in the ethiopian region of tigray is now affecting more than 400,000 people. the meeting was also told that the conflict in the region between tigrayan forces and ethiopian federal troops may rapidly deteriorate. president biden says the withdrawal of us troops from afghanistan is on track but won't be completed in the next few days. he was speaking hours after the us left its main military base in the country, bagram, which served as the hub of us—led operations for almost two decades. brazil's prosecutor general has asked for authorisation to investigate president bolsonaro for failing to act on allegations of corruption in the purchase of coronavirus vaccines. a whistle—blower said government officials had agreed to take bribes to buy 400 million doses of the indian—made jab covaxin. now on bbc news, click.

this week—it is a mobiles and music special, we are sending a robot into explore a show floor, right winning song, play the violin and make a pop video. and is that nick rhodes from duran duran? hey, welcome to click. hope you're well. for months now, we've been fantasising about places that we cannot go to at the moment. stories that we cannot, in and the big tech expos that are not happening in the real world, they've moved online. but not this week!

lara, would you care to accompany me to barcelona? absolutely! i've been looking forward to this! now, mobile world congress was one of the first events to be cancelled last year and this year, despite a few delays, it is actually happening. but we're not heading off to sunny spain. well, not really, but we will be there in spirit, 0k? well, not really, but we will be there in spirit, ok? you read ? be there in spirit, ok? you ready? absolutely. - be there in spirit, ok? you ready? absolutely. let's i be there in spirit, 0k? 7m. ready? absolutely. let's do be there in spirit, 0k? 7m, ready? absolutely. let's do it. ok, let's go. we've arrived. this is barcelona and this is megawatts c and we're not here in spirit but we are here in body invisible. my face on it. and the good thing is this robot is connected to the same zoom call as me and you so you can ride along pillion.— can ride along pillion. crikey, ou are can ride along pillion. crikey, you are in — can ride along pillion. crikey, you are in control. _ can ride along pillion. crikey, you are in control. am - can ride along pillion. crikey, you are in control. am i - can ride along pillion. crikey, you are in control. am i in . you are in control. am i in safe hands?— you are in control. am i in safe hands? ~ , ., safe hands? absolutely not! so i can look _ safe hands? absolutely not! so i can look uo — safe hands? absolutely not! so i can look up and _ safe hands? absolutely not! so i can look up and down - safe hands? absolutely not! so i can look up and down like - i can look up and down like

that. and i can go left and right. and forward and backwards with the arrow keys. here we go, look, it's working. i left, they are saying to go left. �* , ~ i left, they are saying to go left. 3 ~ ,., ., left. it's like when someone tries to help _ left. it's like when someone tries to help you _ left. it's like when someone tries to help you park. - left. it's like when someone tries to help you park. so i left. it's like when someone tries to help you park. so if| tries to help you park. so if ever the — tries to help you park. so if ever the robots _ tries to help you park. so if ever the robots do - tries to help you park. so if ever the robots do try - tries to help you park. so if ever the robots do try to i tries to help you park. so if l ever the robots do try to take over, we'rejust ever the robots do try to take over, we're just going to ever the robots do try to take over, we'rejust going to be ever the robots do try to take over, we're just going to be a pushover because we will all be filming it on our phones rather than running up the way. so true. than running up the way. so true- the — than running up the way. sr true. the good thing is do you see down the bottom of the screen you get a birds eye view.., nice! fix, screen you get a birds eye view.., nice!— screen you get a birds eye view.., nice! view. . , nice! a birds eye view ofthe view. . , nice! a birds eye view of the bottom _ view. . , nice! a birds eye view of the bottom of _ view. . , nice! a birds eye view of the bottom of the - view. . , nice! a birds eye view of the bottom of the robots . view. . , nice! a birds eye view| of the bottom of the robots so hopefully we won't run over someone's feet. it does feel great to be back at a tech show, even if it is inside a tablet stuck to the top of a segway, and i guess this is what we are now calling a hybrid show, with some exhibitors and attendees in barcelona and some online. and we will bring you a couple of the stories from mwc in a few minute's time. it looks nice, doesn't it, mwc looks nice. it doesn't it, mwc looks nice. it does look nice, look at those actual people there, actually

