kor_sent
stringlengths 1
5.4k
β | eng_sent
stringlengths 1
17.5k
β | source
stringclasses 4
values | similarity
float64 0.12
1
β | from
stringclasses 5
values | __index_level_0__
float64 0
1.78M
β |
---|---|---|---|---|---|
λ³μμΈ‘μ΄ μ²μμ μ΄κ±Έ λ°κ²¬νμ λ μ΄λ° μ¬μ€μ 무μνκ³ μ΄ μμ μλλ₯Ό λ€μ κΉμμ΅λλ€. | Now, when the hospital saw this originally, they tried to lay turf back over it, ignore it. | IWSLT2017 | null | null | null |
νμ§λ§ μΌλ§ νμ, μ΄ κΈΈμ΄ μ€μν μꡬμ¬νμμ μκ² λμμ£ . λ³μμΈ‘μ νμλ€κ³Ό νμλ₯Ό κ°μ‘κ³ κ·Έ νμ μ΄ κ³³μ λλ‘λ₯Ό λ§λ€μμ΅λλ€. | But after a while, they realized it was an important need they were meeting for their patients, so they paved it. | IWSLT2017 | null | null | null |
μ°λ¦¬κ° ν μΌμ μ΄λ° ν¬λ§κ²½λ‘κ° λλ¬λλ©΄ κ·Έκ±Έ ν¬μ₯νλ κ²μ
λλ€. | And I think our job is often to pave these emerging desire paths. | IWSLT2017 | null | null | null |
λ°λ λΆλΆμ μ¬λ‘λ‘ λ€μ λμκ°λ³΄λ©΄ μ΄ ν¬λ§κ²½λ‘κ° λ μ΄ μ리λ μλμμ΄μ. | If we look back at the one in North London again, that desire path hasn't always been there. | IWSLT2017 | null | null | null |
κ·Έ μ΄μ λ μΆκ΅¬ κ²½κΈ°κ° μλ λ μλ μ¬λλ€μ μ€λ₯Έμͺ½ μλμ μ§νμ² μμμ μμ€λ ν΄λ½ κ²½κΈ°μ₯μΌλ‘ μ΄λνμ΅λλ€. | The reason it sprung up is people were traveling to the mighty Arsenal Football Club stadium on game days, from the Underground station you see on the bottom right. | IWSLT2017 | null | null | null |
κ·Έ ν¬λ§κ²½λ‘κ° λ³΄μ΄μμ£ . | So you see the desire path. | IWSLT2017 | null | null | null |
λͺ λ
μ μΌλ‘ κ±°μ¬λ¬ μ¬λΌκ° κ²½κΈ°μ₯μ μ§κ³ μμ λλ μ΄ ν¬λ§κ²½λ‘λ μμμ΅λλ€. | If we just wind the clock back a few years, when the stadium was being constructed, there is no desire path. | IWSLT2017 | null | null | null |
μ°λ¦¬κ° ν΄μΌ ν μΌμ μ΄λ° ν¬λ§κ²½λ‘κ° λλ¬λλ κ±Έ μ°Ύμλ΄κ³ μ λΉν μμΉμ κ·Έ κΈΈμ λ§λ€μ΄ μ£Όλ κ²μ
λλ€. λκ΅°κ°κ° μ΄κ³³μ ν κ²μ²λΌμ. | So our job is to watch for these desire paths emerging, and, where appropriate, pave them, as someone did here. | IWSLT2017 | null | null | null |
λκ΅°κ° μ¬κΈ° μ₯μ λ¬Όμ λμκ³ μ¬λλ€μ μλλ‘ λμκ°κΈ° μμνμ κ·Έ κΈΈμ ν¬μ₯λλ‘λ‘ λ§λ€μμ£ . | Someone installed a barrier, people started walking across and round the bottom as you see, and they paved it. | IWSLT2017 | null | null | null |
κ·Έλ°λ° μ΄κ²λ λλΌμ΄ κ΅νμ μ€λ€κ³ μκ°ν©λλ€. μ€μ λ‘ μΈμμ λμμμ΄ λ³ννλ€λ κ±°μμ. | But I think this is a wonderful reminder as well, that, actually, the world is in flux. | IWSLT2017 | null | null | null |
κ³μν΄μ λ°λκ³ μμ΄μ. μ΄ μ¬μ§μ μμͺ½μ 보μλ©΄ λ λ€λ₯Έ ν¬λ§κ²½λ‘κ° λ§λ€μ΄μ§κ³ μκ±°λ μ. | It's constantly changing, because if you look at the top of this image, there's another desire path forming. | IWSLT2017 | null | null | null |
μ΄λ€ μΈκ°μ§ μΌνκ° μΌκΉ¨μ μ£Όλ κ²μ μ¬λλ€μ μ€μ μꡬμ λ°λΌ λμμΈν΄μΌ νλ€λ κ²μ
λλ€. | So these three desire paths remind me we need to design for real human needs. | IWSLT2017 | null | null | null |
