Ever used an online poemgenerator, perhaps to get started with an English assignment? They normally have a template and some word lists you can fill in, with a simple algorithm that randomly selects from the word lists to fill out the template. “I wandered lonely as a cloud” might become “I zoomed destitute as a rainbow” or I danced homeless as a tree”. It would all depend on those word lists. Artificial Intelligence and machine learning researchers are aiming to be more creative.
Stanford University, the University of Massachusetts and Google have created works that look like poems, by accident. They were using a machine learning Artificial Intelligence they had previously ‘trained’ on romantic novels to research the creation of captions for images, and how to translate text into different languages. They fed it a start and end sentence, and let the AI fill in the gap. The results made sense though were ‘rather dramatic’: for example
“he was silent for a long moment he was silent for a moment it was quiet for a moment it was dark and cold there was a pause it was my turn”
Is this a real poem? What makes a poem a poem is in itself an area of research, with some saying that to create a poem, you need a poet and the poet should do certain things in their ‘creative act’. Researchers from Imperial College London and University College Dublin used this idea to evaluate their own poetry system. They checked to see if the poems they generated met the requirements of a special model for comparing creative systems. This involved things like checking whether the work formed a concept, and including measures such as flamboyance and lyricism.
Read some poems written by humans and compare them to poems created by online poetry generators. What makes it creativity? Maybe that’s up to you!
See also this article about Christopher Strachey, who came up with the first example of a computer program that could create lines of text (from lists of words) to make up love poems.
Claude Shannon, inventor of the rocket powered Frisbee, gasoline powered pogo stick, a calculator that worked using roman numerals, and discoverer of the fundamental equation of juggling! Oh yeah, and founder of the most important theory underpinning all digital communication: information theory.
Claude Shannon is perhaps one of the most important engineers of the 20th century, but he did it for fun. Though his work changed the world, he was always playing with and designing things, simply because it amused him. Like his contemporary Richard Feynman, he did it for ‘the pleasure of finding things out.’
As a boy, Claude liked to build model planes and radio-controlled boats. He once built a telegraph system to a friend’s house half a mile away, though he got in trouble for using the barbed wires around a nearby pasture. He earned pocket money delivering telegrams and repairing radios.
He went to the University of Michigan, and then worked on his Masters at MIT. While there, he thought that the logic he learned in his maths classes could be applied to the electronic circuits he studied in engineering. This became his Masters thesis, published in 1938. It was described as ‘one of the most important Master’s theses ever written… helped to change digital circuit design from an art to a science.’
Claude Shannon is known for his serious research, but a lot of his work was whimsical. He invented a calculator called THROBAC (Thrifty Roman numerical BACkward looking computer), that performs all its operations in the Roman numeral system. His home was full of mechanical turtles that would wander around, turning at obstacles; a gasoline-powered pogostick and rocket-powered Frisbee; a machine that juggled three balls with two mechanical hands; a machine to solve the Rubik’s cube; and the ‘Ultimate Machine’, which was just a box that when turned on, would make an angry, annoyed sound, reach out a hand and turn itself off. As Claude once explained with a smile, ‘I’ve spent lots of time on totally useless things.’
A lot of the early psychology experiments used to involve getting a mouse to run through a maze to reach some food at the end. By performing these experiments over and over in different ways, they could figure out how a mouse learns. So Claude built a mouse-shaped robot called Theseus. Theseus could search a maze until he solved it, and then use this knowledge to find its way through the maze from any starting point.
Oh, and there’s one other paper of his that needs mentioning. No, not the one on the science of juggling, or even the one describing his ‘mind reading’ machine. In 1948 he published ‘A mathematical theory of communication.’ Quite simply, this changed the world, and changed how we think about information. It laid the groundwork for a lot of important theory used in developing modern cryptography, satellite navigation, mobile phone networks… and the internet.
Undersea telecommunications cables let the world communicate and led to the world spanning Internet. It was all started by the Victorians. Continents were connected, but closer islands were too including the Scilly Isles.
Autumn 1869. There were great celebrations as the 31 mile long telecommunications cable was finally hauled up the shore and into the hut. The Scilly Isles now had a direct cable communication link to the mainland. But would it work? Several tests messages were sent and it was announced that all was fine. The journalists filed their story. The celebrations could begin.