there. oh, dear, we have stalled. connecting to mwc is frozen but you are moving normally. frozen but you are moving normally-— normally. actually if you remember _ normally. actually if you remember the _ normally. actually if you remember the times - normally. actually if you remember the times we | normally. actually if you - remember the times we have normally. actually if you _ remember the times we have been to mwc and have struggled to get a mobile phone signal, as you say, and look at what we're trying to do now. all right these people, these people had better up the way. i'm sure i have got right—of—way. fire better up the way. i'm sure i have got right-of-way. are they oblivious to _ have got right-of-way. are they oblivious to us? _ have got right-of-way. are they oblivious to us? i— have got right-of-way. are they oblivious to us? i have - have got right-of-way. are they oblivious to us? i have gone - have got right-of-way. are they oblivious to us? i have gone up| oblivious to us? i have gone up to the bar- _ oblivious to us? i have gone up to the bar. can _ oblivious to us? i have gone up to the bar. can i _ oblivious to us? i have gone up to the bar. can i have - oblivious to us? i have gone up to the bar. can i have an - to the bar. can i have an orangejuice, please? here we 90, orangejuice, please? here we go, look, she is pouring me an orangejuice! right! now you've got a problem!— got a problem! where do we stick it? it _ got a problem! where do we stick it? it will— got a problem! where do we stick it? it will go _ got a problem! where do we stick it? it will go all- got a problem! where do we stick it? it will go all over i stick it? it will go all over the screen. 0k, there were some other robots i wanted to visit, some little ones and they are here somewhere.— some little ones and they are here somewhere. 0k, shall we go look for them? _ here somewhere. 0k, shall we go look for them? let's _ here somewhere. 0k, shall we go look for them? let's try. - here somewhere. 0k, shall we go look for them? let's try. 0k, - look for them? let's try. 0k, and we are — look for them? let's try. 0k, and we are off. _ look for them? let's try. 0k, and we are off. hello. - look for them? let's try. 0k, and we are off. hello. so - look for them? let's try. 0k, l and we are off. hello. so these are pretty _ and we are off. hello. so these are pretty cute _ and we are off. hello. so these are pretty cute robots - and we are off. hello. so these are pretty cute robots and - and we are off. hello. so these l are pretty cute robots and some playing cards on the table as well. what's going on here, please? 50 well. what's going on here, lease? , ., , please? so this one we can use our robot _

please? so this one we can use our robot to — please? so this one we can use our robot to teach _ please? so this one we can use our robot to teach folding, - our robot to teach folding, steam _ our robot to teach folding, steam and artificial intelligence education from age four to — intelligence education from age four to 16. this one is speaking up to ten different language. so students can play with the — language. so students can play with the robot at the same time they can — with the robot at the same time they can learn.— with the robot at the same time they can learn. launched in may 2020, the — they can learn. launched in may 2020, the bot's _ they can learn. launched in may 2020, the bot's optic— they can learn. launched in may 2020, the bot's optic centres i 2020, the bot's optic centres and algorithms help to get kids coding across subjects like maths and music. apparently they even play football better than me. 50 they even play football better than me. ., ., , ,, than me. so we have four steps, the first one _ than me. so we have four steps, the first one is _ than me. so we have four steps, the first one is the _ than me. so we have four steps, the first one is the activity - the first one is the activity by parts, _ the first one is the activity by parts, the second application, third is microbeads and switch and the last one — microbeads and switch and the last one is _ microbeads and switch and the last one is piper.— last one is piper. there are not many _ last one is piper. there are not many kids _ last one is piper. there are not many kids there - last one is piper. there are not many kids there at - last one is piper. there are | not many kids there at mwc last one is piper. there are - not many kids there at mwc so i wonder what the best reactions from the adults have been. fine from the adults have been. one thin , from the adults have been. one thin, i from the adults have been. one thing. i make — from the adults have been. one thing, i make them _ from the adults have been. one thing, i make them surprised, even _ thing, i make them surprised,