μλΉμμ μꡬλ₯Ό 곡κ°νλ κ²μ΄ μ¬μ
μ±κ³΅μ κ°μ₯ ν° μ²λκ° λλ€κ³ μκ°ν©λλ€. | I think empathy for what your customers want is probably the biggest leading indicator of business success. | IWSLT2017 | null | null | null |
μ€μ μꡬμ λ°λΌ λμμΈνκ³ λ§μ°°μ μ΅μννλ λμμΈμ νμΈμ. λ§μ°°μ μ΅μννλ λμμΈμ μ 곡νμ§ λͺ»νλ©΄ λκ΅°κ°, λλ‘λ μλΉμκ° κ·Έ μΌμ ν κ²μ΄κΈ° λλ¬Έμ
λλ€. | Design for real needs and design them in low friction, because if you don't offer them in low friction, someone else will, often the customer. | IWSLT2017 | null | null | null |
λ λ²μ§Έλ‘, λμ€μ μ€μ μꡬλ₯Ό νμ
νλ κ°μ₯ μ’μ λ°©λ²μ μλΉμ€λ₯Ό μΌλ¨ μ 곡ν΄λ³΄λ κ²μ
λλ€. | Secondly, often the best way to learn what people really want is to launch your service. | IWSLT2017 | null | null | null |
ν΄λ΅μ΄ μ±
μ머리μμ λμ€λ κ²μ μλκ±°λ μ. | The answer is rarely inside the building. | IWSLT2017 | null | null | null |
λ°μΌλ‘ λκ°μ μ¬λλ€μ΄ μ§μ§ μνλ κ²μ λμΌλ‘ νμΈνμΈμ. | Get out there and see what people really want. | IWSLT2017 | null | null | null |
κ·Έλ¦¬κ³ λ§μ§λ§μΌλ‘, κΈ°μ μ νμ
μ΄μ μΈμμ μμκ°μ λλΌμ΄ λ³νλ₯Ό 보μ
λλ€. | And finally, in part because of technology, the world is incredibly flux at the moment. | IWSLT2017 | null | null | null |
κ³μν΄μ λ³ννκ³ μμ΄μ. | It's changing constantly. | IWSLT2017 | null | null | null |
μ΄λ° ν¬λ§κ²½λ‘λ λμμ΄ λΉ λ₯΄κ² λνλκ³ μμ΅λλ€. | These desire paths are going to spring up faster than ever. | IWSLT2017 | null | null | null |
μ°λ¦¬μ μν μ κ·Έ μ€ μ λΉν κ²μ κ³ λ₯΄κ³ κ·Έ κΈΈμ ν¬μ₯μ μμμ£Όλ κ²μ
λλ€. | Our job is to pick the appropriate ones and pave over them. | IWSLT2017 | null | null | null |
κ²½μ²ν΄ μ£Όμ
μ κ°μ¬ν©λλ€. | Thank you very much. | IWSLT2017 | null | null | null |
μ λ ꡬκΈμμ κΈ°κ³μ§λ₯νμ μ΄λκ³ μμ΅λλ€. λ€λ₯΄κ² νννλ©΄, μ»΄ν¨ν°μ μ₯μΉλ₯Ό 곡νμ μΌλ‘ νλ ¨μμΌ λκ° νλ μΌμ ν μ μκ² ν©λλ€. | So, I lead a team at Google that works on machine intelligence; in other words, the engineering discipline of making computers and devices able to do some of the things that brains do. | IWSLT2017 | null | null | null |
κ·Έλ¦¬κ³ μ΄ μΌμ νλ©΄μ μ ν¬λ μ€μ λμ μ κ²½κ³Όνμ κ΄μ¬μ λκ² λμμ΅λλ€. νΉν κ΄μ¬ μλ λΆλΆμ μ°λ¦¬μ λκ° νλ μΌ μ€μ μμ§ μ»΄ν¨ν°λ³΄λ€ ν¨μ¬ λ°μ΄λ λΆλΆμ λν κ²μ
λλ€. | And this makes us interested in real brains and neuroscience as well, and especially interested in the things that our brains do that are still far superior to the performance of computers. | IWSLT2017 | null | null | null |
μμ¬μ μΌλ‘ μ΄λ° λΆλΆ μ€μ νλλ‘ μΈμμ΄ μΈκΈλΌ μμ΅λλ€. μΈμμ μ‘΄μ¬νλ μ리λ μ΄λ―Έμ§λ₯Ό κ³Όμ μ ν΅ν΄ λ§μμμ κ°λ
νμν€λ κ²μ
λλ€. | Historically, one of those areas has been perception, the process by which things out there in the world -- sounds and images -- can turn into concepts in the mind. | IWSLT2017 | null | null | null |
μ΄κ²μ μ°λ¦¬ λμ νμμ μΈ κΈ°λ₯μ΄κ³ μ»΄ν¨ν°μλ κ½€ μ μ©ν©λλ€. | This is essential for our own brains, and it's also pretty useful on a computer. | IWSLT2017 | null | null | null |
