant and he used to divide pounds, shillings and pence by hand. all

those jobs have gone shillings and pence by hand. all thosejobs have gone now shillings and pence by hand. all those jobs have gone now but there are more jobs those jobs have gone now but there are morejob5 in the those jobs have gone now but there are more jobs in the finance industry than you could ever imagine because of the creative things we have learnt to do, using computers and we will do the same. when we have software that helps us write and absorb information in a form thati5 and absorb information in a form that is a natural language, we will be able to be much more creative around that and i am so excited about that and we will be able to have personalised tutoring, so that children... at the moment, the model of teaching is one person and a0 people in a class, 30 people in a class. we will be able to have personalised tutor5 for kids. my worry more is that making sure there is an equality of... a level playing field, so that all kids get the same of unity. field, so that all kids get the same of uni . ., ., field, so that all kids get the same of uni . . ~ ., ., of unity. talking of a level playing field, one thing _ of unity. talking of a level playing field, one thing we _ of unity. talking of a level playing field, one thing we seem - of unity. talking of a level playing field, one thing we seem to - of unity. talking of a level playing field, one thing we seem to be i field, one thing we seem to be hearing from the industry around the world is regulation. it feels like they are crying out for regulation. we know that you advise the uk

government and you have certainly helped to shape the uk'5 ai 5trategy. what is it that you are telling them? indie strategy. what is it that you are telling them?— strategy. what is it that you are tellin: them? ~ ., ., ., ., telling them? we have to learn how to reuulate telling them? we have to learn how to regulate and _ telling them? we have to learn how to regulate and i _ telling them? we have to learn how to regulate and i think _ telling them? we have to learn how to regulate and i think we _ telling them? we have to learn how to regulate and i think we have - telling them? we have to learn how to regulate and i think we have got| to regulate and i think we have got time to do that but we need to start now and it has to be done globally. there is no point... the eu are going straight to law about this but i personally think that will stifle innovation and it is a good start. that is the tricky part. regulation and not qua5hing innovation. the that is the tricky part. regulation and not quashing innovation. the uk white paper — and not quashing innovation. the uk white paper released _ and not quashing innovation. the uk white paper released in _ and not quashing innovation. the uk white paper released in march - and not quashing innovation. the uk white paper released in march was l white paper released in march was much more about allowing for innovation and using all the things we have already to help us start on the regulation journey. we have already to help us start on the regulationjourney. all of these are possible. then you have got the fact that actually in the eu, we don't have any companie5, fact that actually in the eu, we don't have any companies, they are in the us or china. we really, the us has got to be in line with this,

because in the us, the companies are lobbying washington to get the laws that they want to help them grow as companies, but that does not mean they will take any notice of what we're doing in regulation here. and then we have to bring china to the table. we have to learn how to get it right was if we ru5h table. we have to learn how to get it right was if we rush to do it, we could well get it wrong and we could be regulating for the year and now but not for the future. let be regulating for the year and now but not for the future.— but not for the future. let me end on this. but not for the future. let me end on this- as _ but not for the future. let me end on this- as a _ but not for the future. let me end on this. as a cutting-edge - but not for the future. let me end i on this. as a cutting-edge computer on this. as a cutting—edge computer 5cienti5t, get the crystal ball out, where is this technology in five years' time? i where is this technology in five years" time?— where is this technology in five years' time? i would answer your auestion years' time? i would answer your question by _ years' time? i would answer your question by saying, _ years' time? i would answer your question by saying, what - years' time? i would answer your question by saying, what was - years' time? i would answer your question by saying, what was it l years' time? i would answer your i question by saying, what was it like five years ago? chat gpt did not exist but everything else we have been talking about exi5ted. we will see incremental advances. it will 5ee incremental advances. it will feel like they've waves but if you think back, we did not have a world wide web, 30 years ahead, they will be huge advantage. but i am hoping that we will be able to put in place frameworks for regulating in the future that are flexible and agile

enough to cope with new technologies that come along and we can work out how to manage them in the future. even though we don't know what they are yet. even though we don't know what they are et. ., ., ,.. ,., are yet. you are scaring me in some wa s, but are yet. you are scaring me in some ways. but i — are yet. you are scaring me in some ways. but i will— are yet. you are scaring me in some ways, but i will end _ are yet. you are scaring me in some ways, but i will end on _ are yet. you are scaring me in some ways, but i will end on that. - are yet. you are scaring me in some ways, but i will end on that. a - are yet. you are scaring me in some ways, but i will end on that. a real i ways, but i will end on that. a real pleasure having you on the show. thank you for your time. 0k, thank you for your time. ok, so far thank you foryourtime. ok, so far in thank you for your time. ok, so far in the show, we have been judging the risks and rewards of artificial intelligence and my next gue5t says artificial intelligence and my next guest 5ay5 companies artificial intelligence and my next gue5t says companies that failed to capitalise on al tool5 gue5t says companies that failed to capitalise on al tools will risk losing out. you should know, he created the worlds largest advertising company from a standing start. these days, he is using technology to disrupt the old ways of working at his new company. martin, a real pleasure having won the show. let's start with this, because the show this week is all about balancing the positive and the negatives of ai. in advertising, your business, what are some of the positives and negatives, as you see them? , , , .,

