文化與技術三部曲 · SF-10｜專訪 Dennis Allison

---

Dennis Allison: I went to school at UCLA for a while and then moved to Berkeley, where I studied physics. Physics was an exciting field in 1960. And I left the university with an undergraduate degree in 1961 and did other things for a little while. I lived in Berkeley and in the summer of 1962, I went and took a job with SRI as a physicist to study geophysics. I was responsible for a number of projects there. The 1960s in the United States was a little chaotic. We were deeply involved in war in Vietnam. We had the view of Russia SRs or rather the Soviet Union of the time has a well, they were friendly in some sense. We found a way to deal with them and they’re with us. But we couldn’t really figure that was going to last for long. And the defense industry was very significant here in California and elsewhere. So one of the things which you had to contend with is a recent young person, recent graduate, was that you had the potential for being drafted. And I was drafted and found a job which would allow me to stay without outside of the military, working for the military industrial complex, as it were. I work for SRI, which was at that time a part of Stanford University. It had a separate facility operated primarily with contracts with outside of the University. The contracts were usually with the the Defense Department or something like that. And so I did that. We were doing research basically and the job I was hired to work on was to be a physicist, but not in high energy physics but rather in geophysics. So we would tell sorts of pre-physical things. In particular I learned a lot about the ionosphere and a lot about radio way propagation through interesting complex media.

丹尼斯·艾里森：我曾在UCLA上學一段時間，然後搬到伯克利，學習物理學。1960年，物理學是一個令人興奮的領域。我在1961年獲得了本科學位，然後做了一些其他的事情。我住在伯克利，在1962年的夏天，我去了SRI工作，擔任物理學家研究地球物理學。我在那裏負責了許多項目。1960年代的美國有些混亂。我們深陷越南戰爭。我們對當時的蘇聯持有友好的態度。我們找到了一種與他們打交道的方式，他們也與我們保持聯繫。但我們無法確定這種情況能持續多久。國防工業在加利福尼亞和其他地方非常重要。所以作爲一個年輕人，作爲一個剛剛畢業的人，你必須面對的一個問題是有可能被徵召入伍。我被徵召了，但找到了一份工作，讓我能夠在軍隊之外工作，爲軍工複合體工作。我在當時是史丹佛大學的一部分的SRI工作。 它有一個獨立的設施，主要是通過與大學外部的合同運營。這些合同通常是與國防部之類的機構簽訂的。所以我就這樣做了。我們基本上是在進行研究，而我被聘請的工作是作爲一名物理學家，但不是在高能物理學方面，而是在地球物理學方面。所以我們會研究各種預物理的事情。特別是我對電離層和通過有趣的複雜介質的無線電波傳播有了很多瞭解。

When did research? Of course our research attempt was based on having large amounts of data and processing that data. And that processing had been done for a long time by hand, but it was very slow and not very useful and so we computerize things. We originally, like the laboratory I work in, I got a CDC 160 EA which was a little tiny machine and I got the job of making it work. I rewrote pieces of the Fortran compiler that came with the machine so that in fact we could use it. We I worked on other machines as well as part of the same program. If you’ve been down at Stanford and look on the skyline, you can still see if there’s a 150 foot dish up there, the control program of which at least in one of its instantiations, I wrote (05:00 unclear) to have a right programs that took originally a fruit PDP5 and then later a PDP8, the controls for this very larger dish which tracks celestial bodies and could actually examine things by sending radio signals out and getting response back–that is a big radar. Or we’ll just receive radio signals from other sources or whatever. And it was interesting research and we looked at rotation rates of planets and did reasonable research there. This was exciting stuff. And this is a good thing. You’re young and ambitious and you’re willing to stay up all night to watch the (unclear). It does its thing. That’s good. The the mid 60s was a transition for me. I was married in a 1966 and we examine many different other things. In between that period the ACM had a national conference. Now ACM is the association for Computing Machine Net and I had been active in that organization. Because as time went on, my physics became less significant with what I did and the computer work which I had began at Berkeley and I became a major portion of what I did.

我們進行研究的時間是什麼時候？當然，我們的研究嘗試是基於擁有大量的數據並處理這些數據。而且這種處理過程很長時間都是手工完成的，但速度非常慢且不太有用，所以我們開始計算機化。最初，就像我所在的實驗室一樣，我得到了一臺名爲CDC 160 EA的小型機器，我負責讓它運作起來。我重寫了機器附帶的Fortran編譯程序的部分代碼，以便我們能夠使用它。我還在同一個計劃的其他機器上工作過。如果你曾經去過史丹佛大學並仰望天際線，你仍然可以看到那裏有一個150英尺的天線，其中的控制程序至少在其中一個實例中是我寫的（05:00不清楚），它擁有正確的程序，最初使用的是一臺Fruit PDP5，後來是PDP8，用於控制這個非常大的天線，該天線可以追蹤天體並通過發射無線電信號並接收回應來觀察事物-這是一個大型雷達。或者我們只是接收來自其他來源的無線電信號等等。這是一個有趣的研究，我們研究了行星的旋轉速率，並在那裏進行了合理的研究。 這真是令人興奮的事情。而這是一件好事。你年輕有抱負，願意熬夜觀看（不清楚的事物）。它發揮作用。這很好。對我來說，60年代中期是一個轉折點。我在1966年結婚，我們研究了許多其他不同的事情。在這段期間，ACM舉辦了一次全國性的會議。現在ACM是計算器網絡協會，我在該組織中非常活躍。隨着時間的推移，我的物理學在我所做的事情中變得不那麼重要，而我在伯克利開始的計算器工作成爲我所做的主要部分。