Except it didn’t actually work! The cable wasn’t connected at all. The ship laying the cable had gone off course. Either that or someone’s maths had been shaky. The cable had actually run out 5 miles off the islands. Not wanting to spoil the party, the captain ordered the line to be cut. Then, unknown to the crowd watching, they just dragged the cut off end of the cable up the beach and pretended to do the tests. The Scilly Isles weren’t actually connected to Cornwall until the following year.
Paul Curzon, Queen Mary University of London (from the archive)
What was the first technology for recording music: CDs? Records? 78s, The phonograph? No. Trained songbirds came before all of them.
Composer, musician, engineer and visiting fellow at Goldsmiths University, Sarah Angliss, usually has a robot on stage performing live with her. These robots are not slick high tech cyber-beings, but junk modelled automata. One, named Hugo, sports a spooky ventriloquist dolls head! Sarah builds and programs her robots, herself.
She is also a sound historian, and worked on a Radio 4 documentary, ‘The Bird Fancyer’s Delight‘, uncovering how birds have been used to provide music across the ages. During the 1700’s people trained songbirds to sing human invented tunes in their homes. You could buy special manuals showing how to train your pet bird. By playing young birds a tune over and over again, and in the absence of other birds to put them right, they would adopt that song as their own. Playing the recorder was one way to train them, but special instruments were also invented to do the job automatically.
With the invention of the phonograph, home songbird popularity plummeted but it didn’t completely die out. Blackbirds, thrushes, canaries, budgies, bullfinches and other songbirds have continued to be schooled to learn songs that they would never sing in the wild.
The nurse types in a dose of 100.1 mg [milligrams] of a powerful drug and presses start. It duly injects 1001 mg into the patient without telling the nurse that it didn’t do what it was told. You wouldn’t want to be that patient!
Designing a medical device is difficult. It’s not creating the physical machine that causes problems so much as writing the software that controls everything that that machine does. The software is complex and it has to be right. But what do we mean by “right”? The most obvious thing is that when a nurse sets it to do something, that is exactly what it does.
Getting it right is subtler than that though. It must also be easy to use and not mislead the nurse: the human-computer interface has to be right too. It is the software that allows you to interact with a gadget – what buttons you press to get things done and what feedback you are given. There are some basic principles to follow when designing interfaces. One is that the person using it should always be clearly told what it is doing.
Manufacturers need ways to check their devices meet these principles: to know that they got it right.
It’s not just the manufacturers, though. Regulators have the job of checking that machines that might harm people are ‘right’ before they allow them to be sold. That’s really difficult given the software could be millions of lines long. Worse they only have a short time to give an answer.
Million to one chances are guaranteed to happen.
Problems may only happen once in a million times a device is used. They are virtually impossible to find by having someone try possibilities to see what happens, the traditional way software is checked. Of course, if a million devices are bought, then a million to one chance will happen to someone, somewhere almost immediately!
Paolo Masci at Queen Mary University of London, has come up with a way to help and in doing so found a curious problem. He’s been working with the US regulator for medical devices – the FDA – and developed a way to use maths to find problems. It involves creating a mathematical description of what critical parts of the interface program do. Properties, like the user always knowing what is going on, can then be checked using maths. Paolo tried it out on the code for entering numbers of a real medical device and found some subtle problems. He showed that if you typed in certain numbers, the machine actually treated them as a number ten times bigger. Type in a dose of 100.1 and the machine really did set the dose to be 1001. It ignored the decimal point because on such a large dose it assumed small fractions are irrelevant. However another part of the code allows you to continue typing digits. Worse still the device ignores that decimal point silently. It doesn’t make any attempt to help a nurse notice the change. A busy nurse would need to be extremely vigilant to see the tiny decimal point was missing given the lack of warning.
A useful thing about Paolo’s approach is that it gives you the button presses that lead to the problem. With that you can check other devices very quickly. He found that medical devices from three other manufacturers had exactly the same problem. Different teams had all programmed in the same problem. None had thought that if their code ignored a decimal point, it ought to warn the nurse about it rather than create a number ten times bigger. It turns out that different programmers are likely to think the same way and so make the same mistakes (see ‘Double or Nothing‘).