even if— thing, i make them surprised, even if we _ thing, i make them surprised, even if we don't inaudible they can leorn— even if we don't inaudible they can learn to use a robot. my robot also — can learn to use a robot. my robot also did _ can learn to use a robot. ij�*i robot also did not can learn to use a robot. ij�*i: robot also did not need can learn to use a robot. m; robot also did not need a manual but maybe i should have taken a driving test first. meanwhile while i leave spencer to play the part of a robot, or is it a robot playing the part of spencer? 0ther is it a robot playing the part of spencer? other machines are also trying to take some work away from us. like creating music. chris fox has been finding out more. electronic version of— finding out more. electronic version or eurovision - finding out more. electronic. version or eurovision theme version of eurovision theme plays. the ai version of eurovision theme plays. the a! song contest is back for a second year and unlike lusty�*s show which was eurovision — themed, this year anything goes. i eurovision - themed, this year anything goes-— anything goes. i feelthe wires, anything goes. i feelthe wires. i _ anything goes. ifeelthe wires, ifeel— anything goes. i feelthe wires, i feel the - anything goes. i feelthe wires, i feelthe wires. l anything goes. i feel the i wires, ifeelthe wires. my wires, ifeel the wires. my blood flows into my heart breeds. ~' ., , ., breeds. like eurovision, the coal breeds. like eurovision, the oal is breeds. like eurovision, the goal is to — breeds. like eurovision, the goal is to create _ breeds. like eurovision, the goal is to create a _ breeds. like eurovision, the goal is to create a song - breeds. like eurovision, the goal is to create a song thatj goal is to create a song that will impress the public and a jury will impress the public and a jury of experts.

will impress the public and a jury of exoerts-_ jury of experts. like a bird fl in jury of experts. like a bird flying in — jury of experts. like a bird flying in the _ jury of experts. like a bird flying in the sky. - jury of experts. like a bird flying in the sky. but - jury of experts. like a bird flying in the sky. but each | jury of experts. like a bird i flying in the sky. but each of the 38 entries _ flying in the sky. but each of the 38 entries had _ flying in the sky. but each of the 38 entries had to - flying in the sky. but each of the 38 entries had to use - flying in the sky. but each of i the 38 entries had to use some artificial intelligence in the process, it was to create lyrics or the melody or even virtual instruments.- lyrics or the melody or even virtual instruments. can't you tell if i could _ virtual instruments. can't you tell if i could fly _ virtual instruments. can't you tell if i could fly i'd _ virtual instruments. can't you tell if i could fly i'd take - tell if i could fly i'd take you high... tell if i could fly i'd take you high---_ tell if i could fly i'd take you high... tell if i could fly i'd take ouhih... ., _ tell if i could fly i'd take ou hih... . ,, . you high... that song was by a rou you high... that song was by a group called — you high... that song was by a group called smorgasbord - you high... that song was by a group called smorgasbord and| you high... that song was by a. group called smorgasbord and if you recognise the pixelated faith that is because it is click�*s very own lj rich who has enter the contest. how do you even get started with creating a song using ai? when i worked alone _ creating a song using ai? when i worked alone the _ creating a song using ai? when i worked alone the music - creating a song using ai? when | i worked alone the music always comes first, the melody, the courts come first but when i'm working with other people i feel inspired by their lyrics and so with this, it felt very natural to get the ai and so with this, it felt very natural to get the al to generate some lyrics for me, which is what we did. we used dpt-3, it's which is what we did. we used dpt—3, it's incredibly good at generating text. you can do anything from writing a question and it would give you

an answer to writing a few sentences and it is completing their —— gpt—3. sentences and it is completing their -- gpt-3. i sentences and it is completing their —— gpt—3. i wanted to write a dance truck, i don't know why, i got into my head that eurovision requires a dance track and so i put in the following. and i thought it would come up with a sort of proper dance— type thing and it returned walking like a zombie in the dark as a completion. and i thought that's really strange! and unearthly and odd! 0nce strange! and unearthly and odd! once you had your lyrics, how did you turn it into a song? how did you decide if it would be a ballad or a banger? it felt naturally that it was falling into a ballad. a few years ago, i was commissioned to analyse the eurovision song contest so i had a lot of data about what is successful in your revision. if you've got a song that is too happy in a major key it will not do that