My my computer work at Berkeley was sort of interesting. Berkeley got a computer and I got a part time job working for Dudley Hirschbach who got the Nobel prize one year–this was before he got the Nobel Prize. In any case, one of the things that what he hired me for was to write a computer program to simulate the shape of the plasma beam that was coming out of a long thin tube going into a vacuum chamber, and I did that successfully eventually. And eventually it is being started out with learning program in Fortran and then learning how to modify the operating system in real time so that you would do what it’s supposed to do, and then finding a hardware bug in a new computer and things like that. But anyhow eventually that was my first paid computing job, I guess, and it was pretty exciting. In any case, jumping back to the mid 60s. I worked at SRI. I had control of and ran a little data center for some work we were doing. I as (08:30 unclear) physics and radio physics in general. And so I learned how to operate and maintain and extend the XDS 930 machine, which is a mini computer. And we use this essentially the same machine as the Engelbart group did, except that ours was bigger and faster and we had a disk and they didn’t. So it was an exciting thing we did. Again, I wrote interesting programs there. For example, I used integers instead of floating point numbers to do some computations and signal processing, which we’re fairly interesting and I did wrote a beam forming program where we combine a number of parameters to build a special radar signal which we use to identify particular properties, which are all very exciting. And sort of about the same time I was the chairman of the local chapter of the ACM. And we had an ACM national conference here and at that conference I met a guy named Bob Albert. He was giving a talk on computers and education, something I was interested in and we spoke afterwards and he said, well, I need you to be on the board of the new company I’m forming, People’s Computer Company, PCC. And so he and I and Lois Brand, who is Stewart Brand’s soon-to-be ex-wife at the time. And I think the attorney who originally wrote the papers were all on the the original incorporation. I served as the chairman of the corporation for about 15 or 16 years.

我在伯克利的計算機工作有點有趣。伯克利買了一臺計算機，我找到了一份在杜德利·赫什巴赫那裏的兼職工作，他在一年前獲得了諾貝爾獎-這是在他獲得諾貝爾獎之前。無論如何，他僱用我其中一個原因是讓我寫一個計算機程序來仿真從一個長而細的管子進入真空室的等離子束的形狀，最終我成功地做到了。最終，我開始學習Fortran編程，然後學習如何實時修改操作系統，以使其按照預期運行，還發現了新計算機的硬件錯誤等等。但無論如何，那可能是我第一份有薪水的計算工作，非常令人興奮。無論如何，回到60年代中期。我在SRI工作。我負責並運營一個小型數據中心，用於我們的一些工作。我對物理學和無線電物理學有興趣。因此，我學會了如何操作、維護和擴展XDS 930機器，這是一臺小型計算器。 我們使用的機器基本上與 Engelbart 團隊使用的一樣，只是我們的機器更大更快，而且我們有一個磁盤，而他們沒有。所以我們做的事情很令人興奮。同樣地，我在那裏寫了一些有趣的程序。例如，我使用整數而不是浮點數來進行一些計算和信號處理，這些都相當有趣。我還寫了一個波束形成程序，將多個參數結合起來建立一個特殊的雷達信號，用於識別特定的特性，這些都非常令人興奮。大約在同一時間，我擔任當地 ACM 分會的主席。我們在這裏舉辦了一次 ACM 全國會議，在那次會議上，我遇到了一個叫 Bob Albert 的人。他正在講述關於計算機和教育的話題，這是我感興趣的，我們事後交談，他說，嗯，我需要你加入我正在成立的新公司的董事會，這家公司叫做 People's Computer Company，簡稱 PCC。於是他、我和當時還是 Stewart Brand 的前妻 Lois Brand，以及最初起草文件的律師，都成爲了公司的創始人之一。 我在這家公司擔任主席大約15或16年。

PCC started out with the goal of teaching people about computers. And we were exceptionally successful, very surprising myself. I think we can make a claim too, and somehow invented personal computing in a sense, at least a lot of what we wanted to, we would say today is important for personal computing.

PCC的起初目標是教導人們有關計算機的知識。令人驚訝的是，我們取得了非常成功的成果，甚至讓我自己都感到意外。我認爲我們也可以提出一項主張，某種程度上可以說我們在某種意義上發明了個人計算機，至少我們當時所追求的很多內容，今天被視爲個人計算機的重要元素。

The the impact it had was fairly significant. First of all, PCC didn’t have any funding. We weren’t sponsored by anybody. We had a microscopic amount of capital that came from from Portola Institute and others. We had access to two machines that Dymax have gotten. This is before the microprocessor. So these were (12:10unclear) PDP8. PDP8 is a little tiny machine. It was sort of cheap. It’s like a toy, well, not a toy. More powerful than Apple, less powerful than whatever. Any case, it was a bargain at about 25 thousand dollars each and they would have a terminal was a ESR (12:40 unclear) It take characters as a second. I was very interesting came with a basic and we use that to take and demonstrate programming in schools.

它所產生的影響相當顯着。首先，PCC沒有任何資金支持。我們沒有任何贊助商。我們只有一點點來自Portola Institute和其他機構的微薄資本。我們可以使用Dymax獲得的兩臺機器。這是在微處理器出現之前。所以這些機器是PDP8。PDP8是一臺很小的機器。它有點便宜。不是玩具，比Apple更強大，比其他機器更弱。無論如何，每臺機器大約2.5萬美元，它們都有一個終端，是一個每秒可以接收字符的ESR。我非常有興趣，它還帶有一個基本編程語言，我們用它來在學校進行編程演示。

During the time that PCC was actively teaching kids we would go to classrooms usually 4th and 5th grade. You could get somebody to come with a computer and gameplay with with kids and birthday parties we started and ran a Community Computer Center. Now let’s say collection of people and to toys and books and documents which was available for kids to come to after school. And they could buy computer time. For a quarter you could get an hour. I shared PDP8 computer and people did that and older people came too. We had people doing writing programs and so forth and the staff at CCC, the Community Computer Center was volunteer and they were really amazing people. we had (unclear), I don’t know what her title was, but it was I think organizer or whatever who ran classes and organized the classes and kept everything going. But we had other people too, we had a came of people well, you could buy a video terminal–characters only, sort of video color type which was made by ADM. And you could buy it as a completed unit or you could buy it as a kit, which means that you got the parts and you had to solder them in. And so PCC actually or rather PCC actually made some of its budget by making and selling the terminals to people who are interested. And people would come into PCC, rent computer time and create a game which was later published in PCC’s journals. Now PCC was really interested in promoting computing. And so what we did, the reason it sort of work had, was serendipitous, sense of surprising. Menlo Park had a publisher in town who published a basically newsletters and ads and that sort of thing, but they had a web press, a big one. And for a very small amount of money, a couple thousand dollars, we could print 5000 copies of a 10 or 15 pages newspaper. And it’d be done and we could give them away to people for free which is what we did. So PCC was basically a way of promoting computers and training people and teaching classes and it just sort of happened.