Now the problem is known, nurses can be warned to be extra careful and the manufacturers can update the software. Better still they and the regulators now have an easy way to check their programmers haven’t made the same mistake in future devices. In future, whether vigilant or not, a nurse won’t be able to get it wrong.
Ariane 5 on the launch pad. Photo Credit: (NASA/Chris Gunn) Public Domain via Wikimedia Commons.
If you spent billions of dollars on a gadget you’d probably like it to last more than a minute before it blows up. That’s what happened to a European Space Agency rocket. How do you make sure the worst doesn’t happen to you? How do you make machines reliable?
A powerful way to improve reliability is to use redundancy: double things up. A plane with four engines can keep flying if one fails. Worried about a flat tyre? You carry a spare in the boot. These situations are about making physical parts reliable. Most machines are a combination of hardware and software though. What about software redundancy?
You can have spare copies of software too. Rather than a single version of a program you can have several copies running on different machines. If one program goes wrong another can take over. It would be nice if it was that simple, but software is different to hardware. Two identical programs will fail in the same way at the same time: they are both following the same instructions so if one goes wrong the other will too. That was vividly shown by the maiden flight of the Ariane 5 rocket. Less than 40 seconds from launch things went wrong. The problem was to do with a big number that needed 64 bits of storage space to hold it. The program’s instructions moved it to a storage place with only 16 bits. With not enough space, the number was mangled to fit. That led to calculations by its guidance system going wrong. The rocket veered off course and exploded. The program was duplicated, but both versions were the same so both agreed on the same wrong answers. Seven billion dollars went up in smoke.
Can you get round this? One solution is to get different teams to write programs to do the same thing. The separate teams may make mistakes but surely they won’t all get the same thing wrong! Run them on different machines and let them vote on what to do. Then as long as more than half agree on the right answer the system as a whole will do the right thing. That’s the theory anyway. Unfortunately in practice it doesn’t always work. Nancy Leveson, an expert in software safety from MIT, ran an experiment where different programmers were given programs to write. She found they wrote code that gave the same wrong answers. Even if it had used independently written redundant code it’s still possible Ariane 5 would have exploded.
Redundancy is a big help but it can’t guarantee software works correctly. When designing systems to be highly reliable you have to assume things will still go wrong. You must still have ways to check for problems and to deal with them so that a mistake (whether by human or machine) won’t turn into a disaster.
The Ancient Egyptians, Romans and Greeks used dice with various shapes and markings; some even believed they could be used to predict the future. Using just a few invisible dice, which you can easily make at home, you can amaze your friends with a transparent feat of magical prediction.
The presentation
You can’t really predict the future with dice, but you can do some clever magic tricks with them. For this trick first you need some invisible dice, they are easy to make, it’s all in the imagination. You take your empty hand and state to your friend that it contains two invisible dice. Of course it doesn’t, but that’s where the performance come in. You set up the story of ancient ways to predict the future. You can have lots of fun as you hand the ‘dice’ over and get your friend to do some test rolls to check the dice aren’t loaded. On the test rolls ask them what numbers the dice are showing (remember a dice can only show numbers 1 through 6), this gets them used to things. Then on the final throw, tell them to decide what numbers are showing, but not to tell you! You are going to play a game where you use these numbers to create a large ‘mystical’ number.
To start, they choose one of the dice and move it closer to them, remembering the number on this die. You may want to have them whisper the numbers to another friend in case they forget, as that sort of ruins the trick ending!
Next you take two more ‘invisible dice’ from your pocket; these will be your dice. You roll them a bit, giving random answers and then finally say that they have come up as a 5 and a 5. Push one of the 5s next to the dice your friend selected, and tell them to secretly add these numbers together, i.e. their number plus 5. Then push your second 5 over and suggest, to make it even harder, to multiply their current number by 5+5 (i.e. 10 – that’s a nice easy multiplication to do) and remember that new number. Then finally turn attention to your friend’s remaining unused die, and get them to add that last number to give a grand total. Ask them now to tell you that grand total. Almost instantly you can predict exactly the unspoken numbers on each of their two invisible dice. If they ask how it you did it, say it was easy – they left the dice in plain sight on the table. You just needed to look at them.