κΈ°κ³ μΈμ μκ³ λ¦¬μ¦μ μλ‘ μ ν¬ νμμ ν μΌμ κ΅¬κΈ ν¬ν μ€μ μ¬λ¦° μ¬μ§μ λκ° μ°νλμ λ°λΌ κ²μμ΄ κ°λ₯νκ² ν κ²μ΄μ£ . | The machine perception algorithms, for example, that our team makes, are what enable your pictures on Google Photos to become searchable, based on what's in them. | IWSLT2017 | null | null | null |
μΈμμ λ°λλ§μ μ°½μμ±μ
λλ€. κ°λ
μ μΈμμ μ‘΄μ¬νλ κ²μΌλ‘ λ°κΎΈλ κ²μ
λλ€. | The flip side of perception is creativity: turning a concept into something out there into the world. | IWSLT2017 | null | null | null |
μ§λ λͺ λ
λμ κΈ°κ³ μΈμμ λν΄ μ ν¬κ° ν΄μ¨ μΌλ€μ λ»λ°μλ κΈ°κ³μ μ°½μλ ₯κ³Ό κΈ°κ³ μμ μ μ°κ²°νμ΅λλ€. | So over the past year, our work on machine perception has also unexpectedly connected with the world of machine creativity and machine art. | IWSLT2017 | null | null | null |
μ λ λ―ΈμΌλμ €λ‘κ° μΈμκ³Ό μ°½μμ± κ°μ μ΄μ€ κ΄κ³λ₯Ό κΏ°λ«μ΄ 보μλ€κ³ μκ°ν©λλ€. | I think Michelangelo had a penetrating insight into to this dual relationship between perception and creativity. | IWSLT2017 | null | null | null |
μ΄κ²μ κ·Έμ μ λͺ
ν μΈμ©κ΅¬μ
λλ€. "λͺ¨λ λλ©μ΄λ κ·Έ μμ μ‘°κ°μμ κ°μ§κ³ μκ³ κ·Έκ²μ λ°κ²¬νλ κ²μ΄ μ‘°κ°κ°μ κ³Όμ
μ΄λ€." | This is a famous quote of his: "Every block of stone has a statue inside of it, and the job of the sculptor is to discover it." | IWSLT2017 | null | null | null |
κ·Έλμ μ λ λ―ΈμΌλμ €λ‘μ μκ°μ μ°λ¦¬λ μΈμνλ κ²μΌλ‘ μ°½μ‘°νκ³ κ·Έ μΈμ μμ²΄κ° μμνλ νμμ΄λ©° μ°½μμ±μ΄λΌ μ¬κΉλλ€. | So I think that what Michelangelo was getting at is that we create by perceiving, and that perception itself is an act of imagination and is the stuff of creativity. | IWSLT2017 | null | null | null |
μκ°νκ³ μΈμνκ³ μμνλ κΈ°κ΄μ λ¬Όλ‘ λμ
λλ€. | The organ that does all the thinking and perceiving and imagining, of course, is the brain. | IWSLT2017 | null | null | null |
κ·Έλ¦¬κ³ μ λ κ°λ΅νκ² λμ λν μ°κ΅¬μ μμ¬μ λν΄μ μ΄μΌκΈ°νκ³ μΆμ΅λλ€. | And I'd like to begin with a brief bit of history about what we know about brains. | IWSLT2017 | null | null | null |
μλνλ©΄ μ¬μ₯μ΄λ μ₯κ³Ό λ¬λ¦¬ λ³΄κΈ°λ§ ν΄μ λμ λν΄ μ΄μΌκΈ°ν κ² μκΈ° λλ¬Έμ
λλ€. κ²μΌλ‘ 보기μ λ§μ΄μ£ . | Because unlike, say, the heart or the intestines, you really can't say very much about a brain by just looking at it, at least with the naked eye. | IWSLT2017 | null | null | null |
μ΄κΈ° ν΄λΆνμλ€μ λλ₯Ό λ³΄κ³ νλ©΄μμ ꡬ쑰μ μ¨κ° κΈ°λ°ν μ΄λ¦μ λΆμμ΅λλ€. ν΄λ§κ°μ΄ λ§μ΄μ£ , λ»μ "μμ μμ°"μ
λλ€. | The early anatomists who looked at brains gave the superficial structures of this thing all kinds of fanciful names, like hippocampus, meaning "little shrimp." | IWSLT2017 | null | null | null |
νμ§λ§ λ¬Όλ‘ μ΄λ° μ΄λ¦λ€μ΄ μ€μ λ‘ λ¬΄μ¨μΌμ νλμ§ λ§ν΄ μ£Όμ§λ μμ΅λλ€. | But of course that sort of thing doesn't tell us very much about what's actually going on inside. | IWSLT2017 | null | null | null |