在PCC積極教授孩子們的時候，我們通常會去四年級和五年級的教室。你可以找人帶着一臺計算機來和孩子們一起玩遊戲，我們還開始了一個小區計算機中心，舉辦生日派對。現在，我們擁有一個人們捐贈的玩具、書籍和文件的收藏，供孩子們放學後使用。他們可以購買計算機使用時間，一個季度只需一塊錢就可以使用一小時。我們共享一臺PDP8計算機，年長的人也會來使用。我們有人負責編寫程序等等，小區計算機中心的工作人員都是志願者，他們真的非常出色。我們有一位（不清楚）的人，我不知道她的職位是什麼，但我想她是組織者或者其他什麼，她負責開設課程、組織課程並保持一切運轉良好。但我們還有其他人，我們有一羣人，你可以購買一個只能顯示字符的視頻終端，類似彩色視頻類型，這是由ADM公司生產的。你可以買一個成品單元，也可以買一個套件，意味着你得到零件，然後自己焊接。 所以PCC實際上或者更確切地說PCC實際上通過製造和銷售終端設備來賺取一部分預算。人們會來到PCC，租用計算機時間並創建遊戲，這些遊戲後來會在PCC的期刊上發表。現在PCC對推廣計算非常感興趣。所以我們所做的事情，之所以有點奏效，是因爲它是偶然的，具有令人驚訝的意義。門洛帕克鎮上有一家出版商，主要出版通訊和廣告之類的東西，但他們有一臺大型網絡印刷機。只需花費很少的錢，幾千美元，我們就可以印刷5000份10到15頁的報紙。然後我們可以免費向人們分發這些報紙，這就是我們所做的。所以PCC基本上是一種推廣計算機、培訓人員和教授課程的方式，這只是一種偶然發生的情況。

Much of the material that we published came from our readers who had taken a previous article, done something or had an idea and done something, forbidden it down, set it in polish so it was very definitely a grassroots operation which was very much the kind of community growth that we were trying to throw. In some ways, that was the most creative moments of PCC. But we did well there. And the newspaper ended up having a fairly substantial circulation, but it generated no revenue. So we decided that what we wanted to do was to migrate the newspaper, PCC newspaper, to People’s Computers. A new computer magazine targeted the same audience that People’s Computers were the newspaper, but with a slightly different format as the standard magazine format, printed on slightly better paper and there was a subscription fee. And so we set out the notice when we got lots of subscriptions and we became sort of self sustaining.

我們發表的許多內容都來自於我們的讀者，他們從先前的文章中獲得靈感，做了一些事情或者有了一個想法並付諸實踐，將它們整理出來，使之更加完善。這絕對是一個基層運作，非常符合我們試圖推動的小區成長的方式。在某種程度上，這是PCC最具創造力的時刻。我們在那裏做得很好。這份報紙最終擁有相當大的發行量，但並未產生任何收入。因此，我們決定將我們想做的事情轉移到People's Computers上。這是一本新的計算機雜誌，針對與PCC報紙相同的讀者羣，但格式稍有不同，印刷在稍微好一些的紙上，並且需要訂閱費。因此，我們發出通知後收到了很多訂閱，我們變得有點自給自足了。

This is an operation which really was run by people who cannot seem to require a large amounts of money to sustain themselves. People were there because they wanted to be there. They bought into the the idea that people should learn about computers and why they were willing to take the time and effort to learn. And they did very well. We did very well. And People’s Computers were very good. (unclear) . She likes a magazine. The editor for that was Phyllis Cole. And they will see this as we continue on.

這是一個由那些似乎無法要求大量金錢來維持自己的人運營的操作。人們在那裏是因爲他們想在那裏。他們相信人們應該學習關於計算機以及爲什麼他們願意花時間和精力去學習。他們做得非常好。我們做得非常好。而且人民的計算機非常好。（不清楚）。編輯是菲利斯·科爾喜歡雜誌，隨着我們的發展，他們將會看到這一點。

The whole community was interesting because we had people in academia and people in the industry. We have the personal computer people, the hobbyists and so forth and lots of other groups, all of which sort of merge together and it turns out that most of the people were in two or three of the groups. So people who worked at HP would be eventually interested in building hardware for their personal computer that they bought from somebody else who was making personal computers like Steve Wozniak. So people were experimenting, looking into things, citing how to do the things and so forth.

