drawing up the rules? that's piqued your interest, hasn't it? that's piqued your interest, hasn't it? sport and for a full round—up, from the bbc sport centre, here's mark edwards. hello from the bbc sport centre. we're going to start in the english premier league where chelsea and tottenham hotspur go head to head at stamford bridge, the hosts looking to climb into the european places while spurs still on the hunt for a champions league spot. chelsea out of the blocks the fastest bossing the first half an hour, the pressure telling as trevor chalobah headed the hosts in front after 2a minutes. tottenham growing as the half time whistle approached but it's 1—0 at the break. its the semi finals of the europa league and marseille hosts serie a side atalanta. the conquerors of liverpool in the last round taking the lead at orange velodrome. gianlucca scammacca with the opener afterjust 11 minutes, but marseille equalising

through chancel mbemba nine minutes later. while in the other semi final bundesliga champions bayer leverkusen are up against roma in the italian capital 1—0. lets not forget that leverkusen are on a european record—extending 46 games without a loss. and in the europa conference league, the third tier of competition, greek side olympiacos lead aston villa 2—0 at villa park morrocco international yaoub el kaabi with the only — both goals goal of the game so far, while it's 1—1 in florence between fiorentina and belgium's club brugge. english coach will still has parted ways with ligue i club stade de reims with immediate effect following a meeting between himself, the club president and general manager. still took charge of the club on an interim basis in october 2022

following the departure of oscar garcia. he was named full—time coach and oversaw an unbeaten run of 17 matches to begin his tenure at the club — a liguei record. reims are 11th but have only won two of their last 14 games. the semi—finals of snooker�*s world championship are under way with three qualifiers in the last four for the first time since 1977. stuart bingham is the only former winner left. he is playing jakjones right now. this is the latest from th crucible in sheffield. world number 29 bingham flew out of the blocks, sweeping up the first three frames with the welshman missing chances and the 2015 champion leads 3—0. david gilbert is up against kayren wilson in the other semi, it's the first to 17 frames. 4—1; at the break, 11 reds and ii blacks for wilson but he missed out on a second crucible 147, still taking the frame though to level this one nicely. it resumes on friday. and that's all the sport for now.

you are watching the context. it is time for our our regular weekly segment — ai decoded. welcome to ai decoded, it is that time of the week when we dive deep into the world of artificial intellligence. last week we looked at the advance of ai chatbots and the role they can welcome to ai decoded, it is that time of the week when we dive deep into the world of artificial intellligence. last week we looked at the advance of ai chatbots and the role they can play in disseminating false, misleading or harmful information. tonight were looking at the benefits and risks posed by ai on the battlefield. the leverage that the west will get from its advantage in technology, specifically ai over adversaries that may have greater mass is going to be the future of warfare. we will bring you a fascinating interview tonight of the men building the pentagon's most advance systems, palantir�*s european vice—president louis mosley. but first this from reuters...

that's according to a senior us official this week who has urged china and russia to match declarations made by the united states and others that only humans, and never artificial intelligence, would make decisions on deploying nuclear weapons. and what about on battlefield itself? the atlantic profiles boston dynamics one of the most advanced robotics companies in the world, and makes robots for the us military. we'll show you a clip of the companies latest atlas humanoid which moves like no robot you've ever seen before. stephanie here, our regular contributor and technology author. we are going to talk about al in the military. and a system that palantir has build for the pentagon known as titan. full disclosure you once worked for palantir so help us understand what it does? it's been a while, 2017 and i have not personally worked on the system are seen it. it's important to

clarify that. what i read, was publicly available is really interesting. it is very advanced software. it is gone very big on al. it is integrating software with hardware to come up with almost a full technology system using input from sensors, that's an example of hardware, into other data streams. it's all good to be integrated to give a 360 degrees view of the battlefield. we sent our north america corresponded to dc, it's a meeting of some of the most important figures in national security. there she met palantir�*s vice president. we make ai both corporate and government but in defence the