well at eurovision. you want something in a minor key. that's going to do a lot better than... and so, i decided actually, let's do this at 80 beats a minute in d minor which is an incredibly popular key and give it the sort of feel of and give it the sort of feel of a wide— reaching slightly unearthly love song. a wide- reaching slightly unearthly love song. and when it came to _ unearthly love song. and when it came to writing _ unearthly love song. and when it came to writing your- unearthly love song. and when | it came to writing your melody, i believe you had a little bit of a helping hand from the chart topper adele? i mean, kind of! she _ chart topper adele? i mean, kind of! she was _ chart topper adele? i mean, kind of! she was the - chart topper adele? i mean, kind of! she was the dataset that i trained and ai kind of! she was the dataset that i trained and al on to work out where the melody should go next. jukebox ai is a programme that you can run which you can train on particular bands or artists so adele, i think, particular bands or artists so adele, ithink, is particular bands or artists so adele, i think, is an incredible song writer. why not train and ai incredible song writer. why not train and al on adele and enter the lyrics in and see if the music that the machine creates

can give me some inspiration as to where to go with the melody? machine generates a few seconds of audio and gives you three or four different choices as to where you wanted to go next so it might play... 0r where you wanted to go next so it might play... or it might go... 0r it might play... or it might go... or it might go... 0ne it might play... or it might go... or it might go... one of thoseis go... or it might go... one of those is definitely wrong, right? but the other two are probably 0kc every three seconds, which would take 12 minutes or so of processing time, i would then say yes, i want option one and set it running again. 0k, want option one and set it running again. ok, i will choose option two for this. and i would slowly build this melody with the machine that felt like a better fit. melody with the machine that felt like a betterfit. i’ll melody with the machine that felt like a better fit.— felt like a better fit. i'll be the one — felt like a better fit. i'll be the one to _ felt like a better fit. i'll be the one to watch - felt like a better fit. i'll be the one to watch you - felt like a better fit. i'll be. the one to watch you shine. good — the one to watch you shine. good iuck— the one to watch you shine. good luck lj and everyone in the competition, public voting has closed but you can listen to all of the songs at al has closed but you can listen to all of the songs at a! song contest .com and the winner will be revealed onjuly six.

hello and welcome to the week in tech. it was the week elon musk announced that his satellite internet service starlink can reach 500,000 customers over the coming year. speaking at the mobile world congress in barcelona, the stakes boss also said customers would soon be able to access the internet anywhere on earth, excluding the north and south carolina —— space x boss. also at the mobile world congress, tcl revealed a pair of smart glasses that operate like a display by projecting videos onto them. the glasses will go onto them. the glasses will go on sale later this year for about £500. facebook�*s standalone newsletter bulletin launched and mit and harvard engineers designed face mask can diagnose covid—19 within about 90 minutes. a flying car

has completed a 35 minute flight between two airports in slovakia. pierre cardin takes just over two minutes to transform from a car into an aircraft. the company says a future preproduction model would be road and our legal —— the air car. and robots are entering the world of pop. boston dynamics's robot dog spot challenged korean boy band to be applicable bts to a dance. athletic performances like dance help to improve the programming of the robot by stressing the mechanical design and testing the algorithms and the software. no rusty screws here! violinists actually playing one string at a time and i thought they were doing cords with the bow, the stick? i should have asked. bow? cords with the bow, the stick? ishould have asked. bow? i'm getting a crash course in how to play the violin here at pompeo university in barcelona.

researchers have created an artificial intelligence system that helps. it's called tell me which stands for technology enhanced learning of musical instrument performance. the tiny finger — instrument performance. the tiny finger is _ instrument performance. the tiny finger is there... - instrument performance. the tiny finger is there... using l tiny finger is there... using trackers — tiny finger is there... using trackers on _ tiny finger is there... using trackers on the _ tiny finger is there... using trackers on the violin - tiny finger is there... using trackers on the violin and i tiny finger is there... using i trackers on the violin and the bow as well as an xbox connect, they can monitor how fast i'm playing and if i'm tilting the boat and the quality of contact with the strings but first, a quick lesson. he! —— hey! hey! so i've had my crash course in how to play the violin or hold the violin and the computer software there is showing me whether my bow is perpendicular to the string. that's helping me keep the bow perfectly perpendicular and stay on that one string so let's try another string. 0k. beautiful music this is not and i apologise.