整個小區非常有趣，因爲我們有學界的人和業界的人。我們有個人計算機的人，愛好者等等，還有很多其他的團體，所有這些團體都有點融合在一起，結果發現大部分的人都同時身處於兩個或三個團體之中。所以在惠普工作的人最終會對從像史蒂夫·沃茲尼亞克那樣的個人計算機製造商那裏購買的個人計算機進行硬件建造方面的興趣。所以人們在實驗、探索事物，並研究如何做這些事情等等。

I worked at SRI. And at that point, I’m a good SRI. I was involved with the People’s Computer Company and with couple of other organizations, ACM for example and I was a multi-group person and so it’s a little hard to say that it was some the hobbyists from PCC who did things because the hobbyists from PCC might actually have been working on the boards for Apple or something like that. Of course that was later but not pretty much deal. So we’re talking about sort of period 1966 to 1972 or 1973. Now I have left SRI in the early seventies and worked briefly for a company that I was a part of called Polymorphic. Polymorphic provided consulting services and one of the things we did was to tell them what they needed to do in the next 10 years. And the one that I sort of really like was that we worked with a company in San Francisco. And we were doing a long range plans for a company in France, a mini computer manufacturer. And we wrote an amazingly interesting report with projections about what this company needed to do. And we were watching that technology. This is sort of early-mid 70s. And we looked at the technologies and we decided that what they needed to do is invest in learning how to build semiconductor fabs in France, to make designs of mini computers on a chip, because that was what was going to happen. They needed to look into networking, look into the interfaces with communications and printing into 3D graphics and 2D graphics and all the sorts of things that we think of today as being indicative of the future today, either in hand or coming up soon. But this was in 1974 and we had numbers and we had figured it out when this market was going to emerge. They read the report and said, this is science fiction, which I could (unclear) pay you? And they were, 15 years later, right on the money. So it made me feel good, but we had to fight for payment. It is very interesting for me to see how naive people are in the face of tradition(?).

我在SRI工作。那時候，我是一個出色的SRI員工。我參與了人民計算機公司和其他幾個組織的工作，比如ACM。我是一個多組織的人，所以很難說是PCC的業餘愛好者做了一些事情，因爲PCC的業餘愛好者可能實際上正在爲蘋果公司等公司製作電路板。當然，那是後來的事情，但並不是非常重要。所以我們談論的是大約1966年到1972年或1973年的時期。現在，我在七十年代初離開了SRI，並短暫地爲一個叫做Polymorphic的公司工作。Polymorphic提供諮詢服務，我們做的其中一件事就是告訴他們未來十年需要做什麼。我特別喜歡的一個項目是我們與舊金山的一家公司合作。我們爲一家法國的小型計算機製造商制定了長期計劃，並撰寫了一份非常有趣的報告，預測了這家公司需要做的事情。我們一直在關注那項技術。這是七十年代初中期的事情。 我們研究了各種技術，並決定他們需要投資學習如何在法國建立半導體晶圓廠，設計芯片上的迷你計算機，因爲那就是即將發生的事情。他們需要研究網絡技術，研究與通訊和印刷相關的接口，以及3D圖形和2D圖形等我們今天認爲是未來的指標性技術，無論是現在還是即將到來。但這是在1974年，我們有數據，我們已經計算出這個市場何時會出現。他們讀了報告，說這是科幻小說，我能給你付錢嗎？而15年後，他們完全正確。這讓我感到很好，但我們必須爲付款而奮鬥。對我來說，看到人們在傳統面前是多麼天真，這是非常有趣的。

So anyhow, in 1972, a team in Intel made a chip called the 4004. This is a four mid micro processor. Jalen Pmos (unclear) and as I remember it was 35 volt pin loss. He was not really compatible with modern tech. Concurrently, they did another chip as a special for a Texas terminal manufacturer (unclear), which became the Intel 8008. And I got hired by Intel as a contractor to write the assembler for the 4004. 4004 was the usual machine (unclear). But it was really the first microprocessor single chip. In truth, it was actually three different chips, but the passes one chip twice(unclear) It is fairly slow. It was quite expensive and become big and terrible popular. It was processor built for a company called Busy Com and the architecture was done by Busy Com in part and the implementation at Intel was really very nice. The 8008 was sort of done concurrently again by the same team at Intel and by a team of Texas Instruments. The Intel processor became the one that was most successful and it was done in Enmos and had a more easily compatible (27:07 unclear) structure. The 8008 was designed by it by a team including Federico Feigene (unclear) and Tel Hoff (unclear) and Stan Lazer (unclear).

所以，無論如何，在1972年，Intel的一個團隊製造了一款名爲4004的芯片。這是一款四位微處理器。Jalen Pmos（不清楚）和我記得它是35伏的引腳電壓。它與現代技術並不真正兼容。與此同時，他們爲德克薩斯終端製造商（不清楚）製作了另一款芯片，後來成爲Intel 8008。我被Intel聘爲承包商，爲4004編寫彙編程序。4004是一臺普通的機器（不清楚）。但它實際上是三個不同的芯片，但通過一次芯片傳遞兩次（不清楚）。它相當慢。它非常昂貴，但卻非常受歡迎。它是爲一家名爲Busy Com的公司構建的處理器，Busy Com部分完成了架構設計，而Intel的實施非常出色。8008是由Intel的同一團隊和德州儀器的一個團隊同時完成的。Intel的處理器成爲最成功的處理器，它採用了Enmos技術，並具有更易於兼容的結構（27:07不清楚）。 8008 是由包括 Federico Feigene（不清楚）、Tel Hoff（不清楚）和 Stan Lazer（不清楚）在內的團隊設計的。

Later on, Stan and no one(unclear), some of the thousand eighteen(unclear) went to form Zilog and Zilog made the Z80 which was a slight extension of the 8080, which was the extension of the 8008 and that became sort of the standard processor and this is in mid seventies. The IBM processor which was based on the 8088 which was a variation on the 8080 which had a 8 pairs supposed to 16 bit pass.

後來，Stan和某人（不清楚）在一千零十八年（不清楚）組成了Zilog，並且Zilog製造了Z80，這是8080的輕微擴展，而8080則是8008的擴展，這成爲了一種標準處理器，這是在七十年代中期。IBM處理器則是基於8088，這是8080的變體，具有8對16位通道。

I have the privilege of being the keynote speaker at the kick off for the IBM PC project. (unclear) where everything was super secret. At that time well, the 8080 came out, before that the 8008 was available but it had some architectural issues. So some people have been working on how you make that into a reasonable processor. Kirei (unclear) We have worked on and started a company called Digital Research and John Turo (unclear) , he is a friend from University of Washington I had started another company called Digital Systems. Both of which were very successful Visual Systems, in fact was sold tremendous number of systems into the Pacific Rim years ago. Digital Research was quite famous for its CPM, which is a operating system and compilers he wrote compilers for several of those systems. Concurrent to the same when the 8080 came out, and this was in 1976, there are parts I pull you. The processors 8080 was the processor to use. And Intel sold a bunch of them. The 8080 was used by a company and (30:45 unclear) and used to be the processor in the Altair personal computer.