leverage that the west will get from its advantage in technology specifically ai over adversaries that may have greater mass is going to be the future of warfare. it is in use right now in ukraine. what promising opportunities have you seen there? because we know that ukraine has been something of an experiment as well for some of these technologies. it is absolutely the laboratory. it is the front line. it's where we're seeing the newest and most most interesting technologies get tested. and in fact, the speed at which we're learning is extraordinary. so technologies like drones, for example, and by the way, those are the essential technologies on the front line right now. they are changing every 6 to eight weeks fundamentally. so the design itself of the drone, the obsolescence is measured in weeks and months. and i don't think there's been a iteration like that, an innovation cycle since the second world war. what is the challenge that you face in employing or deploying some of that technology on the battlefield in ukraine?

so, the challenge is russian countermeasures. so jamming technologies, never underestimate the russian capabilities. and they are learning also very, very fast. and that's what drives that cat and mouse game between the ukrainians and the russians. if you look at the situation on the battlefield right now, despite the use of this technology, the battlefield has stalled and that is for a lack of conventional weapons. quite frankly, the ukrainians have run out of bullets. does that show us the limits that this type of technology has? so obviously there are limits. in the end, it will come down to ammunition stocks, number of people. but this is where technology can also add that advantage that we discussed before, because if you are in the ukrainian situation where you're outnumbered, russia is three and a half times the population. it has far bigger manufacturing capacity. you're going to have to make each soldier count a great deal more. each of those bullets, each of those missiles count a lot more. and that is where with the addition

of ai and technology, you can ensure that those all go as far as they need to go. really important. thank you for that interview. what he describes is an arms race. that line that drone technology is increasing and improving every six to eight weeks, it's terrifying. haw to eight weeks, it's terrifying. how do ou to eight weeks, it's terrifying. how do you control _ to eight weeks, it's terrifying. how do you control that? _ to eight weeks, it's terrifying. how do you control that? we don't control that necessarily. were talking about in iteration cycle. so palantir developing its weaponry, pala ntir developing its weaponry, all defence palantir developing its weaponry, all defence companies doing the same, so as russia, china, filling the blank of any country you want on earth. untilwe the blank of any country you want on earth. until we have some sort of international arms regulation for this weaponry it is an open playing field. ~ ., this weaponry it is an open playing field. ~ . , ., , ., ., field. what needs to be ironed out? obviously unintended _ field. what needs to be ironed out? | obviously unintended consequences, how do you disengage a ultimately, who is overseeing that? is it

defence or is there independent oversight? defence or is there independent oversiuht? . , ~ , defence or is there independent oversiuht? . , , ., oversight? that is the key question. the us department _ oversight? that is the key question. the us department of _ oversight? that is the key question. the us department of defence - oversight? that is the key question. the us department of defence has i oversight? that is the key question. | the us department of defence has ai fix. they had their principles that anyone can read. the question is is it the companies marking their own homework, the us department of defence deciding if it's ethical or not or somebody who is independent? at the moment it is not really anyone. right now it's marketing for all parties concerned. everyone can point to policy but there's no enforcement.— point to policy but there's no enforcement. a, . _, point to policy but there's no enforcement. . _, . ~ enforcement. okay. we will come back to that after — enforcement. okay. we will come back to that after the _ enforcement. okay. we will come back to that after the break. _ enforcement. okay. we will come back to that after the break. in _ to that after the break. in the company of this man... humanity will no longer be the only or even the most intelligent species on the planet. and everything will change. and if we don't control them then the future will belong to them, not to us. that is conor leahy, ceo of conjecture, an ai safety company, who is warning that al is on a trajectory that

could quickly overwhelm our ability to control it. and nowhere is that more concerning than in defence and robotics. we will be right back.

