Information technology grows in an exponential manner. It's not linear. And our intuition is linear. When we walked through the savanna a thousand years ago we made linear predictions where that animal would be, and that worked fine. It's hardwired in our brains. But the pace of exponential growth is really what describes information technologies. And it's not just computation. There is a big difference between linear and exponential growth. If I take 30 steps linearly -- one, two, three, four, five -- I get to 30. If I take 30 steps exponentially -- two, four, eight, 16 -- I get to a billion. It makes a huge difference. And that really describes information technology.
資訊科技正在以指數的幅度發展 它並不是線性的。可是對我們來講,直覺知識卻是線性的 一千年以前,當我們走過熱帶草原 我們直接推斷獵物會在哪邊 這樣的推斷是行得通的。我們已經習慣利用線性的方式來估計 但是指數發展的速度 才能準確地形容目前的資訊科技. 這不僅僅是計算方式的差異. 線性和指數增長有著很大的不同. 假如我直線地走個30步, 1, 2, 3, 4, 5 我到達30. 假如我以指數方式走30步, 2, 4, 8, 16, 我到達10億多. 這相差了十萬八千里. 指數增長確切地描述了資訊科技
When I was a student at MIT, we all shared one computer that took up a whole building. The computer in your cellphone today is a million times cheaper, a million times smaller, a thousand times more powerful. That's a billion-fold increase in capability per dollar that we've actually experienced since I was a student. And we're going to do it again in the next 25 years. Information technology progresses through a series of S-curves where each one is a different paradigm. So people say, "What's going to happen when Moore's Law comes to an end?" Which will happen around 2020. We'll then go to the next paradigm. And Moore's Law was not the first paradigm to bring exponential growth to computing. The exponential growth of computing started decades before Gordon Moore was even born. And it doesn't just apply to computation. It's really any technology where we can measure the underlying information properties.
當年我還在麻省理工學院上學的時候, 我們班上共用的一台電腦就佔掉了整棟樓的能量資源. 現在手機裡面的電腦程式便宜了一百萬倍, 小了一百萬倍, 強大了一百萬倍. 這相當於一美元就有一億倍的增長能力 從我還是個學生至今, 這就是我們所經歷的. 在未來, 這樣的快速發展還會持續25年. 通過一系列的S-曲線 資訊科技將會持續進步 到不同的模式. 所以人們問, "當摩爾定律到達終點, 這世界會變成怎樣?" 當摩爾定律在2020到達終點, 我們會進入下一個發展模式. 但是摩爾定律並不是第一個導致 資訊科技指數發展的思維模式. 資訊科技指數性的進步發生於 戈登.摩爾出生幾十年前 科技的指數發展並不限於電腦科技, 它包含任何一樣 我們所知道到的科技.
Here we have 49 famous computers. I put them in a logarithmic graph. The logarithmic scale hides the scale of the increase, because this represents trillions-fold increase since the 1890 census. In 1950s they were shrinking vacuum tubes, making them smaller and smaller. They finally hit a wall; they couldn't shrink the vacuum tube any more and keep the vacuum. And that was the end of the shrinking of vacuum tubes, but it was not the end of the exponential growth of computing. We went to the fourth paradigm, transistors, and finally integrated circuits. When that comes to an end we'll go to the sixth paradigm; three-dimensional self-organizing molecular circuits.
這裡有49台不同年代的電腦,我用對數線圖做個整理 對數線的大小影藏了真正增長的比率. 但是這圖表描繪了自1890以來 科技億萬倍的增長. 在50年代, 電腦工程師盡可能的縮小真空管, 他們一直改良又改良, 最後到達了極限. 他們不能再縮小真空管,只能保留真空部分 而那就是真空管縮小技術的終點 但那可不是資訊科技指數發展的結局. 我們到了第四個發展模式, 改良電晶體 然後我們又去整合電路. 當上個步驟結束了, 我們將到達第六個發展模式, 開發三維自組織分子電路.
But what's even more amazing, really, than this fantastic scale of progress, is that -- look at how predictable this is. I mean this went through thick and thin, through war and peace, through boom times and recessions. The Great Depression made not a dent in this exponential progression. We'll see the same thing in the economic recession we're having now. At least the exponential growth of information technology capability will continue unabated.
但比這個驚人的進步更難以置信的, 我說真的, 是科技的發展有多麼好預測. 科技的發展經過大跟小, 戰爭跟和平, 繁榮跟衰退. 1930年的經濟大蕭條根本沒影響到科技的指數發展. 在這金融危機裡我們會見識到一樣的結果. 至少資訊科技的指數增長的能力 將不會減弱.
And I just updated these graphs. Because I had them through 2002 in my book, "The Singularity is Near." So we updated them, so I could present it here, to 2007. And I was asked, "Well aren't you nervous? Maybe it kind of didn't stay on this exponential progression." I was a little nervous because maybe the data wouldn't be right, but I've done this now for 30 years, and it has stayed on this exponential progression.