0nce this is not and i apologise. once you've mastered the movements, the system can tell you whether you are playing nicely oozing artificial intelligence so it's been trained what's —— what nice smooth movement sound like and trained what bad playing sound like that if i play a note here. 0k. like that if i play a note here. ok. i like that if i play a note here. 0k. iwas like that if i play a note here. ok. i was sitting to strings at the same time there. it came up yellow to say that was a bad note to let me try again. i got a green circle because i was playing like a pro. because i was playing like a ro. because i was playing like a pro. ., because i was playing like a ro. ., ,., , pro. so we took four professors of music and — pro. so we took four professors of music and they _ pro. so we took four professors of music and they recorded - pro. so we took four professors of music and they recorded a i of music and they recorded a series of exercises from very easy to very complicated and then we ran algorithms in the computer to understand what makes good playing and then trying to then compare these two beginners.— two beginners. presumably ou've two beginners. presumably you've also _ two beginners. presumably you've also taught - two beginners. presumably you've also taught it - two beginners. presumably you've also taught it what l two beginners. presumably i you've also taught it what bad players sounds like. is that why you have invited me here today? it why you have invited me here toda ? ., , ., why you have invited me here toda ? ., , today? it was not the purpose but actually. _ today? it was not the purpose but actually, we _ today? it was not the purpose but actually, we recorded - today? it was not the purpose but actually, we recorded you | but actually, we recorded you and we will use you!-

but actually, we recorded you and we will use you! using the s stem, and we will use you! using the system. i _ and we will use you! using the system. i was _ and we will use you! using the system, i was able _ and we will use you! using the system, i was able to - and we will use you! using the system, i was able to get - and we will use you! using the system, i was able to get an l system, i was able to get an idea of how well i was playing. it helped me self adjust and play the note little bit more in tune. i cannot say how easy it would be to learn an instrument completely from scratch using software like this but i think i am improving. and i can get real—time feedback on whether i am getting it right or wrong. nailed it. of course raphael turned on the style and showed me up about the point of the system is that it can help players of all skill levels. irate players of all skill levels. we were players of all skill levels. - were targeting professional players, very high level players, very high level players, but the solutions all the technology have developed is also impact did the ability of beginners, so the whole spectrum of players. we always say, this is not to replace a teacher, but it is to complement the teachers. i can see how i _

complement the teachers. i can see how i set — complement the teachers. i can see how i set up _ complement the teachers. i can see how i set up like _ complement the teachers. i can see how i set up like this - see how i set up like this would help someone learn how to play an instrument, but is going to take a few more lessons before violin maestro is a string on my bow. amazing, that was chris fox. but here at mwc, the tech isn'tjust teaching us music, it is using music to learn about you. time to trundle over to musimaps, where thomas is about to do a deep dive into my music tastes. hey spencer! 0ne deep dive into my music tastes. hey spencer! one thing that we do is analyse news and —— moods and emotions in the music, but we can also analyse the peoples personality profile based on the music they are listening to. the music they are listening to, ., ., , the music they are listening to. ., ., , , the music they are listening to. ., .,, to. so what does his system make of that _ to. so what does his system make of that song _ to. so what does his system make of that song by - to. so what does his system| make of that song by marina to. so what does his system - make of that song by marina and the diamonds? musimaps produces an array of graphs, picking up sale, tempo, genre and emotional cues as well. we perceive the song is friendly,