welcome back to ai decoded. we are joined now by conor leahy who's the founder and ceo of conjecture, an ai startup working on controlling ai systems and aligning them to human values. i got the right, have i? your not into features. you spoken out quite a bit about the threats coming from al. in fact, a bit about the threats coming from al. infact, you a bit about the threats coming from al. in fact, you were in dabbles ai. in fact, you were in dabbles earlier this year armed with an array of policy solutions to control the advances we're talking about. what did you hear that interview with the chief executive or vice president of palantir? what alarms you about what he was talking about?

it was definitely nothing i didn't expect. what we're seeing now what you've said accurately is an arms race bows on the battlefield and off. an arms race, no winner and there's only one loser and that his humanity. what we're doing here is signing over more and more control, more speed, more speed, more of our systems or decision—making to machines without machines that are not necessarily understand. it is important thing to understand about al systems they are not like normal ai systems they are not like normal software. normal software is like computer code, it's written by humans. writing lines of code of what the computer supposed to do. and the computer does it step—by—step. this is not how ai functions. ai is more like grown. you have huge piles of data and you use massive supercomputers to grow a programme to solve your data. and this programme does not look like code written by a human, it looks like a bunch of numbers, a huge pile

of billions of numbers. if you execute these numbers on your computer they can do amazing things fundamentally, to this day even top scientists don't really know what's inside of the systems. is that the kind of system you want making battlefield decision or civilian decisions?— battlefield decision or civilian decisions? . ~ ., decisions? picking through it, the wa ou decisions? picking through it, the way you describe _ decisions? picking through it, the way you describe that, _ decisions? picking through it, the way you describe that, i've - decisions? picking through it, the way you describe that, i've never| way you describe that, i've never heard it described like that. in real time battle scenarios who is going to go through and pick out the numbers that count? the going to go through and pick out the numbers that count?— numbers that count? the truth is more and more, _ numbers that count? the truth is more and more, nobody. - numbers that count? the truth is more and more, nobody. what's| numbers that count? the truth is - more and more, nobody. what's most likely going to happen and it's already starting to happen with drone programmes and the like is it starts out with systems. it gives advice to the human. more more the human starts becoming a rubber stamp machine. and mormor the system seems to detect good enough just click the button, click the button. the future of life institute as a called artificial escalation where they go through a hypothetical scenario

where could be where countries are unsure of the situation. we have some signal but maybe don't know what it is. you come down to the second, down to the minute you might not have time to bake a human decision. and whoever has more ai decision—making can act faster and not even know what's going on for them in 1983 there was a man named lieutenant colonel, russian soldiers stationed in a bunker outside of moscow and one day while he was surveilling for american nuclear strikes these our missile appear on his radar. he had very clear orders, to escalate up the chain of command in order to trigger a nuclear counterstrike. and therefore ending the world as we know it. but he decided that this is unlikely, to see a single two single missile this must be effective for that he decided, to god not to push the button. we could be a very different

state of mr petrov have push that button. ., �* , ., state of mr petrov have push that button. . �* , ., ., state of mr petrov have push that button. . �*, . ., ., , ., button. that's a great example of what they call _ button. that's a great example of what they call human _ button. that's a great example of what they call human and - button. that's a great example of what they call human and a - button. that's a great example of what they call human and a leap l what they call human and a leap decision—making for that were often reassured _ decision—making for that were often reassured that we could use ai in the battlefield because will always keep a _ the battlefield because will always keep a human and luke, he'll always be a human— keep a human and luke, he'll always be a human that makes the final decision— be a human that makes the final decision to — be a human that makes the final decision to whatever iai recommends. i'd be _ decision to whatever iai recommends. i'd be interested to hear what you think_ i'd be interested to hear what you think about this based on notjust the palantir interview but everything we've been hearing in ukraine, — everything we've been hearing in ukraine, gaza, othertheatres of war~ _ ukraine, gaza, othertheatres of war~ do— ukraine, gaza, othertheatres of war~ do you— ukraine, gaza, othertheatres of war. do you think keeping a human in the loop— war. do you think keeping a human in the loop should give us comfort? fortunately, i don't think that is a practical political reality. it's a race to the bottom. whoever has the least humans in the loop can act the fastest. humans are slow they make mistake, they gets tired. we're ready now have simulated dog fights between fighter jets controlled ready now have simulated dog fights between fighterjets controlled by ai systems that completely destroy even our top gun pilots. even the best pilots cannot hold up with how