The computing behind
This trick works by hiding some simple algebra in the presentation. You have no idea what two numbers your friend has chosen, but let’s call the number on the die they select A and the other number B. If we call the running total X then as the trick progresses the following happens: to begin with X=0, but then we add 5 to their secret number A, so X= A+5. We then get the volunteer to multiply this total by 5+5 (i.e. 10) so now X=10*(A+5). Then we finally add the second secret number B to give X=10(A+5)+B. If we expand this out, X= 10A+50+B. We know that A and B will be in the range 1-6 so this means that when your friend announces the grand total all you need to do is subtract 50 from that number. The number left (10*A+B) means that the value in the 10s column is the number A and the units column is B, and we can announce these out loud. For example if A=2 and B=4, we have the grand total as 10(2+5)+4 = 74, and 74 – 50= is 24, so A is 2, and B is 4.
In what are called procedural computer languages this idea of having a running total that changes as we go through well-defined steps in a computer program is a key element. The running total X is called a variable, to start in the trick, as in a program, we need to initialise this variable, that is we need to know what it is right at the start, in this case X=0. At each stage of the trick (program) we do something to change the ‘state’ of this variable X, ie there are rules to decide what it changes to and when, like adding 5 to the first secret number changes X from 0 to X=(A+5). A here isn’t a variable because your friend knows exactly what it is, A is 2 in the example above, and it won’t change at any time during the trick so it’s called a constant (even if we as the magician don’t know what that constant is). When the final value of the variable X is announced, we can use the algebra of the trick to recover the two constants A and B.
Other ways to do the trick
Of course there are other ways you could perform the trick using different ways to combine the numbers, as long as you end up with A being multiplied by 10 and B just being added. But you want to hide that fact as much as possible. For example you could use three ‘invisible dice’ yourself showing 5, 2 and 5 and go for 5*(A*2+5) + B if you feel confident your friend can quickly multiply by 5. Then you just need to subtract 25 from their grand total (10A+25+B), and you have their numbers. The secret here is to play with the presentation to get one that suits you and your audience, while not putting too much of a mental strain on you or your friend to have to do difficult maths in their head as they calculate the state changes of that ever-growing variable X.
How can you tell if someone looks trustworthy? Could it have anything to do with their facial expression? Some new research suggests that people are less likely to trust someone if their smile looks fake. Of course, that seems like common sense – you’d never think to yourself ‘wow, what a phoney’ and then decide to trust someone anyway. But we’re talking about very subtle clues here. The kind of thing that might only produce a bit of a gut feeling, or you might never be conscious of at all.
To do this experiment, researchers at Cardiff University told volunteers to pick someone to play a trust game with. The scientists told the volunteers to make their choice based on a short video of each person smiling – but they didn’t know the scientists could control certain aspects of each smile, and could make some smiles look more genuine than others.
One of the ways that computers could be more like humans – and maybe pass the Turing test – is by responding to emotion. But how could a computer learn to read human emotions out of words? Matthew Purver of Queen Mary University of London tells us how.
Have you ever thought about why you add emoticons to your text messages – symbols like 🙂 and :-@? Why do we do this with some messages but not with others? And why do we use different words, symbols and abbreviations in texts, Twitter messages, Facebook status updates and formal writing?
In face-to-face conversation, we get a lot of information from the way someone sounds, their facial expressions, and their gestures. In particular, this is the way we convey much of our emotional information – how happy or annoyed we’re feeling about what we’re saying. But when we’re sending a written message, these audio-visual cues are lost – so we have to think of other ways to convey the same information. The ways we choose to do this depend on the space we have available, and on what we think other people will understand. If we’re writing a book or an article, with lots of space and time available, we can use extra words to fully describe our point of view. But if we’re writing an SMS message when we’re short of time and the phone keypad takes time to use, or if we’re writing on Twitter and only have 140 characters of space, then we need to think of other conventions. Humans are very good at this – we can invent and understand new symbols, words or abbreviations quite easily. If you hadn’t seen the 😀 symbol before, you can probably guess what it means – especially if you know something about the person texting you, and what you’re talking about.