warmhearted, passionate, sensible, happy and also a little bit melodramatic. just like me, then. let's go for something more sombre, paranoid android please. irate something more sombre, paranoid android please.— android please. we have some happiness. _ android please. we have some happiness. it — android please. we have some happiness, it says _ android please. we have some happiness, it says it _ android please. we have some happiness, it says it is - android please. we have some happiness, it says it is a - happiness, it says it is a friendly song but it also says innocent, warmhearted, cool, sentimental.— sentimental. there is some happiness _ sentimental. there is some happiness in _ sentimental. there is some happiness in paranoid - sentimental. there is some - happiness in paranoid android, thatis happiness in paranoid android, that is what your system is saying? that is what your system is sa in ? ., that is what your system is sa in? . ., �* that is what your system is sa in? . .,�* ., saying? yeah, we don't do i rics saying? yeah, we don't do lyrics analysis _ saying? yeah, we don't do lyrics analysis at _ saying? yeah, we don't do lyrics analysis at this - saying? yeah, we don't do| lyrics analysis at this point. so this is more like really looking to the audio signal. i would still call that an optimistic reading. 0k, would still call that an optimistic reading. ok, i would still call that an optimistic reading. 0k, ifeel the moment has come, i am giving you my spotify playlists for you to tell me about my personality. and here are my results. plenty of character analysis, or assassinations, complete with myers briggs style personality types. i am

not sure i am that agreeable though. not sure i am that agreeable thou~h. ., . . not sure i am that agreeable thouh. ., . ., though. you are a challenger, and i feel— though. you are a challenger, and i feel bad. _ though. you are a challenger, and i feel bad. it _ though. you are a challenger, and i feel bad. it is _ though. you are a challenger, and i feel bad. it is my- though. you are a challenger, and i feel bad. it is my job! i and i feel bad. it is my “ob! and numberfi and i feel bad. it is my “ob! and number three * and i feel bad. it is my job! and number three type - and i feel bad. it is my job! and number three type is l and i feel bad. it is my job! l and number three type is the performer. and number three type is the performer-— all. performer. can't see that at all- laugtis _ performer. can't see that at all. laughs. _ performer. can't see that at all. laughs. i— performer. can't see that at all. laughs. i can - performer. can't see that at all. laughs. i can see - performer. can't see that at all. laughs. i can see that! now! all. laughs. i can see that! now i thought _ all. laughs. i can see that! now i thought that - all. laughs. i can see that! | now i thought that streaming services like spotify already provide many mood playlists, so haven't they done this already? for the mood —based playlist today, it is still a lot of manual curators involved, but more than that we are building right now and emotional base recommendation system so that the ai would sense you already, since what you would like to play in, what you would like to hear, we would find the music that fits a certain environment or a certain advertising, and even more so, by knowing what music you are interested do, we could also recommend you other products that you are interested in. 50 products that you are interested in. , ,. interested in. so moody music production. — interested in. so moody music production, great _ interested in. so moody music production, great - _ interested in. so moody music production, great - although l production, great — although when it comes to serving up my

emotions to advertisers? maybe i am a paranoid android. 0f i am a paranoid android. of course, ever since the invention of the music video, music icons haven'tjust had to think about how their music sounds, but also how it looks. they have sold over 100 million records, and their fifth decade of performing, and have shared a stage with many a great act. but duran duran have been trying a new type of collaboration, letting artificial intelligence create the music video for their new song, invisible.— song, invisible. how is a machine _ song, invisible. how is a machine brain _ song, invisible. how is a machine brain going - song, invisible. how is a machine brain going to l song, invisible. how is a - machine brain going to make a video? how is it going to construct it? how is it going to conceive ideas? how can we feed in enough information to get something valid back? and... than i was introduced to

huxley. and... than i was introduced to huxle . , and... than i was introduced to huxle. , , �* and... than i was introduced to huxle. , , ~ ., huxley. huxley is the ai that was trained _ huxley. huxley is the ai that was trained to _ huxley. huxley is the ai that was trained to get _ huxley. huxley is the ai that was trained to get creative. l was trained to get creative. this is where it all starts. the process of making an ai the process of making an a! video involves teaching the ai how a human would think. so first of all, you've got to really take a close look at what is going on inside a human's rain, and that can happen right here. it’s human's rain, and that can happen right here. it's med scanner. — happen right here. it's med scanner, and _ happen right here. it's med scanner, and what - happen right here. it's med scanner, and what it - happen right here. it's med scanner, and what it does l happen right here. it's med| scanner, and what it does is happen right here. it's med i scanner, and what it does is it uses really strong magnets to measure very sensitive signals firing in the top part of the brain, and that is where we think a lot of the information for creativity and things like other tasks are encoded. so if you want to imbue them into a! we need to first measure them in humans and find out what is going on and then take that over to the code.— going on and then take that over to the code. this sort of data was _ over to the code. this sort of data was used _ over to the code. this sort of data was used to _ over to the code. this sort of data was used to teach - over to the code. this sort of. data was used to teach huxley to interpret what it sees in the same way a human brain would. then images of the band,