fast the systems can react. sooner or later whoever is the first one to dig humans entirely is going to be the one that's can it get the superiority. this is a very tempting thing for people to do. i expect people to do this. i was at da bows and i companies are talked to many enterprises and civilian actors about this. —— dabbles another thing i disappointedly hear a lot is various large countries across the globe is publicly they will say ai is to assist humans, to enable more creativity. at the negotiating table, the sales table the only thing they want to hear is how many people cannot lay off if i your product? people cannot lay off if i your roduct? �* , people cannot lay off if i your roduct? �*, people cannot lay off if i your roduct? �* , ., people cannot lay off if i your roduct? �*, ., ., product? let's expand that thought into this story _ product? let's expand that thought into this story from _ product? let's expand that thought into this story from reuters - product? let's expand that thought into this story from reuters today. | into this story from reuters today. the white house is now pushing for a nonproliferation treaty. specifically when it comes to nuclear weapons. specifically when it comes to nuclearweapons. is specifically when it comes to nuclear weapons. is that usable? i nuclear weapons. is that usable? i think many things are feasible and possible. it is a proliferation

problem. it's a proliferation problem. it's a proliferation problem similar to nuclear and the soviets was a very hard diplomatic challenge. at some point we as humanity have to make a choice. the choices, do we continue this race to the bottom, give more to the machines until one day we're just not in control any more? or do we get together somehow, not in control any more? or do we get togethersomehow, i'm not in control any more? or do we get together somehow, i'm not saying i have a solution. irate get together somehow, i'm not saying i have a solution.— i have a solution. we can't even auree i have a solution. we can't even agree on _ i have a solution. we can't even agree on ukraine _ i have a solution. we can't even agree on ukraine or— i have a solution. we can't even agree on ukraine or matters - i have a solution. we can't even agree on ukraine or matters in. i have a solution. we can't even i agree on ukraine or matters in the middle east. there surely isn't the bandwidth for a conversation such as that on al. ., , bandwidth for a conversation such as that on al. . , ., ., ., that on al. last time, a while ago is in the house _ that on al. last time, a while ago is in the house of _ that on al. last time, a while ago is in the house of lords _ that on al. last time, a while ago is in the house of lords and - is in the house of lords and spoke on this topic. i had similar christians raise. surely the chinese or russians will not be willing to talk, it's impossible. then lord browne stood up and said, we did it with the soviets during the cold war. it was hard, yes it was hard

but were still here. i'm not saying it's easy by any means. we're lucky that at least at the current moment the west is very technologically advanced compared to its competitors on the ai sphere. ai is currently very bottlenecked by a very tangible resource which is computing power. to build powerful ai systems we need very large supercomputers was very big, expensive and are only made by three companies. 50 big, expensive and are only made by three companies.— three companies. so far we talked about missiles, _ three companies. so far we talked about missiles, we _ three companies. so far we talked about missiles, we talked - three companies. so far we talked about missiles, we talked all- about missiles, we talked all fighterjets about missiles, we talked all fighter jets that can about missiles, we talked all fighterjets that can attack one another. what about the soldiers of the future? i want to show you something. this is from a company called boston dynamics, they are the most advanced robotics companies in the world, they make robots for the us military. and this is their latest model.