μ μκ°μ μ΅μ΄λ‘ λμμ λ¬΄μ¨ μΌμ΄ μΌμ΄λλμ§μ λν΄ ν° κ³΅νμ ν μ¬λμ μ€νμΈμ μλν μ κ²½ ν΄λΆνμμΈ μ°ν°μκ³ λΌλͺ¬ μ΄ μΉ΄ν μ
λλ€. 19μΈκΈ°μ νλ―Έκ²½ κ΄μ°°κ³Ό νΉμν μ°©μμ μ΄μ©ν΄ μ νμ μΌλ‘ κ°κ°μ λμΈν¬λ₯Ό μ±μ°κ±°λ λμ λλΉλ₯Ό λ§λ€μ΄ λ΄ νννμ μΈ μ΄ν΄λ₯Ό ν μ μκ² ν μ¬λμ
λλ€. | The first person who, I think, really developed some kind of insight into what was going on in the brain was the great Spanish neuroanatomist, Santiago RamΓ³n y Cajal, in the 19th century, who used microscopy and special stains that could selectively fill in or render in very high contrast the individual cells in the brain, in order to start to understand their morphologies. | IWSLT2017 | null | null | null |
μ΄κ²λ€μ κ·Έκ° μ κ²½ μΈν¬λ‘ λ§λ κ·Έλ¦Όλ€μ
λλ€. 19μΈκΈ°μ λ§μ΄μ£ . | And these are the kinds of drawings that he made of neurons in the 19th century. | IWSLT2017 | null | null | null |
μ΄κ²μ μμ λ κ·Έλ¦Όμ
λλ€. | This is from a bird brain. | IWSLT2017 | null | null | null |
κ·Έλ¦¬κ³ κ΅μ₯ν λ€μν μΈν¬λ₯Ό λ³Ό μ μμ΅λλ€. μ¬μ§μ΄ μΈν¬μ΄λ‘ λ μλ €μ§ μ§ μΌλ§ μ λ λμμ΅λλ€. | And you see this incredible variety of different sorts of cells, even the cellular theory itself was quite new at this point. | IWSLT2017 | null | null | null |
κ·Έλ¦¬κ³ μ΄ κ΅¬μ‘°λ μμ§μλΆλ₯Ό κ°μ§κ³ μλ μΈν¬λ€μ κ°μ§λ μμ£Ό λ©λ¦¬κΉμ§ λ»μ μ μλλ° λΉμ λ§€μ° μλ‘μ μ΅λλ€. | And these structures, these cells that have these arborizations, these branches that can go very, very long distances -- this was very novel at the time. | IWSLT2017 | null | null | null |
μ΄ κ΅¬μ‘°λ μ μ μ μ°μμν΅λλ€. | They're reminiscent, of course, of wires. | IWSLT2017 | null | null | null |
μ μ κ³Ό μ κΈ°μ νλͺ
μ΄ μΌμ΄λλ 19μΈκΈ° μ¬λλ€μ λΉμ°ν κ·Έλ κ² λ³Ό μ μμμ κ²μ
λλ€. | That might have been obvious to some people in the 19th century; the revolutions of wiring and electricity were just getting underway. | IWSLT2017 | null | null | null |
νμ§λ§ μ¬λ¬ κ°μ§ λ©΄μμ μ΄λ° λΌλͺ¬ μ΄ μΉ΄ν μ μ‘°μ§νμ κ·Έλ¦Όμ μ€λλ μλ μ΅κ³ λ‘ μ¬κ²¨μ§λλ€. | But in many ways, these microanatomical drawings of RamΓ³n y Cajal's, like this one, they're still in some ways unsurpassed. | IWSLT2017 | null | null | null |
μ°λ¦¬λ μ§λ ν μΈκΈ° λμ λΌλͺ¬ μ΄ μΉ΄ν μ΄ μμν μΌμ λλ΄λ €κ³ λ
Έλ ₯νκ³ μμ΅λλ€. | We're still more than a century later, trying to finish the job that RamΓ³n y Cajal started. | IWSLT2017 | null | null | null |
μ΄κ²λ€μ λ§μ€νλν¬ μ κ²½κ³Όν μ°κ΅¬μ νλ ₯μλ€μ κΈ°μ΄ λ°μ΄νμ
λλ€. | These are raw data from our collaborators at the Max Planck Institute of Neuroscience. | IWSLT2017 | null | null | null |
κ·Έλ¦¬κ³ μ ν¬ νλ ₯μλ€μ΄ ν κ²μ λμΈν¬μ μμ λΆλΆμ μ‘°λͺ
ν κ² μ
λλ€. | And what our collaborators have done is to image little pieces of brain tissue. | IWSLT2017 | null | null | null |
μ΄ μνμ μ 체 ν¬κΈ°λ λλ΅ 1 μ
λ°© λ°λ¦¬λ―Έν°μ΄κ³ κ²°κ³Όλ¬Όμ μμ£Ό μμ λΆλΆμ λ³΄κ³ κ³μ κ²μ