lyrics and video footage were all put into the system, leaving the ai all put into the system, leaving the al to generate its version of that, creating a collaborative process between the band and the computer. this is what it came outwith, eventually. # you can't hear a word i say... # you can't hear a word i sa ... ., # you can't hear a word i sa ., ., say... so what did nick make of it? thrilled. _ say... so what did nick make of it? thrilled, astonished, - it? thrilled, astonished, amazed. _ it? thrilled, astonished, amazed, stunned - it? thrilled, astonished, amazed, stunned and i it? thrilled, astonished, - amazed, stunned and bewildered! how was it to work with an ai? i quite like collaborating with an ai, doesn't answer back as much for sure, and it doesn't ask very difficult questions. but it does come up with surprising results. ijust love the way that some of the thought processes were working. you didn't understand what on earth that was coming up with sometimes, and it became obsessed with the dogs for several days, but when it was

fed in some information, some keyboards, some imagery of keyboards, some imagery of keyboards, it came back with a zebra. that is my kind of surrealist game. of course, stripes, it makes perfect sense. fora stripes, it makes perfect sense. for a bass guitar it came back with a fish, but actually if you think about it, i suppose, it had been through its dictionary, seabass. ihmd its dictionary, seabass. and then there _ its dictionary, seabass. and then there were _ its dictionary, seabass. and then there were the - its dictionary, seabass. and then there were the band's faces, created by broken down data on their images being totally rebuilt by huxley. it made the images of us, we did it without iphones, just literally lip syncing so that could see and learn the dimensions of ourfaces could see and learn the dimensions of our faces and what all the measurements were as ourjaws moved, and it looked at him, and made its versions of us.— versions of us. but a music video also _ versions of us. but a music video also needs _ versions of us. but a music video also needs to - versions of us. but a music video also needs to tell - versions of us. but a music| video also needs to tell the story with feeling. it video also needs to tell the story with feeling.- story with feeling. it is not 'ust story with feeling. it is not just "here _ story with feeling. it is not just "here is— story with feeling. it is not just "here is a _

story with feeling. it is not just "here is a sentence, l just "here is a sentence, generate an image". if a generated image that was dark or really happy, we would classified as such, so you bring emotion into this thing. we went to the band and we generated each frame from the lyrics itself, so the lyrics with the input for each frame, but the first output was a bit random, and not to call. so we spoke to simon and he wrote the song, and he wrote down the emotions that were going through his mind next to each sentence and we added that in there as well and we really started getting something really interesting. it certainly captured the emotion of the song, but don't you miss being in the actual music video yourself? i being in the actual music video ourself? ~ ., ,., . yourself? i quite like robotic me, i yourself? i quite like robotic me. i could _ yourself? i quite like robotic me, i could send _ yourself? i quite like robotic me, i could send robotic- yourself? i quite like robotic me, i could send robotic mej yourself? i quite like robotic. me, i could send robotic me to do a lot of things, i can tell you. laughs. we will make videos with us and them again, of course. it was just thoroughly appealing to do something different. moving forward, something different. moving forward. is _ something different. moving forward, is it _ something different. moving forward, is it something - something different. moving forward, is it something you| forward, is it something you would want to use again? how do