imagine that, georgia with a weapon. when it comes to human intelligence — the one thing we should all remember is that we are mortal. the intelligence dies with us. but as this article forbes points out to, that robot is going to soak up every piece of human intelligence and will grow exponentially. into what? the answer is we don't know. i don't think that particular robot hopefully be very dangerous. if we look at where things are heading, more advanced algorithms, better intelligence, systems that can process data in ways humans can possibly imagine, computers can think on orders of millions or thousand times faster than humans. imagine if we managed to build an ai

system that is as smart as a smart human or scientist for the there thousand times feet of a human. that means for every day it can do two years of research. two years of thinking, two years of planning per day. we know this is possible given not much more that we already have today for that we're to have systems that can read every book ever written in an afternoon. we arty have systems that can generate thousands of photorealistic pictures just like that. we already have dog fights, systems that can control fighter planes beyond any human capability. this is already possible today. it's only speeding up. find today. it's only speeding up. and seedin: today. it's only speeding up. and speeding no _ today. it's only speeding up. and speeding up with the wars. one of the things— speeding up with the wars. one of the things i'm thinking about is after_ the things i'm thinking about is after the — the things i'm thinking about is after the second world war the allies _ after the second world war the allies had to come together and create _ allies had to come together and create a — allies had to come together and create a new type of war crimes, crimes _ create a new type of war crimes, crimes against humanity in order to bring _ crimes against humanity in order to bring to— crimes against humanity in order to bring tojustice the crimes against humanity in order to bring to justice the axis powers which _ bring to justice the axis powers which were germany and japan to hold them accountable for the wondering if them accountable for the wondering it were _ them accountable for the wondering if were actually going to have to

update — if were actually going to have to update some of our war crimes legislation to take account what's going _ legislation to take account what's going to — legislation to take account what's going to happen when the first machine — going to happen when the first machine kills a human, either intentionally or accidentally. who's reaponsible? _ intentionally or accidentally. who's responsible? that's _ intentionally or accidentally. who's responsible? that's what _ intentionally or accidentally. who's responsible? that's what i - intentionally or accidentally. who's responsible? that's what i was - intentionally or accidentally. who's l responsible? that's what i was good to say. responsible? that's what i was good to sa . ~ ., , responsible? that's what i was good tosa .~ ., , ., to say. who depended on? manufacturer, _ to say. who depended on? manufacturer, the - to say. who depended on? manufacturer, the person l to say. who depended on? | manufacturer, the person or programme day? you and i both know this is— programme day? you and i both know this is recycled code or many authors _ this is recycled code or many authors of— this is recycled code or many authors of code. the sources of the data, _ authors of code. the sources of the data. the _ authors of code. the sources of the data, the government who bought it? this is— data, the government who bought it? this is going to be in absolute nightmare forjurors. can this is going to be in absolute nightmare forjurors.- this is going to be in absolute nightmare forjurors. nightmare for “urors. can we talk about nightmare forjurors. can we talk about oversight. _ the wall streetjournal reports today that the ceos from some of the biggest companies, the like of sam altman willjoin the new artificial intelligence safety board to advise on regulation. should they be on it? it depends what the purpose of this institution really is. if its purposeis institution really is. if its purpose is to help these companies grade their own homework and get good government contracts, it's doing its purpose properly. it's

like putting oil companies on your climate change committee. we laughed at but this is not uncommon. this is quite a common practice. who would know more about climate than the oil execs? they have all the information. it a similar thing here. would you put on your ai safety panel? the people knowing about iai, right? but we've learned such as putting tobacco companies on your cancer research committee or putting oil companies on your climate change committee, they might be a conflict of interest here. i think we terrified everybody watching at home. i'm terrified just listening. all right, let's lighten the mood, shall we? it is not all doom and gloom. in fact there is a new ai music app, previewed by rolling stone magazine a few weeks ago, called udio. and i am going to introduce you to comedian and musician ashley frieze who does a lot of stand up, and a large part