我有幸成爲IBM PC項目啓動儀式的主講人。那時候一切都是超級機密的。在那個時候，8080處理器問世了，之前有8008可用，但它存在一些架構問題。所以有些人一直在研究如何將其打造成一個合理的處理器。Kirei（不清楚）我們曾經合作並創辦了一家叫做Digital Research的公司，而John Turo（不清楚）是我在華盛頓大學的朋友，他創辦了另一家叫做Digital Systems的公司。這兩家公司都非常成功，特別是Visual Systems，多年前它向太平洋地區銷售了大量的系統。Digital Research因其操作系統CPM而非常有名，他還爲幾個系統編寫了編譯器。與此同時，8080處理器問世了，那是在1976年，有些部分我會跳過。8080處理器成爲了主流處理器，Intel賣出了大量的8080處理器。一家公司（30:45不清楚）使用了8080處理器，並將其作爲Altair個人電腦的處理器。

Now PCC at that time had been running computer centers, talking about how wonderful BASIC was teaching people about BASIC. And when the Altair came out and you could be able to actually think about buying your own machine as opposed to spending a quarter per hour to use a shared PDP8. It was quite appealing. So the the Altair was an interesting machine for what it would seem like an outrageously large price today you could get a processor and 4000 bytes of memory. And that was not really very much. With that time, Bob and I were talking one day, and I said, Bob, it’s not too hard to write a BASIC for this machine and make it very small and we can throw things out of BASIC and make it even smaller. And so we decided to do that. So I got a request from PCC to write an article about doing this tiny BASIC as what we call. Well, Tiny BASIC got written up and it was published in the newspaper and also in discussing People’s Computers. And suddenly there’s a lot of interest. With a lot of interest, we got ambitious. We said, okay we have and I think we had 4,500 letters in the space of the first two weeks talking details about tiny BASIC and what they wanted and so forth. And so we decided we would start up, and instead of putting this in the newspaper, we would have a separate journal called Tiny BASIC Notes and that lasted about two weeks. And then we decided we’re gonna do a bigger journal because we had too much material to put in the load Xerox to version of this thing so we started another journal. It became Doctor Dobb’s Journal. Our pipe setter rebeka linski (unclear) decided that it should be called Doctor Dobb’s. And I was the (unclear) part of it and Bob was the Ops and Bob and I were the founders of this journal, which he also took some of the pictures of some photographs of submotions photography and used a tag on that, called Running Light without overbite, which in English is a bit of a pun because the pictures were of a nude running man. But it was fine and it set us aside from the journals like head letters, words in their titles.

當時的PCC一直在運營計算機中心，談論着BASIC的優點，教導人們如何使用BASIC。當Altair推出時，你可以考慮購買自己的機器，而不是每小時花一刻錢使用共享的PDP8，這非常吸引人。所以Altair是一臺有趣的機器，以今天來看，價格可能看起來非常高，你可以獲得一個處理器和4000字節的內存，這實際上並不多。在那個時候，Bob和我有一天在談話，我說，Bob，寫一個這臺機器的BASIC並使其非常小並不太難，我們可以從BASIC中刪除一些東西，使其更小。於是我們決定這樣做。所以我收到了PCC的一封請求，要求我寫一篇關於編寫這個小型BASIC的文章，我們稱之爲Tiny BASIC。Tiny BASIC被寫了出來，並發表在報紙上，也在People's Computers上進行了討論。突然間，引起了很多人的興趣。由於興趣很大，我們變得雄心勃勃。我們說，好吧，我們在前兩個星期內收到了4500封信，關於Tiny BASIC的細節和他們的需求等等。 所以我們決定開始行動，而不是將這個放在報紙上，我們會有一個單獨的期刊叫做《微型BASIC筆記》，那個持續了大約兩個星期。然後我們決定要做一個更大的期刊，因爲我們有太多的材料無法放在這個東西的Xerox版本中，所以我們開始了另一個期刊。它變成了《Doctor Dobb's Journal》。我們的管道設置者Rebeka Linski（不清楚）決定它應該叫做《Doctor Dobb's》。我是其中的一部分，Bob是運營部分，而Bob和我是這個期刊的創始人，他還拍了一些次運動攝影的照片，並在上面使用了一個標籤，叫做《Running Light without overbite》，這在英語中有點雙關意味，因爲這些照片是一個裸體跑步的人。但這沒關係，它讓我們與那些標題中有字母的期刊區別開來。

In any case, Doctor Dobb’s Journal is interesting because it first of all opened up with a quite a large population of subscribers. And we’ve got fairly large and I think it’ll be that very large number of subscribers. And I actually don’t know how many. But it was popular and it had published code unlike anybody else. Nobody else had started. Code was worth publishing. But we didn’t publish code, we published code systems, and we publish compilers and we publish those programs that did silly things and things that did complicated things. And when I was teaching at Stanford and I would have students who submit their projects as possible things. One of my students at Stanford group, a real time operating system for the 8080 in Gary Kildall’s programming language PLM. And we publish that and its code. And we publish a number of things. Tiny BASIC, well, when we publish the description of Tiny BASIC, it was precisely very popular.