you perceive the future of music videos and ai, for you? obviously it is a joy to be obviously it is a joy to be first with something, because it is nice to... jump off a bridge and realise you can land somewhere with something magical. # we all become invisible... so sounds like they loved being the first ones to do this, but they won't be doing another one. , ., ., ., one. sounds that way, although mick was very — one. sounds that way, although mick was very into _ one. sounds that way, although mick was very into ai _ one. sounds that way, although mick was very into ai in - mick was very into ai in general, and robots, though he didn't— general, and robots, though he didn't need to try and make this— didn't need to try and make this episode of click.- this episode of click. which has been — this episode of click. which has been challenging - this episode of click. which has been challenging to - this episode of click. which | has been challenging to say this episode of click. which - has been challenging to say the least. trust me, well, we are not the stage yet where driving a robot wirelessly around the crowded trade so is easy or reliable, but we made it, just. and i am objecting... reliable, but we made it, 'ust. and i am objecting. . .fi and i am ob'ecting. .. here we to. and i am objecting... here we go- welcome _ and i am objecting... here we go. welcome back! _ and i am objecting... here we go. welcome back! and - and i am objecting... here we go. welcome back! and as - and i am objecting... here we. go. welcome back! and as ever you can — go. welcome back! and as ever you can keep up with the team throughout the week, find us on youtube. — throughout the week, find us on youtube, instagram, facebook and twitter at bbc click. and don't forget _ and twitter at bbc click. and don't forget you _ and twitter at bbc click. and don't forget you can - and twitter at bbc click. fific don't forget you can see all the entries for the ai

don't forget you can see all the entries for the a! song contest at al the entries for the a! song contest at a! song contest .com. and a special treat this week, we are going to play out with an acoustic version of rj's ai with an acoustic version of rj's a! song. so enjoy this, links for watching and we will see you soon stop # can't you tell if i could fly, i take you high... #i high... # i give you all the things you need to make it through the lonely night... # i'll be your star... well, it's going to be one of these days where the weather might turn like that, and you'll have to run for cover if you're not prepared. big showers and thunder and lightning on the way. not everywhere — in fact, many of us will miss the storms altogether, but where they occur, they could even bring some flash flooding. now, early in the morning, there will be already some heavy rain spreading across the southern half of the uk, moving northwards. to the north of that, across scotland, it will have

been been generally a dry night. a bit fresher here — 11 degrees — but in the south, it's muggy — 16 celsius first thing. now already, that band of rain could bring 30mm or so in a short space of time, so even with that, there could be a bit of flash flooding here and there. but later in the afternoon, that's when we'll see the really heavy showers and thunderstorms developing. and the thinking is somewhere in the south—west of england, into western england, we could see this line of storms, mostly inland. in the extreme case, there could be around 80mm of rain. hence the risk of the flash floods. but the storms will be developing in other parts of the country as well, across the midlands, perhaps the south—east, east anglia and also further north in england, so very hit—or—miss. again, some of us will miss the storms altogether, and it's actually going to be a relatively bright, if not sunny, day. but some of these storms may actually linger into the evening hours as well. how about the other side of the world in the mediterranean? in rome for the football, well, here, of course it's going to be warm and sunny, with temperatures up to 28 celsius. now, here's a look at

sunday's weather forecast. and again, showers on the cards, again spread right across the country. you can see how they develop during the course of the afternoon. thunder and lightning possible as well. but again, i think particularly some coastal areas and maybe down towards the south—west may miss the storms altogether. temperatures a little below par — 19 in london, possibly touching 21 degrees where the sun does pop out for any lengthy period of time. now, into next week, we've got low pressure swinging in off the atlantic. that means strengthening winds across the english channel and possibly the south coast of england as well. and with that also will come a spell of rain, so i think monday and tuesday will be fairly changeable. so here's the outlook for the next few days. i know there's a lot of shower clouds, rain clouds, in the outlook here, but i think at times there will be some sunshine as well, so it's not all bad. enjoy that. bye— bye.

welcome to bbc news. i'm lewis vaughan jones. our top stories: the un security council is told that a famine in the ethiopian region of tigray is now affecting more than 400,000 people, and the situation could get a lot worse. if the parties to the conflict fail to seize this moment, the consequences for the people of ethiopia will be devastating — more fighting, more famine. the last us troops leave bagram, their main military base in afghanistan. the biden administration says it's on course to withdraw completely from the country by the end of august. brazil's prosecutor general wants to investigate president bolsonaro for failing to act on allegations of corruption in the purchase of coronavirus vaccines.

and spain and italy are through to the semi—finals