我更新了這些圖 因為在我的書"奇點迫近"(The Singularity is Near), 數據只延伸到2002年, 所以我們更新了資料 讓我才能夠在2007年發表. 很多人問我, "你不緊張嗎? 說不定數據並不證明你所說的指數發展." 我是有點緊張. 害怕數據可能會不合. 可是我做這行30多年了, 數據總是證明科技是朝向指數發展的.
Look at this graph here.You could buy one transistor for a dollar in 1968. You can buy half a billion today, and they are actually better, because they are faster. But look at how predictable this is. And I'd say this knowledge is over-fitting to past data. I've been making these forward-looking predictions for about 30 years. And the cost of a transistor cycle, which is a measure of the price performance of electronics, comes down about every year. That's a 50 percent deflation rate. And it's also true of other examples, like DNA data or brain data. But we more than make up for that. We actually ship more than twice as much of every form of information technology. We've had 18 percent growth in constant dollars in every form of information technology for the last half-century, despite the fact that you can get twice as much of it each year.
看. 在1968年你要花一美元才能買一個電晶體 今天一美元可以買五千萬個電晶體 實際上今天的晶體管更好, 更快. 看科技的發展有多麼好預測. 我會說這資訊是過去式了. 我做了超過30年的前瞻性預測. 電晶體的費用, 相應地呈現了電子的市場價格, 每年都下降. 那說明了百分之五十的下降. 而且它也適用於其他的例子 例如DNA數據或大腦的數據. 但是我們的社會進步的更快. 實際上我們生產一倍以上 一種同樣的科技. 過去半個世紀,不管哪種資訊科技, 衡定價值都有百分之十八的增長 儘管你每年都可以得到一倍以上的回報
This is a completely different example. This is not Moore's Law. The amount of DNA data we've sequenced has doubled every year. The cost has come down by half every year. And this has been a smooth progression since the beginning of the genome project. And halfway through the project, skeptics said, "Well, this is not working out. You're halfway through the genome project and you've finished one percent of the project." But that was really right on schedule. Because if you double one percent seven more times, which is exactly what happened, you get 100 percent. And the project was finished on time.
這是個完全不同的例子. 這不是摩爾定律. 我們所獲得DNA數據的總量 總是增加一倍以上. 而每年費用卻下跌一半. 自從人類基因定序計劃(Human Genome Project), 這已經成為了一個持續的發展定律. 當這計劃進行到一半時, 有人懷疑 "這不會成功的. 已過了一半的計劃時間, 你卻只完成了百分之一的任務." 可是那工程是如期進行. 因為如果你將百分之一乘兩倍,並連乘七次以上 實際上所產生的, 就是百分之百. 如此工程按照時間地完成了.
Communication technologies: 50 different ways to measure this, the number of bits being moved around, the size of the Internet. But this has progressed at an exponential pace. This is deeply democratizing. I wrote, over 20 years ago in "The Age of Intelligent Machines," when the Soviet Union was going strong, that it would be swept away by this growth of decentralized communication.
傳播科技 可用50種不同的方式來評量 正在移動的位元數目, 網路的大小. 但科技正在以指數的步伐進步. 這是強烈地民主化 20年前,我在我的書"誰會代替人類:智能簡史" (The Age of Intelligent Machines) 中寫到, 當蘇聯正強大的時候, 它會被這鼓增長的非主流通訊勢力瓦解
And we will have plenty of computation as we go through the 21st century to do things like simulate regions of the human brain. But where will we get the software? Some critics say, "Oh, well software is stuck in the mud." But we are learning more and more about the human brain. Spatial resolution of brain scanning is doubling every year. The amount of data we're getting about the brain is doubling every year. And we're showing that we can actually turn this data into working models and simulations of brain regions.
當我們經過21世紀, 我們能運用大量電腦科技 來做些事,例如模擬人類大腦區域 但是我們要從哪裡得到這科技? 有寫評論家說, "喔, 科技還沒那麼發達." 事實上, 我們越來越了解人類大腦 每年腦部掃描的空間分辨率都比前年高了一倍. 每年我們所得到有關人類大腦的訊息都增加了一倍. 我們證明,事實上可以轉化這個數據 便成大腦區域的模型和模擬
There is about 20 regions of the brain that have been modeled, simulated and tested: the auditory cortex, regions of the visual cortex; cerebellum, where we do our skill formation; slices of the cerebral cortex, where we do our rational thinking. And all of this has fueled an increase, very smooth and predictable, of productivity. We've gone from 30 dollars to 130 dollars in constant dollars in the value of an average hour of human labor, fueled by this information technology.
目前人類大概建構,模擬並測試了 20個大腦區域: 不同的聽覺和視覺皮層區域, 構成不同能力的小腦, 做理性思考的大腦等. 所有的發現, 以相當平穩可預測的模式,增加了生產力. 因為資訊科技的進步, 我們的工作價值從每小時30元美金 到每小時130元美金.