無論如何，Doctor Dobb's Journal之所以有趣，首先是因爲它開始時就有相當多的訂閱者。我們擁有相當大的訂閱羣衆，我認爲這個數字將會非常龐大。實際上，我不知道有多少人訂閱了。但它很受歡迎，並且發表了其他人沒有的程序代碼。沒有其他人開始發表程序代碼。程序代碼是值得發表的。但我們不僅發表程序代碼，還發表程序代碼系統，發表編譯程序，發表那些做蠢事和複雜事情的程序。當我在史丹佛教書時，有些學生會提交他們的項目作爲可能的事情。我在史丹佛的一個學生小組，使用Gary Kildall的PLM編程語言爲8080開發了一個實時操作系統。我們發表了它和它的程序代碼。我們發表了很多東西。Tiny BASIC，當我們發表Tiny BASIC的描述時，它非常受歡迎。

Anyhow, we had Tiny BASIC running. I had written a three part article. The first part of the article sketched out the design. It didn’t have any much detail. And two weeks after we published it, we got two copies of the Tiny BASIC setting by people who had implemented it from the description we had made. And some of them actually implemented a little larger language and so forth. And they all were very small. They were all tiny, which is what the goal here was. And at that point, it was amazing. They republish both of them. They were published in split optical, which is two hexadecimal digits as two sets of actual numbers but of course they were bigger. So there are three digit numbers and so forth. And the reason was that was convenient for keying it in to the Altair front panel, because there was no hardware device for storage. And so if you wanted to have an Altair which ran BASIC, you had to put the code in to the machine by hand. And that was roughly 2000 entries. And it’s amazing. So there were enormous numbers people who did this. The interest was so high in making this thing work and getting something really reasoning or not.

不管怎樣，我們成功運行了Tiny BASIC。我寫了一篇三部分的文章。文章的第一部分勾勒了設計，但沒有太多細節。在我們發表文章兩週後，我們收到了兩份Tiny BASIC的設置副本，這些副本是根據我們的描述實現的。其中一些人實際上實現了一個稍微大一點的語言等等。它們都非常小，這也是我們的目標。當時真是令人驚訝。他們重新出版了這兩份副本。它們以分割光學的形式出版，即兩個十六進制數字作爲兩組實際數字，但當然它們更大。所以有三位數等等。原因是這樣方便將其輸入Altair的前面板，因爲沒有存儲的硬件設備。所以如果你想要一個運行BASIC的Altair，你必須手動將代碼輸入機器。大約有2000個條目，這真是令人驚歎。有很多人這樣做，對使這個東西工作並得到真正的推理或不推理的東西的興趣非常高。

In any case, we were talking earlier about the changes that were happening at the computing area. The whole personal computing area was expanding at a great rate and lots of people were involved in lots of different things. And it all happened sort of immediately. And most of the groups that were doing things had people in them that were part of other groups too. So it’s not easy to put your finger on a particular thing.

無論如何，我們之前談到了計算領域正在發生的變化。整個個人計算領域正在以驚人的速度擴展，許多人蔘與了許多不同的事情。而且這一切似乎都發生得很突然。大多數進行各種活動的團體中，成員也同時是其他團體的一部分。所以很難確定具體的事情。

When Doctor Dobb’s Journal started, I hired a student of mine, Jim Warren, who was at University of California of at San Francisco and at Stanford. Well, it was serendipitous because I was teaching in both places. Jim was a student for a PhD in computer science and he had thought that he would do it in medical informatics and I was teaching in The Medical Informatics program in UC San Francisco. He decided it to move to Stanford and became a student there and I inherited him in Stanford for a time. (unclear) you see San Francisco he had Stanford very deeply involved in the counterculture. Jim was perhaps larger than life. A marvelous friend, very interesting guy, very non-standard his belief structure and how the world work was different than most. And he found an admiring community in the personal computer world. So he was a perfect Doctor Dobb’s Journal editor and he enjoyed running around and doing things and going to meetings and transcribing stuff that people said in various places, telling Steve Jobs or Steve Wozniak that he had an article that he needed to have written and getting them to right things(unclear), and stuff like that. He was also the chair of a counterculture group running it. It’s in the area called the Free University. The idea was to replace the at university educations with things that are much more relevant. And the Free University was exciting and it had classes and all sorts of things. Some things that probably, well, let me see if I can think of some juicy ones. There were lots of classes in how the mind works and how do people make decisions and what sort of foods you should eat and what sort of food you shouldn’t eat, and some discussions of drugs and alcohol and that sort of thing, and psychotic experiences. And, well, there was one group, I remember that, whose goal was to have an opportunity to have a meal of human flesh. Well. Somewhat unusual, but anyway, it was a great community of people exploring things. And this is, this all happens sort of about the same time that the personal computer industry group is starting, people are starting companies like Wozniak and Jobs starting Apple. And other people, graduate students at Stanford having a good idea as to how to recognize speech, building speech recognizers that work with Apple computers and people doing terminals and so forth. I had a couple of students who built a machine which was essentially equivalent to the Xerox Park machines, but ran on an 8080 and were easily accessible on the source code. And it was lovely. They later (unclear), founded a company called Logitech, still around today. But in any case, this was lots of things and people with sort of different ideas as to how society ought to work. We’re built into that and many people explore different alternatives. And you look within the the community of people in the personal computer area, you find people who have very divergent ideas about how society should work, shouldn’t work and so forth. There are people who coined some of the words that we think of as being interesting today, like “the Internet wants to be free”. It will route around problems and things of that nature. So these people have always been there and they are an integral part of the creative process that has gone into building computing and access to computing.

