BBC 6 minute English-What can’t computers do

Transcript of the podcast

Note: This is not a word-for-word transcript

Neil: Welcome to 6 Minute English, where we bring you an intelligent topic and six related items of vocabulary. I’m Neil

Tim: And I’m Tim. And today we’re talking about AI – or Artificial Intelligence

Neil: Artificial Intelligence is the ability of machines to copy human intelligent behaviour – for example, an intelligent machine can learn from its own mistakes, and make decisions based on what’s happened in the past

Tim: There’s a lot of talk about AI these days, Neil, but it’s still just science fiction, isn’t it

Neil: That’s not true – AI is everywhere. Machine thinking is in our homes, offices, schools and hospitals. Computer algorithms are helping us drive our cars. They’re diagnosing what’s wrong with us in hospitals. They’re marking student essays… They’re telling us what to read on our smartphones

Tim: Well, that really does sound like science fiction – but it’s happening already, you say, Neil

Neil: It’s definitely happening, Tim. And an algorithm, by the way, is a set of steps a computer follows in order to solve a problem. So can you tell me what was the name of the computer which famously beat world chess champion Garry Kasparov using algorithms in 1997? Was it

a) Hal

b) Alpha 60 or

c) Deep Blue

Tim: I’ll say Deep Blue. Although I’m just guessing

Neil: Was it an educated guess, Tim

Tim: I know a bit about chess

Neil: An educated guess is based on knowledge and experience and is therefore likely to be correct. Well, we’ll find out later on how educated your guess was in this case, Tim

Tim: Indeed. But getting back to AI and what machines can do – are they any good at solving real-life problems? Computers think in zeros and ones don’t they? That sounds like a pretty limited language when it comes to life experience

Neil: You would be surprised to what those zeroes and ones can do, Tim. Although you’re right that AI does have its limitations at the moment. And if something has limitations there’s a limit on what it can do or how good it can be

Tim: OK – well now might be a good time to listen to Zoubin Bharhramani, Professor of Information Engineering at the University of Cambridge and deputy director of the Leverhulme Centre for the Future of Intelligence. He’s talking about what limitations AI has at the moment

INSERT Zoubin Bharhramani, Professor of Information Engineering at the University of Cambridge and deputy director of the Leverhulme Centre for the Future of Intelligence

I think it’s very interesting how many of the things that we take for granted – we humans take for granted – as being sort of things we don’t even think about like how do we walk, how do we reach, how do we recognize our mother. You know, all these things. When you start to think how to implement them on a computer, you realize that it’s those things that are incredibly difficult to get computers to do, and that’s where the current cutting edge of research is

Neil: If we take something for granted we don’t realise how important something is

Tim: You sometimes take me for granted, I think, Neil

Neil: No – I never take you for granted, Tim! You’re far too important for that

Tim: Good to hear! So things we take for granted are doing every day tasks like walking, picking something up, or recognizing somebody. We implement – or perform – these things without thinking – Whereas it’s cutting edge research to try and program a machine to do them

Neil: Cutting edge means very new and advanced. It’s interesting isn’t it, that over ten years ago a computer beat a chess grand master – but the same computer would find it incredibly difficult to pick up a chess piece

Tim: I know. It’s very strange. But now you’ve reminded me that we need the answer to today’s question

Neil: Which was: What was the name of the computer which famously beat world chess champion Garry Kasparov in 1997? Now, you said Deep Blue, Tim, and … that was the right answer

Tim: You see, my educated guess was based on knowledge and experience

Neil: Or maybe you were just lucky. So, the IBM supercomputer Deep Blue played against US world chess champion Garry Kasparov in two chess matches. The first match was played in Philadelphia in 1996 and was won by Kasparov. The second was played in New York City in 1997 and won by Deep Blue. The 1997 match was the first defeat of a reigning world chess champion by a computer under tournament conditions

Tim: Let’s go through the words we learned today. First up was ‘artificial intelligence’ or AI – the ability of machines to copy human intelligent behaviour

Neil: There are AI programs that can write poetry

Tim: Do you have any examples you can recite

Neil: Afraid I don’t! Number two – an algorithm is a set of steps a computer follows in order to solve a problem. For example, Google changes its search algorithm hundreds of times every year

Tim: The adjective is algorithmic – for example, Google has made many algorithmic changes

Neil: Number three – if something has ‘limitations’ – there’s a limit on what it can do or how good it can be. Our show has certain limitations – for example, it’s only six minutes long

Tim: That’s right – there’s only time to present six vocabulary items. Short but sweet

Neil: And very intelligent, too. OK, the next item is ‘take something for granted’ – which is when we don’t realise how important something is

Tim: We take our smart phones for granted these days – but before 1995 hardly anyone owned one

Neil: Number five – ‘to implement’ – means to perform a task, or take action

Tim: Neil implemented some changes to the show

Neil: The final item is ‘cutting edge’ – new and advanced – This software is cutting edge

Tim: The software uses cutting edge technology

Neil: OK – that’s all we have time for on today’s cutting edge show. But please check out our Instagram, Twitter, Facebook and YouTube pages

Tim: Bye-bye

Neil: Goodbye