λλ€. | The entire sample here is about one cubic millimeter in size, and I'm showing you a very, very small piece of it here. | IWSLT2017 | null | null | null |
μΌμͺ½μ μλ λ°λ 1λ―Έν¬λ‘ μ
λλ€. | That bar on the left is about one micron. | IWSLT2017 | null | null | null |
λ³΄κ³ κ³μ ꡬ쑰λ λ―Έν μ½λ리μμ
λλ€. μ΄λ λ°ν
리μλ§νΌ μμ΅λλ€. | The structures you see are mitochondria that are the size of bacteria. | IWSLT2017 | null | null | null |
μ΄κ²μ μμ£Ό μμ μ‘°μ§μΌλ‘ μλ₯Έ μ°μμ μΈ λ¨λ©΄μ
λλ€. | And these are consecutive slices through this very, very tiny block of tissue. | IWSLT2017 | null | null | null |
λΉκ΅λ₯Ό νμλ©΄ 머리카λ½μ νκ· μ§λ¦μ 100 λ―Έν¬λ‘ μ
λλ€. | Just for comparison's sake, the diameter of an average strand of hair is about 100 microns. | IWSLT2017 | null | null | null |
μ ν¬κ° λ³΄κ³ μλ κ²μ λ¨Έλ¦¬μΉ΄λ½ ν κ°λ₯λ³΄λ€ ν¨μ¬ μμ κ²μ
λλ€. | So we're looking at something much, much smaller than a single strand of hair. | IWSLT2017 | null | null | null |
κ·Έλ¦¬κ³ μ΄λ° μ μνλ―Έκ²½μΌλ‘ λλ μΌλ ¨μ μ‘°κ°λ€λ‘ μ κ²½μΈν¬λ₯Ό 3Dλ‘ μ΄λ κ² λ³΅μν μ μμ΅λλ€. | And from these kinds of serial electron microscopy slices, one can start to make reconstructions in 3D of neurons that look like these. | IWSLT2017 | null | null | null |
μ΄κ²μ λΌλͺ¬ μ΄ μΉ΄ν μ λ°©μκ³Ό μ΄λ μ λ κ°μ΅λλ€. | So these are sort of in the same style as RamΓ³n y Cajal. | IWSLT2017 | null | null | null |
μΌλΆ μ κ²½μΈν¬λ§ λΉμΆμμ£ . κ·Έλ μ§ μμΌλ©΄ μ무κ²λ ꡬλΆν μ μμ κ²μ
λλ€. | Only a few neurons lit up, because otherwise we wouldn't be able to see anything here. | IWSLT2017 | null | null | null |
μ¬μ§ κ°λν μ κ²½μΈν¬λΌλ¦¬ μλ‘ μ°κ²°λ κ΅¬μ‘°λ§ λ³΄μΌ κ²μ
λλ€. | It would be so crowded, so full of structure, of wiring all connecting one neuron to another. | IWSLT2017 | null | null | null |
λΌλͺ¬ μ΄ μΉ΄ν μ μλλ₯Ό μμλκ°κ³ κ·Έν μμ λ
λμ λμ μ΄ν΄μ λν μ°κ΅¬λ μμν λ°μ νμ΅λλ€. | So RamΓ³n y Cajal was a little bit ahead of his time, and progress on understanding the brain proceeded slowly over the next few decades. | IWSLT2017 | null | null | null |
κ·Έλ¬λ μ°λ¦¬λ μ κ²½μΈν¬κ° μ κΈ°λ₯Ό μ΄μ©νλ κ²μ μμλκ³ μ 2μ°¨ μΈκ³λμ λ λ°μ ν κΈ°μ λ‘ μ€μ λ‘ μ κ²½μΈν¬μ μ κΈ° μ€νμ ν μ μκ² λκ³ μ κ²½μΈν¬λ₯Ό λ μ΄ν΄ν μ μμμ΅λλ€. | But we knew that neurons used electricity, and by World War II, our technology was advanced enough to start doing real electrical experiments on live neurons to better understand how they worked. | IWSLT2017 | null | null | null |
μ»΄ν¨ν°κ° λ°λͺ
λ κ²λ λ°λ‘ μ΄λμΈλ° λλ₯Ό λͺ¨λΈλ‘ ν μμ΄λμ΄μμ£ . μ¨λ° νλ§μ "μ§λ₯ν κΈ°κ³" λΌκ³ λΆλ μ΅λλ€. μ»΄ν¨ν° 곡νμ μλ²μ§ μ€μ ν λͺ
μ΄μ£ . | This was the very same time when computers were being invented, very much based on the idea of modeling the brain -- of "intelligent machinery," as Alan Turing called it, one of the fathers of computer science. | IWSLT2017 | null | null | null |