of his set is given over to music. he is also a computer nut. what do you zafar? it's much lower than the defence or for that you put in silly lyrics and a tinge of a silly song. it's a lot more fun than the rest of the stuff you've been talking about.— talking about. you've made us a son:. talking about. you've made us a song. dedicated _ talking about. you've made us a song. dedicated to _ talking about. you've made us a song. dedicated to the - talking about. you've made us a song. dedicated to the bbc. - talking about. you've made us a song. dedicated to the bbc. i. song. dedicated to the bbc. i thought it was on message. let's play it. # oh, the bbc # the bbc — it's owned by you and me # funded by the licence fee # if it went away, i'd get the blues # what about my bbc news? # what about all those programmes i adore?# it's like the director general meets michael bull blade. how long did that take you? it

michael bull blade. how long did that take you?— michael bull blade. how long did that take you? it took me about 25 seconds to — that take you? it took me about 25 seconds to write _ that take you? it took me about 25 seconds to write a _ that take you? it took me about 25 seconds to write a lyric and - that take you? it took me about 25 seconds to write a lyric and then i that take you? it took me about 25 seconds to write a lyric and then a | seconds to write a lyric and then a few presses of the button, tweaking parameters until i got something that sounded worth sharing with someone else. um? that sounded worth sharing with someone else.— that sounded worth sharing with someone else. why would this be better than _ someone else. why would this be better than sitting _ someone else. why would this be better than sitting down - someone else. why would this be better than sitting down with - someone else. why would this be better than sitting down with one | someone else. why would this be l better than sitting down with one of those guitars to create something new? �* , ., �* , those guitars to create something new? �*, ., , those guitars to create something new? �*, ., new? it's not. it's absolutely a -a new? it's not. it's absolutely a party trick _ new? it's not. it's absolutely a party trick in — new? it's not. it's absolutely a party trick in a _ new? it's not. it's absolutely a party trick in a gimmick. - new? it's not. it's absolutely a party trick in a gimmick. but . new? it's not. it's absolutely a party trick in a gimmick. but itj new? it's not. it's absolutely a l party trick in a gimmick. but it is the most astonishing thing is a creator of music. it's the most astonishing thing i've ever played with. like opening a toy box and suddenly there is harry potter magic in there that can realise any silly idea into what feels like a produced songin idea into what feels like a produced song in a few seconds. i get addicted to it. it'sjust song in a few seconds. i get addicted to it. it's just been so much fun making these songs. itruiiiiii addicted to it. it'sjust been so much fun making these songs. will we be hearin: much fun making these songs. will we be hearing from _ much fun making these songs. will we be hearing from the _ much fun making these songs. will we be hearing from the estate _ much fun making these songs. will we be hearing from the estate of - much fun making these songs. will we be hearing from the estate of frank. be hearing from the estate of frank sinatra _ be hearing from the estate of frank sinatra after that wonderful tune? i feel that _ sinatra after that wonderful tune? i feel that mr sinatra will be interested.— feel that mr sinatra will be interested. . ~ interested. that said. are you worried about _ interested. that said. are you worried about the _ interested. that said. are you worried about the rights - interested. that said. are you worried about the rights when interested. that said. are you - worried about the rights when you put this music together? we

worried about the rights when you put this music together?— worried about the rights when you put this music together? i've had to take the assumption _ put this music together? i've had to take the assumption that _ put this music together? i've had to take the assumption that this - put this music together? i've had to take the assumption that this is - take the assumption that this is ethically sourced in its reviewed such a large data set there rather than deliberately impersonating one thing it's doing what we all do, which is mimicking of an archetype rather than a specific band. i don't believe that is of frank sinatra song rewritten, ithink believe that is of frank sinatra song rewritten, i think it's just the average of lots of swing music put together. as somebody who makes songs, it's remarkable that a machine can do it better than i can. i could probably get rid of these guitars, it's better than me. great talkin: to guitars, it's better than me. great talking to you- _ guitars, it's better than me. great talking to you. keep _ guitars, it's better than me. great talking to you. keep up _ guitars, it's better than me. great talking to you. keep up the - talking to you. keep up the good work. will you come back and do one for al decoded? anytime you like. just before we end. soumare who comes on sometime he says ourjobs or organ to change, when i could be creative for that working to be curators for that you could see why that would be the case when you see what is possible. $5 a