當《Doctor Dobb's Journal》剛開始時，我聘請了我的學生Jim Warren，他在加州大學舊金山分校和史丹佛大學就讀。這真是一個巧合，因爲我在這兩個地方都教書。Jim正在攻讀計算器科學的博士學位，他原本打算在醫學信息學方面進行研究，而我則在加州大學舊金山分校的醫學信息學項目中任教。他決定轉到史丹佛大學並在那裏繼續學業，我在史丹佛大學接手了他一段時間。（不清楚）你知道，舊金山和史丹佛大學都深深地參與了反文化運動。Jim可能是個傑出的人物。他是一個了不起的朋友，非常有趣，他的信仰體系和對世界運作的看法與大多數人不同。他在個人計算機界找到了一個欽佩他的社羣。所以他是一個完美的《Doctor Dobb's Journal》編輯，他喜歡四處奔波，參加會議，抄寫人們在各個地方說的話，告訴史蒂夫·喬布斯或史蒂夫·沃茲尼亞克他需要寫一篇文章，並讓他們寫一些東西（不清楚），等等。 他還是一個反主流文化團體的主席。它位於一個被稱爲自由大學的地區。這個想法是用更相關的事物取代大學教育。自由大學令人興奮，有各種各樣的課程和活動。有一些可能很有趣的課程，讓我想想有哪些。有很多關於思維如何運作、人們如何做決策、應該喫什麼樣的食物以及不應該喫什麼樣的食物的課程，還有一些關於藥物、酒精和精神病經歷的討論。還有一個我記得的小組，他們的目標是有機會喫人肉。有點不尋常，但無論如何，這是一個探索事物的偉大社區。這一切發生的時間差不多是個人電腦產業集團開始的時候，人們開始像沃茲尼亞克和喬布斯那樣創辦蘋果公司。還有其他人，史丹佛大學的研究生們有了一個關於如何識別語音的好主意，他們正在爲蘋果電腦開發能夠識別語音的軟件，還有人在做終端等等。 我曾經有幾個學生建造了一臺與施樂帕克機器基本相當的機器，但是它是在8080上運行的，並且源代碼很容易獲取。它非常出色。他們後來（不清楚）創辦了一家叫做羅技的公司，至今仍然存在。但無論如何，這是很多事情和人們對社會應該如何運作有不同想法的結合。我們融入其中，許多人探索不同的選擇。在個人電腦領域的社區中，你會發現人們對社會應該如何運作、不應該如何運作等有着非常不同的觀點。有些人創造了我們今天認爲有趣的一些詞語，比如“互聯網想要自由”。它會繞過問題並解決類似的事情。所以這些人一直存在，並且是構建計算和獲取計算的創造過程中不可或缺的一部分。

Cuiyu: Yes, I think then the second question will be will be more about your personal experiences. But maybe you have all already mentioned that.

翠玉：是的，我想第二個問題可能會更多關於你的個人經歷。但也許你已經都提到了。

Dennis Allison: Well, I taught in UC San Francisco. I taught in Stanford. I gave first. It might have been 1977. One of my college’s roommates, Dave Hargers (unclear), was a professor at Berkeley. He later became dean of Engineering at Berkeley and became a recently(unclear)died. It’s very sad. He’s gone now. But in case the two of us gave a tutorial on microprocessors in sort of the year after they were readily available, so that must been about 1976,75. In any case we did for UC and C Extension and we had quite a large number of attendees. I don’t remember quite the number but it was enough to be fulfill the budget for University Extension for quite a long time. We had thousand, maybe. Year after that I gave a tutorial in Japan on micro processors with a bunch of other people and we had problems in the venue because we kept having to increase the size of the facility because we had so many attendees. This was before and you could get video, so people tend to be there to watch it. But we had several thousand attendees in this particular presentation. And basically we went through all of the things that you could do with microprocessors.

丹尼斯·艾里森：嗯，我在加州大學舊金山分校教書。我也在史丹佛教過書。我首次講課可能是在1977年。我的一位大學室友，戴夫·哈格斯（不太清楚），是伯克利的一位教授。他後來成爲了伯克利工程學院的院長，最近不幸去世了。這真是令人傷心。他現在已經離開了。但是在那之後，我們兩個人在微處理器方面進行了一次教學，大概是在它們普及後的一年，所以可能是在1976年或1975年左右。無論如何，我們爲加州大學和C Extension進行了這次教學，參加人數相當多。我不太記得具體數字，但足以爲大學延伸部門的預算提供相當長的時間。可能有上千人吧。之後的一年，我和其他一些人在日本進行了一次有關微處理器的教學，我們在場地上遇到了問題，因爲參加人數太多，我們不得不不斷擴大場地的規模。那時候還沒有視頻，所以人們都會親自到場觀看。但在這次演講中，我們有幾千名參加者。基本上，我們介紹了所有可以用微處理器做的事情。

A few time before that, I was involved with the USA-Japan conference, which was a conference that was trying to build commercial relationships between US and Japanese companies. One of the conferences that I was involved with was something called Nicograph, I think it was 1974. And it had an impact as I think was surprising. I ended up for reasons that I’m not sure, being the program chairman and I work with with my friend, Sawi Shizuka (unclear). (unclear) was a student of (unclear). And in any case we had the job of putting together program and for reasons that escape me even today, we ended up building a computer program about computers and gaming. That was remarkably well populated, had lots of really interesting talks and many Japanese companies decided that they want to be into computer games as a result. And so much of the Russia, the computer, home computer, personal computer computing thing, which really comes in some sense from the impact of Nintendo and a couple of other Japanese computer game companies really comes from having been exposed to these talks and discussions by US people in Japan at the (unclear) conference and I got to meet the guy who invited the. So I saw Spencer for.(unclear) this is this beautifully shaped thing with two holes on the top to you for (unclear). He got a design award from the computer graphics people. I think because this was a design that was done entirely by by computer so this is very interesting but anyhow it’s great conference and remarkably successful and filled with people who who were involved with doing gaming at one in multiple layers.