μλ λ§₯컬λ‘νμ μν° νΌμΈ λ μ΄λλ λΌλͺ¬ μ΄ μΉ΄ν μ μκ° νΌμ§ κ·Έλ¦Όμ 보μμ΅λλ€. μ§κΈ λ³΄κ³ κ³μ κ·Έλ¦Όλ§μ΄μ£ . | Warren McCulloch and Walter Pitts looked at RamΓ³n y Cajal's drawing of visual cortex, which I'm showing here. | IWSLT2017 | null | null | null |
μ΄κ²μ λμ ν΅ν΄ λ€μ΄μ¨ μ΄λ―Έμ§λ₯Ό μ²λ¦¬νλ νΌμ§μ
λλ€. | This is the cortex that processes imagery that comes from the eye. | IWSLT2017 | null | null | null |
κ·Έλ¦¬κ³ κ·Έλ€μκ² μ΄ κ·Έλ¦Όμ λ§μΉ νλ‘λμ²λΌ 보μμ΅λλ€. | And for them, this looked like a circuit diagram. | IWSLT2017 | null | null | null |
λ§₯컬λ‘νμ νΌμΈ μ νλ‘λμλ λ§μ μΈλΆμ¬νμ΄ μμ§λ§ μ ννμ§λ μμ΅λλ€. | So there are a lot of details in McCulloch and Pitts's circuit diagram that are not quite right. | IWSLT2017 | null | null | null |
νμ§λ§ κΈ°λ³Έ μμ΄λμ΄μΈ μκ° νΌμ§μ μλ¦¬κ° μΌλ ¨μ κ³μ° μμλ₯Ό μ°μμ μΌλ‘ νλμμ λ€μμΌλ‘ μ 보λ₯Ό λκΈ΄λ€λ κ²μ κ·Όλ³Έμ μΌλ‘ λ§μ΅λλ€. | But this basic idea that visual cortex works like a series of computational elements that pass information one to the next in a cascade, is essentially correct. | IWSLT2017 | null | null | null |
μ‘°κΈ λ μ΄μΌκΈ°ν΄ λ³΄κ² μ΅λλ€. μκ° μ 보λ₯Ό μ²λ¦¬νλ λͺ¨λΈμ΄ ν΄μΌ νλ μΌμ λν΄μ λ§μ΄μ£ . | Let's talk for a moment about what a model for processing visual information would need to do. | IWSLT2017 | null | null | null |
μΈμμ΄ κΈ°λ³Έμ μΌλ‘ νλ μΌμ μ΄λ° μ΄λ―Έμ§λ₯Ό λ³΄κ³ μ΄λ κ² λ§νλ κ² μ
λλ€. "μ΄κ²μ μμ
λλ€" μ°λ¦¬μκ²λ λ§€μ° μ¬μ΄ μΌμ
λλ€. | The basic task of perception is to take an image like this one and say, "That's a bird," which is a very simple thing for us to do with our brains. | IWSLT2017 | null | null | null |
νμ§λ§ μ¬λ¬λΆ λͺ¨λκ° μμ
μΌ νλ κ²μ΄ λͺ λ
μ κΉμ§ μ»΄ν¨ν°λ‘λ μ΄λ° κ²μ΄ λΆκ°λ₯νμ΅λλ€. | But you should all understand that for a computer, this was pretty much impossible just a few years ago. | IWSLT2017 | null | null | null |
κ³ μ μ μΈ μ»΄ν¨ν
ν¨λ¬λ€μμ μ΄λ° μΌμ μ½κ² ν μ μλ κ²μ΄ μλλλ€. | The classical computing paradigm is not one in which this task is easy to do. | IWSLT2017 | null | null | null |
κ·Έλμ ν½μ
λ€ κ°μ κ΄κ³μ λ§λ€μ΄μ§ μ΄λ―Έμ§μ "μ"λΌλ λ¨μ΄μ κ΄κ³λ κ·Όλ³Έμ μΌλ‘ μ κ²½μΈν¬λ€μ΄ μλ‘ μ°κ²°λμ΄ μ κ²½λ§μ ꡬμΆνκ³ μλ κ²μ
λλ€. μ κ° κ·Έλ¦° λνμ²λΌμ. | So what's going on between the pixels, between the image of the bird and the word "bird," is essentially a set of neurons connected to each other in a neural network, as I'm diagramming here. | IWSLT2017 | null | null | null |
μ΄ μ κ²½λ§μ μκ°νΌμ§ λ΄λΆμ μλ¬Όνμ μΈ κ²μ΄λ μ€λλ μλ μ°λ¦¬μ κΈ°μ λ‘ μ»΄ν¨ν°λ₯Ό ν΅ν΄ μ κ²½λ§μ 그릴 μ μμ΅λλ€. | This neural network could be biological, inside our visual cortices, or, nowadays, we start to have the capability to model such neural networks on the computer. | IWSLT2017 | null | null | null |
κ·Έλ¦¬κ³ μ΄κ²μ΄ μ€μ λͺ¨λΈμ
λλ€. | And I'll show you what that actually looks like. | IWSLT2017 | null | null | null |