them? just briefly, i think there are five big _ them? just briefly, i think there are five big positives. _ them? just briefly, i think there are five big positives. the - them? just briefly, i think there | are five big positives. the verses we are going to see an improvement in productivity of copywriting and visualisation. basically, we will be able to produce copy and adverts faster, two weeks down to two days. we will be able to personalise even more effectively. we are able to do that already but it is going to be super productive, instead of producing 1.5 million assets for a campaign, potentially, we can produce multi—million assets for a campaign thirdly, media planning, i presume will be revolutionised soppy will no longer need 10,000 people around the world. you'll be able to do it algorithmically and the people will have very sophisticated data to base their planning decisions on. we are going to be able to use ai base their planning decisions on. we are going to be able to use al to processes, daily processes and be much more productive. and what i think is the most important positive is that we will be able to

democratise information inside companies. they will become much flatter, much more effective and efficient because knowledge will be able to be disseminated around the company, almost human bouts, humans will be able to access information much more effectively. there are legal issues around copywriting, infringements, of copy. that is an issue that has been worked out in the music industry and other industries as we speak. there will always be the bad actors that use these technologies in evil ways and thatis these technologies in evil ways and that is something that through self—regulation i think companies have got to put pressure... the regulator is not good to be able to keep up. and finally the issue around jobs. this is going to be a net adding jobs, or a around jobs. this is going to be a net addingjobs, ora net diminishing ofjobs? my own view personally, push me against the wall, it will be probably negative

in terms of netjobs. ihti wall, it will be probably negative in terms of net jobs.— wall, it will be probably negative in terms of net jobs. in terms of net “obs. ai can be used to create in terms of net jobs. ai can be used to create complete _ in terms of net jobs. ai can be used to create complete advertising - to create complete advertising campaigns but i am just wondering this, will consumers be able to see the difference between one crafted by advertising agencies or when that has been put together by an algorithm? iii has been put together by an algorithm?— has been put together by an alaorithm? , ., i. ., ., algorithm? if i showed you an advert that was ai generated, _ algorithm? if i showed you an advert that was ai generated, let's - algorithm? if i showed you an advert that was ai generated, let's say - algorithm? if i showed you an advert that was ai generated, let's say a . that was ai generated, let's say a year ago, you would cvjoints. you would see that it was to some extent manufactured. if i showed you the same process today, a year on, you might not be able to see it. and going forward, i think you definitely won't be able to see it. you will not be able to determine what is market fractured algorithmically —— manufactured algorithmically —— manufactured algorithmically and what is done naturally. algorithmically and what is done naturall . ,, ., algorithmically and what is done naturall. ,, ., ., , ., naturally. should advertising and marketin: naturally. should advertising and marketing companies _ naturally. should advertising and marketing companies be - naturally. should advertising and marketing companies be clear. naturally. should advertising and l marketing companies be clear with their clients where they use ai in their clients where they use ai in their work? indie their clients where they use ai in their work?— their clients where they use ai in their work? ~ . ., ., , ., ~ their clients where they use ai in theirwork? ~ . . ., , .,~ ., their work? we have already taken a osition for their work? we have already taken a

position for example _ their work? we have already taken a position for example that _ their work? we have already taken a position for example that we - their work? we have already taken a position for example that we will. position for example that we will not alter pictures. let's say we had a white actor, not convert that white actor into a black actor or vice versa. we have already put in to place ethical protocols that we think are important. this comes back to the self—regulation point. it is hopeless for the regulator in my view to believe that he or she can keep up with this. what the governments have to do, regular to have to do, is to put pressure on the tech companies to behave in a totally responsible way for having said that, there will always be bad actors. do i think that in next years presidential election, some nefarious people will be using ai years presidential election, some nefarious people will be using al to generate fake images or fake adverts orfake interventions? yes, i think that will happen. pare or fake interventions? yes, i think that will happen.— that will happen. are you also sa in: that will happen. are you also saying then — that will happen. are you also saying then that _ that will happen. are you also saying then that you - that will happen. are you also saying then that you don't - that will happen. are you also - saying then that you don't believe that the world will be able to get

together and globally regulate this technology? ida. together and globally regulate this technolo: ? ., ~' together and globally regulate this technolo: ? ., ~ ., , technology? no, i think we have seen it for example — technology? no, i think we have seen it for example the _ technology? no, i think we have seen it for example the american - it for example the american regulator, it is trying to unravel a deal that google did ten years ago. the regulator tends to always be reactive and looking backwards. they lack the resources to do that. if the government was prepared to pile infinite resources into eight, in terms of people and knowledge and data and maybe they might be, but what they have to do is work with the industry. in what they have to do is work with the industry-— what they have to do is work with the indust . . , ., ., the industry. in a nutshell, who are the industry. in a nutshell, who are the winners — the industry. in a nutshell, who are the winners and _ the industry. in a nutshell, who are the winners and losers _ the industry. in a nutshell, who are the winners and losers in _ the industry. in a nutshell, who are the winners and losers in this - the industry. in a nutshell, who are the winners and losers in this ai . the winners and losers in this ai battle? i the winners and losers in this ai battle? ., the winners and losers in this ai battle? ~' ., , ~' the winners and losers in this ai battle? ~' ., , ,, ., battle? i think new models like our own, i'm obviously _ battle? i think new models like our own, i'm obviously biased, - battle? i think new models like our own, i'm obviously biased, but - battle? i think new models like our own, i'm obviously biased, but we| own, i'm obviously biased, but we are a 9000 people to torpedo boat, rather than 100,000 people aircraft carrier, so agility will be key. from the platforms, meta, amazon,

