下载掌阅APP，畅读海量书库
立即打开
《创意代码》这本书阐述了算法是否能做出媲美毕加索和贝多芬那样的伟大创意。
测试中可能遇到的词汇和知识：
outsmart [ˌaʊtˈsmɑːt] v.在智力上超越
sacrosanct [ˈsækrəʊsæŋkt] adj.神圣不可侵犯
synthesis [ˈsɪnθəsɪs] n.合成
阅读马上开始，建议您计算一下阅读整篇文章所用的时间，对照下方的参考值就可以评估出您的英文阅读水平。
As machines outsmart us in ever more domains, we can at least comfort ourselves that one area will remain sacrosanct and uncomputable: human creativity. Or can we?
The comforting argument runs that computers may have already become far smarter than us at crunching big numbers, spotting patterns in messy data, or playing chess or Go. But they will never be able to match the creative genius of a Ludwig van Beethoven, or an Anna Akhmatova, or a Pablo Picasso. If great art is crystallised emotion, then dull, inanimate, electronic computers can never compete with the 100 trillion sparking synapses in a human brain, no matter how powerful they may be.
But in his fascinating exploration of the nature of creativity, Marcus du Sautoy questions many of those assumptions. The Oxford mathematician, who is as adept at explaining complex theories in prose as he is on television, argues that so much of what we consider to be creativity consists of super-smart synthesis rather than the flash of inspiration sparked by falling apples. Fake it until you make it is not just the mantra of self-help gurus but also the basis of computational creativity. “Great things are not done by impulse but by small things brought together,” Vincent van Gogh once said. Computers are very good at that kind of thing.
The Creativity Code opens with a discussion of the three types of creativity defined by the cognitive scientist, Margaret Boden, who likes to refer to computers as “tin cans”. The first is exploratory creativity, which explores the outer edges of the possible. So Bach’s preludes and fugues extended the range of the baroque, opening the way for the classical era of Mozart and Beethoven. Or, in a similar way, Renoir’s reimagining of how we visualised our world led to Monet’s new forms of abstraction. In Boden’s view this type of exploration, which accounts for 97 per cent of creativity, is perfect for computation.
Her second type of creativity is combinatorial, which mashes up different types of conceptual constructs. A classic example is the curvaceous architecture of Zaha Hadid inspired by her love of the Russian painter Kazimir Malevich. There are interesting hints, says du Sautoy, that this sort of creativity, too, might be perfect for the world of AI.
But it is Boden’s third type of transformational creativity that is the most elusive and probably most resistant to algorithmic interpretation. Du Sautoy argues that rare moments in creativity change the game. Think Picasso and Cubism, Schoenberg and atonality, Joyce and Modernism. “These are like phase changes when water suddenly goes from a liquid to a gas,” he says.
'Madame de Belamy', a computer-generated artwork
Even here, though, Du Sautoy suggests that machine learning techniques are becoming so good that, in some fields, it might be possible to attain transformational creativity. It is easy to create a meta-rule in a program that will instruct it to change course. The results may be erratic, but once in a while they may be truly transformational.
Du Sautoy largely debunks the myth of the lone genius. Like the musician Brian Eno, who talks about “scenius”, he believes genius is the product of the collective creativity of a community, more often the result of continuous growth.
'Le Marquis de Belamy'
Or, to quote the writer Joyce Carol Oates: “Creative work, like scientific work, should be greeted as a communal effort — an attempt by an individual to give voice to many voices, an attempt to synthesise and explore and analyse.”
Mathematician that he is, Du Sautoy provides a masterful explanation of the history of the algorithm beginning with Euclid’s 2,000-year-old “recipe” for finding the largest number that divides two numbers, known to school children as the highest common factor. As he shows, this can be surprisingly useful when calculating the size of the largest square tiles to cover your kitchen floor, without cutting any tiles.
We learn that the word “algorithm” is derived from the Latin interpretation of the name of the great 9th-century Persian mathematician Muhammad Al-Khwarizmi, one of the first directors of the House of Wisdom in Baghdad, who invented the language of algebra.
The increasing pervasiveness of algorithms needs to be carefully handled, Du Sautoy concludes. But if we can get that right then our powerful new tools can significantly enhance the human code. There is also the intriguing possibility that machines may themselves one day become “conscious”, allowing them to create paintings, music, novels, and mathematics and (if we can understand them) finally enabling us to feel what it is like to be a machine.
The Creativity Code: How AI is learning to write, paint and think, by Marcus du Sautoy, Fourth Estate, RRP£20, 328 pages
John Thornhill is the FT’s innovation editor
请根据你所读到的文章内容，完成以下自测题目：
A.They are all great musicians.
B.They are all great poets.
C.They are all great painters.
D.They are all creative geniuses.
答案 (1)
A.Bach’s preludes and fugues extended the range of the baroque.
B.The curvaceous architecture of Zaha Hadid is inspired by her love of the Russian painter Kazimir Malevich.
C.Queen's Bohemian Rhapsody combined rock with opera.
D.Picasso and Cubism.
答案 (2)
A.Introducing an interesting idea.
B.Call for action.
C.Report a news event.
D.Tell a story.
答案 (3)
(1) 答案：D They are all creative geniuses.解释：But they will never be able to match the creative genius of a Ludwig van Beethoven, or an Anna Akhmatova, or a Pablo Picasso.
(2) 答案：A Bach’s preludes and fugues extended the range of the baroque.解释：The first is exploratory creativity, which explores the outer edges of the possible. So Bach’s preludes and fugues extended the range of the baroque, opening the way for the classical era of Mozart and Beethoven.
(3) 答案：A Introducing an interesting idea.