ν½μ
μ΄ μ κ²½μΈν¬μ 첫 λ²μ§Έ μΈ΅μ
λλ€. κ·Έλ¦¬κ³ μ΄κ²μ μ€μ λ‘ λμΌλ‘ 보λ κ³Όμ μΌλ‘ 보면 ν½μ
μ΄ λ§λ§μΈ κ²μ
λλ€. | So the pixels you can think about as a first layer of neurons, and that's, in fact, how it works in the eye -- that's the neurons in the retina. | IWSLT2017 | null | null | null |
κ·Έλ¦¬κ³ μ΄ μκ·Ήμ μ κ²½μΈν¬μ ν μΈ΅μμ λ€μ μΈ΅μΌλ‘ μ λ¬ν©λλ€. μ΄λ κ°κ° λ€λ₯Έ λλμ μλ
μ€λ‘ λͺ¨λ μ°κ²°λμ΄μμ΅λλ€. | And those feed forward into one layer after another layer, after another layer of neurons, all connected by synapses of different weights. | IWSLT2017 | null | null | null |
μ΄ λ€νΈμν¬μ λμμ λͺ¨λ μλ
μ€μ κ°λμ μν΄ κ΅¬λΆλ©λλ€. | The behavior of this network is characterized by the strengths of all of those synapses. | IWSLT2017 | null | null | null |
μ΄κ²μΌλ‘ λ€νΈμν¬ λ΄μμ κ³μ°λλ κ²μ νΉμ§μ§μ΅λλ€. | Those characterize the computational properties of this network. | IWSLT2017 | null | null | null |
κ·Έλ¦¬κ³ λ§μ§λ§μ μ κ²½ μΈν¬ νλ λλ ν λ¬΄λ¦¬κ° λ°μ§μ΄λ©° "μ"λΌκ³ λ§ν©λλ€. | And at the end of the day, you have a neuron or a small group of neurons that light up, saying, "bird." | IWSLT2017 | null | null | null |
μ΄μ μ κ° μ΄ μΈκ°μ§λ₯Ό μ
λ ₯λ ν½μ
, μ κ²½λ§μ μλ΅μ€ κ·Έλ¦¬κ³ κ²°κ³Όλ¬ΌμΈ μλ₯Ό μΈ λ³μ x, w, yλΌκ³ νκ² μ΅λλ€. | Now I'm going to represent those three things -- the input pixels and the synapses in the neural network, and bird, the output -- by three variables: x, w and y. | IWSLT2017 | null | null | null |
ν½μ
μ΄ λ°±λ§ κ°λ μμν
λ xλ μ΄λ―Έμ§μ λ°±λ§ κ°μ ν½μ
μ
λλ€. | There are maybe a million or so x's -- a million pixels in that image. | IWSLT2017 | null | null | null |
κ·Έλ¦¬κ³ wλ μμμ΅ νΉμ μμ‘° κ°κ° μμ΅λλ€. μ΄λ μ κ²½λ§μ λͺ¨λ μλ
μ€μ λλλ₯Ό λ§ν©λλ€. | There are billions or trillions of w's, which represent the weights of all these synapses in the neural network. | IWSLT2017 | null | null | null |
κ·Έλ¦¬κ³ μ μ μμ yκ° μμ΅λλ€. μ κ²½λ§μ κ²°κ³Όλ¬Όλ‘μ¨ λ§μ΄μ£ . | And there's a very small number of y's, of outputs that that network has. | IWSLT2017 | null | null | null |
"Bird"λ λ€ κΈμλΏμ΄μμμ. | "Bird" is only four letters, right? | IWSLT2017 | null | null | null |
κ·Έλ¬λ©΄ μ΄κ²μ κ°λ¨ν 곡μμ΄λΌκ³ ν΄λ΄
μλ€. x "x" w = y. | So let's pretend that this is just a simple formula, x "x" w = y. | IWSLT2017 | null | null | null |
μ λ κ³±νκΈ°λ₯Ό ν°λ°μ΄ν μμ λ£μμ΅λλ€. μ€μ λ‘ μ κΈ°μ μΌμ΄λλ μΌμ λ§€μ° λ³΅μ‘ν μΌλ ¨μ μνμ μΈ κ³Όμ μ΄κΈ° λλ¬Έμ
λλ€. | I'm putting the times in scare quotes because what's really going on there, of course, is a very complicated series of mathematical operations. | IWSLT2017 | null | null | null |
μ΄κ²μ ν 곡μμ
λλ€. | That's one equation. | IWSLT2017 | null | null | null |
μΈ κ°μ λ³μκ° μμ΅λλ€. | There are three variables. | IWSLT2017 | null | null | null |
κ·Έλ¦¬κ³ μ°λ¦¬κ° μκ³ μλ κ²μ΄ ν 곡μμμ λ κ°μ λ³μλ₯Ό μλ©΄ λ¨μ ν κ°λ₯Ό μ μ μλ€λ κ²μ