ali baba, tencent, tiktok, plus apple and microsoft, sales force, adobe, oracle, video, they are making the picks and shovels if you like, the chips that are used for all the computing power, and i think only the larger companies will be able to do it at this sort of scale and this cost and i think inevitably, as they are forced to grow organically because of regulation, i think you will see them expand into it. i would say the six platforms would be western, eastern, press apple and microsoft will be the interesting ones. just briefl , will be the interesting ones. just briefly. five _ will be the interesting ones. just briefly, five months were a year's time from now, what is the impact of ai time from now, what is the impact of al on advertising and marketing? i alon advertising and marketing? i wish i knew. it is different, five minutes to five days, even five

years was having said that, i think it will be a much more efficient, much more effective marketing model both bit will be algorithmically driven, the human creative side will be increasingly important but in a different level. it will be those decisions and those views and those judgments will be made with much more effective data than we have ever had before and i think it will be a much more interesting actually environment because we will have much more data, we will have companies that are leaner, that are more efficient, that are flatter, and where people across the organisation are much more involved in decision—making and much more involved in... have much more access to information. information is power. it is going to democratise knowledge inside companies and going to make people much more productive

and much more effective.— and much more effective. always a leasure and much more effective. always a pleasure seeing — and much more effective. always a pleasure seeing you. _ and much more effective. always a pleasure seeing you. thank- and much more effective. always a pleasure seeing you. thank you - and much more effective. always a pleasure seeing you. thank you for coming on the show. we will check in with you soon. coming on the show. we will check in with you soon-— with you soon. thank you very much. that is it for — with you soon. thank you very much. that is it for this _ with you soon. thank you very much. that is it for this week. _ with you soon. thank you very much. that is it for this week. i _ with you soon. thank you very much. that is it for this week. i hope - with you soon. thank you very much. that is it for this week. i hope you i that is it for this week. i hope you enjoy the show both you can keep up with the latest on the bbc website and the smartphone app. you can also follow me on twitter. thank you for watching and i will see you soon. goodbye. hello, there. many places had a soaking on friday, courtesy of low pressure. in fact, some places had a month's worth of rain. now, that low—pressure is still with us and it will bring thundery downpours today, heavy showers probably for all, but also unseasonably windy weather. we have already seen that across southern and western areas during friday. the stronger winds are also developing further east now. we have warnings out as a result of that, as you can see, stretching across the whole of southern england, south wales, part of the midlands and east anglia.

gusts of wind up to 55 mph could bring down the odd weak tree. certainly branches and slates and cause some disruption. we have also got showers to go with and longer spells of rain further north and then thunderstorms developing, so there are warnings out there now scotland and northern england as well, for those thunderstorms, but anywhere could see some heavier showers and some really intense downpours. temperatures around average for the time of year. but a little sunnier potentially later in the day, lifting the temperatures, 22 celsius across parts of east anglia and the south—east, if the showers start to diminish, but not for long. we have more coming receiving and overnight. that low—pressure is just migrating out into the north sea but still some heavy thundery showers around. temperatures fairly academic really overnight. holding into double figures for just about all parts. but it does bring us into a really unsettled picture on sunday. that low—pressure as you can see, drifting out into the north sea. the strongest winds transferring across scotland, into northern ireland. still quite windy in the south but perhaps fewer showers

than today, but still a showery picture and there will be some heavy ones around. but notably so across scotland, coming down through the north channel into northern ireland and the north—west of england, wales as well. again, a very disturbed looking weather picture through the second part of the weekend as well with very similar temperatures to those of today. a5 for next week, early next week, a little ridge of high pressure builds on monday, so perhaps the showers lighter and there will be fewer of them. and then we have got more wet weather to come on tuesday with another area of low pressure. we stay rather cool across our shores but further south, the heat is building again. you can see the oranges and reds, temperatures are above—average and the heat really builds again across western and central europe into next week.

live from london. live from london. this is bbc news. this is bbc news. the belarusian government says the belarusian government says the russian mercenary group wagner is in the country the russian mercenary group and training its troops, three weeks after an attempted mutiny. after an attempted mutiny. the south korean president — yoon suk yeol — is in the ukrainian capital kyiv for a suprise visit. and dozens of cities in southern europe remain on red alert as extremely high temperatures continue to bake the continent. soaring temperatures are impacting countries across southern europe, and are expected to break records next week.