But computers are terrible at this. They’re generally bad at guessing new things, and they’re bad at understanding the way we naturally express ourselves. So if computers need to understand what people are writing to each other in short messages like on Twitter or Facebook, we have a problem. But this is something researchers would really like to do: for example, researchers in France, Germany and Ireland have all found that Twitter opinions can help predict election results, sometimes better than standard exit polls – and if we could accurately understand whether people are feeling happy or angry about a candidate when they tweet about them, we’d have a powerful tool for understanding popular opinion. Similarly we could automatically find out whether people liked a new product when it was launched; and some research even suggests you could even predict the stock market. But how do we teach computers to understand emotional content, and learn to adapt to the new ways we express it?
One answer might be in a class of techniques called semi-supervised learning. By taking some example messages in which the authors have made the emotional content very clear (using emoticons, or specific conventions like Twitter’s #fail or abbreviations like LOL), we can give ourselves a foundation to build on. A computer can learn the words and phrases that seem to be associated with these clear emotions, so it understands this limited set of messages. Then, by allowing it to find new data with the same words and phrases, it can learn new examples for itself. Eventually, it can learn new symbols or phrases if it sees them together with emotional patterns it already knows enough times to be confident, and then we’re on our way towards an emotionally aware computer. However, we’re still a fair way off getting it right all the time, every time.
In school we learn about the maths that others have invented: results that great mathematicians like Euclid, Pythagoras, Newton or Leibniz worked out. We follow algorithms for getting results they devised. Ada Lovelace was actually taught by one of the great mathematicians, Augustus De Morgan, who invented important laws, ‘De Morgan’s laws’ that are a fundamental basis for the logical reasoning computer scientists now use. Real maths is about discovering new results of course not just using old ones, and the way that is done is changing.
We tend to think of maths as something done by individual geniuses: an isolated creative activity, to produce a proof that other mathematicians then check. Perhaps the greatest such feat of recent years was Andrew WIles’ proof of Fermat’s Last Theorem. It was a proof that had evaded the best mathematicians for hundreds of years. Wiles locked himself away for 7 years to finally come up with a proof. Mathematics is now at a remarkable turning point. Computer science is changing the way maths is done. New technology is radically extending the power and limits of individuals. “Crowdsourcing” pulls together diverse experts to solve problems; computers that manipulate symbols can tackle huge routine calculations; and computers, using programs designed to verify hardware, check proofs that are just too long and complicated for any human to understand. Yet these techniques are currently used in stand-alone fashion, lacking integration with each other or with human creativity or fallibility.
‘Social machines’ are a whole new paradigm for viewing a combination of people and computers as a single problem-solving entity. The idea was identified by Tim Berners-Lee, inventor of the world-wide web. A project led by Ursula Martin at the University of Oxford explored how to make this a reality, creating a mathematics social machine – a combination of people, computers, and archives to create and apply mathematics. The idea is to change the way people do mathematics, so transforming the reach, pace, and impact of mathematics research. The first step involves social science rather than maths or computing though – studying what working mathematicians really do when working on new maths, and how they work together when doing crowdsourced maths. Once that is understood it will then be possible to develop tools to help them work as part of such a social machine.
The world changing mathematics results of the future may be made by social machines rather than solo geniuses. Team work, with both humans and computers is the future.
– Ursula Martin, University of Oxford and Paul Curzon, Queen Mary University of London
The history of computational devices: automata, core rope memory (used by NASA in the Moon landings), Charles Babbage’s Analytical Engine (never built) and Difference Engine made of cog wheels and levers, mercury delay lines, standardising the size of machine parts, Mary Coombs and the Lyons tea shop computer, computers made of marbles, i-Ching and binary, Ada Lovelace and music, a computer made of custard, a way of sorting wood samples with index cards and how to work out your own programming origin story.
Subscribe to be notified whenever we publish a new post to the CS4FN blog.
This blog is funded by EPSRC on research agreement EP/W033615/1.
.jpg)
cs4fn 