λλ€. | And we all know that if you have one equation, you can solve one variable by knowing the other two things. | IWSLT2017 | null | null | null |
κ·Έλμ μΆλ‘ ν΄μΌ νλ μμ μ¬μ§μ λ³΄κ³ μλ₯Ό ꡬλΆνλ 곡μμ λ°λ‘ μ΄κ²μ
λλ€. μ΄ κ²½μ°λ yλ μλ €μ§μ§ μκ³ wμ xλ μλ €μ§ κ²½μ°μ΄μ£ | So the problem of inference, that is, figuring out that the picture of a bird is a bird, is this one: it's where y is the unknown and w and x are known. | IWSLT2017 | null | null | null |
μ κ²½λ§κ³Ό ν½μ
μ΄ λ¬΄μμΈμ§λ μκ³ μμ΅λλ€. | You know the neural network, you know the pixels. | IWSLT2017 | null | null | null |
보μλ€μνΌ μ¬μ€ μλμ μΌλ‘ κ°λ¨ν λ¬Έμ μ
λλ€ | As you can see, that's actually a relatively straightforward problem. | IWSLT2017 | null | null | null |
2 κ³±νκΈ° 3μ νλ©΄ λλλ κ±°μ£ | You multiply two times three and you're done. | IWSLT2017 | null | null | null |
μ¬λ¬λΆκ» μ΅κ·Όμ λ§λ μΈκ³΅ μ κ²½λ§μ΄ μ νν μ΄κ²μ νλ κ²μ 보μ¬λλ¦¬κ² μ΅λλ€ | I'll show you an artificial neural network that we've built recently, doing exactly that. | IWSLT2017 | null | null | null |
μ΄κ²μ ν΄λμ νμμ μ€μκ°μΌλ‘ λμκ°λ κ²μ
λλ€. κ·Έλ¦¬κ³ λ¬Όλ‘ ν΄λμ νμμ μ΄λΉ μμμ΅ μμ‘° κ°μ λμμ νλ€λ κ² μ체λ§μΌλ‘λ λλΌμ΄ μΌμ
λλ€ | This is running in real time on a mobile phone, and that's, of course, amazing in its own right, that mobile phones can do so many billions and trillions of operations per second. | IWSLT2017 | null | null | null |
μ¬λ¬λΆμ΄ λ³΄κ³ μλ κ²μ ν΄λμ νκ° λ€λ₯Έ μ μ¬μ§μ λ³΄κ³ βλ€, μ΄κ²μ μμ
λλ€.β νκ³ λλλ κ²μ΄ μλλΌ λ€νΈμν¬ μ λ³΄λ‘ μ’
κΉμ§ λΆλ₯νλ λͺ¨μ΅μ
λλ€. | What you're looking at is a phone looking at one after another picture of a bird, and actually not only saying, "Yes, it's a bird," but identifying the species of bird with a network of this sort. | IWSLT2017 | null | null | null |
μ¬μ§μ 보면 xμ wλ λ°νμ Έ μκ³ yλ λ°νμ§μ§ μμμ΅λλ€. | So in that picture, the x and the w are known, and the y is the unknown. | IWSLT2017 | null | null | null |
μ§κΈ λͺΉμ μ΄λ €μ΄ λΆλΆμ μΌλ²λ¬΄λ¦¬κ³ μ§λκ°κ³ μλλ° κ·Έκ²μ μ°λ¦¬κ° μ΄λ»κ² wλ₯Ό λ°νλμΌλ©° λκ° μ΄λ»κ² κ·Έλ° μΌμ νλ©° | I'm glossing over the very difficult part, of course, which is how on earth do we figure out the w, the brain that can do such a thing? | IWSLT2017 | null | null | null |
μ΄λ»κ² μ΄λ° λͺ¨λΈμ λ°°μΈκΉμ
λλ€. | How would we ever learn such a model? | IWSLT2017 | null | null | null |
wλ₯Ό λ°°μ°κ³ ν΄κ²°νλ κ³Όμ μ κ°λ¨ν 곡μμΌλ‘ λ§λ€μ΄ μ«μλ₯Ό λμ
ν΄λ³΄λ©΄ μ νν μ μ μμ΅λλ€. 6=2 x wλΌκ³ νλ©΄ μλ³μ 2λ‘ λλλ©΄ λλ©λλ€. | So this process of learning, of solving for w, if we were doing this with the simple equation in which we think about these as numbers, we know exactly how to do that: 6 = 2 x w, well, we divide by two and we're done. | IWSLT2017 | null | null | null |
λ¬Έμ μ μ μ΄ μ°μ°μμ | The problem is with this operator. | IWSLT2017 | null | null | null |