And we're all concerned about energy and the environment. Well this is a logarithmic graph. This represents a smooth doubling, every two years, of the amount of solar energy we're creating, particularly as we're now applying nanotechnology, a form of information technology, to solar panels. And we're only eight doublings away from it meeting 100 percent of our energy needs. And there is 10 thousand times more sunlight than we need.
這還只是能源和環境的影響. 嗯, 這是一個對數圖. 每兩年, 我們製造的太陽能持續倍增. 特別是我們現在正在運用奈米科技, 一種資訊科技, 在太陽能電池板上. 我們現在只離我們所需要的百分之百能量 八次的雙倍增長. 而太陽能則超過我們一萬多倍的需求.
We ultimately will merge with this technology. It's already very close to us. When I was a student it was across campus, now it's in our pockets. What used to take up a building now fits in our pockets. What now fits in our pockets would fit in a blood cell in 25 years. And we will begin to actually deeply influence our health and our intelligence, as we get closer and closer to this technology.
最後太陽能會和科技結合。時間就快到了。 當我還是個學生, 它在校園的對面. 現在它可以放進我們的口袋裡. 以前用掉整棟大樓資源的現在適合放進我們的口袋裡. 現在放得進我們口袋裡的,25年後將可以放在一個紅血球裡. 當我們越來越接近這科技, 我們會真正開始左右 我們的健康跟智慧.
Based on that we are announcing, here at TED, in true TED tradition, Singularity University. It's a new university that's founded by Peter Diamandis, who is here in the audience, and myself. It's backed by NASA and Google, and other leaders in the high-tech and science community. And our goal was to assemble the leaders, both teachers and students, in these exponentially growing information technologies, and their application. But Larry Page made an impassioned speech at our organizing meeting, saying we should devote this study to actually addressing some of the major challenges facing humanity. And if we did that, then Google would back this. And so that's what we've done.
所以我們要以TED一貫的傳統,, 在TED這裡宣布,我們要設立優越大學. 這是一所全新的大學 由台下的聽眾,彼得‧岱爾莽第斯先生 和我所創立. 它獲得美國太空總署(NASA)和Google的贊助 還有其他在高科技領域的領袖們的支持. 我們的目標是召集領導人, --老師和學生, 來研究這個指數發展的資訊科技 和它的用途. 裴基(Larry Page)先生在我們的會議上 發表了一段熱烈的演講. 他說我們應致力研究於 真正解決一些人類面臨的重大挑戰. 假如我們做了這選擇, Google會資助我們. 所以我們做了研究上的一些改變.
The last third of the nine-week intensive summer session will be devoted to a group project to address some major challenge of humanity. Like for example, applying the Internet, which is now ubiquitous, in the rural areas of China or in Africa, to bringing health information to developing areas of the world. And these projects will continue past these sessions, using collaborative interactive communication. All the intellectual property that is created and taught will be online and available, and developed online in a collaborative fashion.
在密集的九週暑期學營裡的最後三週, 我們將會分組專門來提出 一些社會上面臨的重大挑戰. 例如將今天已經很普及的網路, 提供給中國和非洲的鄉村地區, 好將健康資訊 傳播到世界的每個發展地區. 這些科研項目會延展到這些學營外, 通過協作地互動溝通討論. 所有萌生和傳授的智慧財產 將會在網路上公開, 並在網路上互相合作發展.
Here is our founding meeting. But this is being announced today. It will be permanently headquartered in Silicon Valley, at the NASA Ames research center. There are different programs for graduate students, for executives at different companies. The first six tracks here -- artificial intelligence, advanced computing technologies, biotechnology, nanotechnology -- are the different core areas of information technology. Then we are going to apply them to the other areas, like energy, ecology, policy law and ethics, entrepreneurship, so that people can bring these new technologies to the world.
這是我們的創校會議的照片. 今天我們在這裡發佈. 優越大學(Singulariy University)將會永久設置在矽谷, 在NASA的艾密斯研究中心. 我們提供不同的課程給研究生, 和不同公司的高階主管. 這裡的六種首要研究方向, 人工智能, 先進的電腦科技,生物科技,奈米科技 分別是資訊科技不同的的核心領域. 然後我們將會將它們應用到其他領域, 例如能源, 生態環境, 政策法律和道德, 企業態度, 使人們可以把這些新技術帶給世界.
So we're very appreciative of the support we've gotten from both the intellectual leaders, the high-tech leaders, particularly Google and NASA. This is an exciting new venture. And we invite you to participate. Thank you very much. (Applause)
我們非常感謝我們所得到, 來自知識份子和高科技領導人們的支持, 特別是Google和NASA. 這是個興奮的全新研究. 我們誠心地邀請你的加入. 謝謝. (鼓掌)