that would be the case when you see what is possible.— what is possible. as a researcher and writer _ what is possible. as a researcher and writer and _ what is possible. as a researcher and writer and get _ what is possible. as a researcher and writer and get a _ what is possible. as a researcher and writer and get a challenge i and writer and get a challenge that and writer and get a challenge that and say— and writer and get a challenge that and say there's still something going — and say there's still something going on— and say there's still something going on in my brain from reading, thinking _ going on in my brain from reading, thinking and writing and editing process — thinking and writing and editing process. i've yet to see any software — process. i've yet to see any software software come close to touching — software software come close to touchinu. ., v software software come close to touchinu. . �* , ., software software come close to touchinu. . �*, ., , , touching. that's it from us this week. thanks to stephanie and conor for coming in this week. next week — ai and energy. how are we going to power these immensely sophisticated system with the limited resource we have? join us for that same time next week. hello there. good evening. it was stormy last night across the south, around 4000 lightning strikes recorded, and still the potential for more thunderstorms overnight tonight and throughout the day tomorrow for some. once again, the best of the warmth and the sunshine for western areas of scotland.

overnight tonight, we've still got that waving front across england and wales, lots of cloud here with some outbreaks of rain. watch out for some heavy, thundery showers pushing into lincolnshire and yorkshire by the time we get to tomorrow morning, perhaps. still with that onshore breeze for north sea—facing coasts, a misty start to the day. plenty of cloud here. and it's a frost—free start to the day tomorrow, of course, as well. now, through friday, we're going to see our waving front slowly push its way a little further northwards, introducing cooler—feeling conditions across the far south coast. but to start with, through the morning, there'll be outbreaks of rain. watch out for these thunderstorms tracking further northwards and westwards through the day, into north—west england, perhaps parts of northern ireland as well. it's a bright start to the day here, but it will cloud over through the afternoon. again, still that onshore breeze — cloudier, coolerfor north sea —facing coasts. a huge drop in temperature for east anglia between thursday and friday. highest temperatures, best of the sunshine across western areas of scotland once again, 21 or 22 degrees. now, as we head through saturday, our front continues to push further northwards across scotland and northern ireland,

so this is where we're likely to see more cloud across the north as we head through saturday. watch out for some showers as well. and i wouldn't totally rule out the possibility of some further showers across the south—east of england. perhaps the best of the sunshine will tend to be across northern england. temperatures more or less the seasonal average. and on saturday night into sunday, we start to see this area of low pressure roll in from the south—west, so possibly the focus for the showers, the cloud will tend to be out towards the south and the west, gradually pushing further northwards and eastwards through the day. so, again, more sunshine for northern england, perhaps more showers and cloud across northern scotland into northern ireland. but it's important not to take the graphics too literally at this stage through the bank holiday weekend, because they're still likely to change. it's a very messy picture all in all. what we do know is that there will be some sunshine at times and, of course, the sun fairly strong at this time of year, so it will feel warm in the best of that, but expect to see more wet

weather, showers and longer spells of rain, too. bye— bye. hello, i'm christian fraser. you're watching the context on bbc news. violent protest is not protected. peaceful protest is. it's against the law when violence occurs. destroying property is not a peaceful protest. police moved in in the early hours of the morning. a massive police operation took

place here, and systematically they formed this ring of steel, really, around the encampment, around the campus. to me, this feels like a place we didn't need to end up- being and something that many of us did not need to go _ through last night. joined by the panel tonight — kezia dugdale, former leader of the scottish labour party, and miles taylor, former republican national security official. we will get to them very shortly. first, the latest headlines — a senior un official says the cost of reconstructing war—ravaged gaza could be as high as $40 billion. the assistant secretary—general said the world hadn't dealt with such destruction since the second world war. russia has denied allegations that it's using chemical weapons in its war on ukraine.

the us state department said this week that russian troops