在那之前的某個時候，我參與了美日會議，這是一場旨在建立美國和日本公司商業關係的會議。我參與的其中一個會議叫做Nicograph，我想是在1974年。它產生了出乎意料的影響。由於某種原因，我最終成爲了項目主席，並與我的朋友沙維·靜香（不清楚）合作。他是（不清楚）的學生。無論如何，我們的任務是組織一個關於計算機和遊戲的計劃。這個計劃非常成功，有很多非常有趣的演講，許多日本公司決定進入電腦遊戲領域。因此，俄羅斯的計算機、家用電腦、個人電腦計算等很大程度上來自於任天堂和其他幾家日本電腦遊戲公司的影響，這些公司在（不清楚）會議上接觸到了美國人的演講和討論。我還有機會見到邀請他們的那個人。 所以我見到了Spencer。(不清楚)這是一個形狀美麗的東西，頂部有兩個孔，可以給你用。(不清楚)他從計算機圖形人員那裏獲得了設計獎。我認爲這是因爲這個設計完全是由計算機完成的，所以非常有趣。但無論如何，這是一個很棒的會議，非常成功，參與其中的人們在多個層面上都與遊戲有關。

Now is the rise of computer gaming and computer competitions of various sorts, online and offline, a part of the counterculture? And I think you probably want to say, yes, it is, in some ways it is similar to color. It definitely changed how people behave in real terms. And if you have teenage kids now or even younger pay the impact of computing at least in the United States is pretty significant. And the great debates and parents have with their children, right, is how many hours of screen time they get a day? It’s interesting. When I first encountered this Screen Time stuff, I thought, well, who cares? Okay, that’s not true

現在是計算機遊戲和各種在線和線下電競的興起，它是反文化的一部分嗎？我想你可能想說，是的，在某種程度上它與顏色相似。它確實改變了人們的行爲方式。如果你現在有十幾歲的孩子，甚至更小，至少在美國，計算器的影響是相當重要的。父母與孩子之間進行的重大辯論是，他們每天能使用多少小時的屏幕時間？這很有趣。當我第一次接觸到這個屏幕時間的東西時，我想，好吧，誰在乎呢？好吧，這不是真的。

Cuiyu: Actually we wanna know some thing about the establishment role organizing of this PCC cover. And can you briefly introduce that especially the cover of your first volume. There is a small manifesto.

翠玉：其實我們想了解一些關於這個PCC封面，你能否簡單介紹一下，特別是第一卷的封面。上面有一個小宣言？

Dennis Allison: Dream is a manifesto. Yes.

丹尼斯·艾里森：夢想是一個宣言。是的。

Cuiyu: When we read that, we keep thinking that why the manifesto happen in 1972? Like is it because of any computer at that time are expensive or are mostly used by, I don’t know, kind of big companies.

翠玉：當我們讀到這段時，我們一直在想爲什麼這份宣言發生在1972年？是因爲當時的計算機很昂貴，或者主要被一些大公司使用嗎？

Dennis Allison: In some sense the observation was, we, that is Bob and I and the other people around PCC at the time, felt that it was important that people who are other than the establishment bosses learn about computing because that is going to be the environment in which they live and work and in fact, that’s what happened. The personal computer became the normal environment. And I was gonna tell you have seen children under the age before children learn to speak, he who have learned to watch their mommy who uses her cell phone in the videos on the cell phone to distract the child. They figured out that the phone has buttons. You push the buttons in the right sequence and that gets you to your video and they grab the phone and they punch the buttons in and watch their video when their mom is distracted. So kids are very smart. They learn and there’s a long debate about whether or not screen time is good or bad. But we don’t have choice anymore. It’s part of our world.

丹尼斯·艾里森：某種意義上，我們，也就是當時在PCC周圍的鮑勃和其他人，覺得讓那些不屬於主流的人瞭解計算機是很重要的，因爲那將是他們生活和工作的環境，事實上，這就是發生的事情。個人電腦成爲了正常的環境。我要告訴你們，在孩子們學會說話之前，他們已經學會觀察媽媽在手機上看視頻的方式。他們發現手機上有按鈕。按照正確的順序按下按鈕，就能看到視頻，他們抓起手機，按下按鈕，趁媽媽不注意時看視頻。所以孩子們非常聰明。他們學習能力很強，關於屏幕時間是好還是壞有很長時間的爭論。但我們已經沒有選擇了。這已經成爲我們世界的一部分。

We didn’t talk about this, but when Jim Warren decided that PCC should have a computer fare. And so he promoted the computer fare within the organization and, number of us, when I was in the the majority here decided that was probably not a good thing for PCC to do, but we encourage Jim to do it. And so Jim divested his some part of his life from PCC and started the computer fair and organized it and it became enormous success. And he made the fortune out of it. Well, small fortune would be made of fortune and he was very good at it. And so lots of interesting things about personal computing can be found not in the proceedings of the computer fares.

我們沒有討論過這個，但當吉姆·沃倫決定PCC應該舉辦一個計算機展時。於是他在組織內推廣了這個計算機展，當時大多數人，包括我在內，認爲這對PCC來說可能不是一件好事，但我們鼓勵吉姆去做。於是吉姆從PCC中撤資了一部分，開始了這個計算機展的籌備工作，並且取得了巨大的成功。他靠這個展覽賺了一筆財富。嗯，小財富可以說是財富，他在這方面非常擅長。所以有很多關於個人計算機的有趣事情可以在這些計算機展的記錄中找到。

These are papers that are and moderately reviewed if at all and sometimes her. Difficult to understand because they don’t seem to make any sense. And probably don’t. And some of them are brilliant pieces of work of working with my friend and colleague Lee Hovel (unclear). We publish a little paper there on our epic computer games and we basically laid out the structure of a lot of the modern computer games in terms of how you would put it together and how you would use AI facilities and puzzles and that sort of thing to, try to to constrain the flow of actions and resolutions to be exciting and it reads pretty well today. It was remarkable so this was one of the good papers. There’s some bad papers too. The impact of the computing industry however was pretty significant and continues to be very significant. It’s new question whether the AI world is going to be as significant or whether is going to simply eliminate any thinking at the level at which we currently applied to solve problems. And I don’t know what the solution to that is yet. But I’m hoping that we’ll figure it out and have an opportunity to think about it.

這些文件可能根本沒有經過仔細審查，有時候甚至是草率地審查，有些文件很難理解，因爲它們似乎毫無意義，而且可能確實如此。其中一些是我和我的朋友兼同事李霍維（不確定）合作的傑出作品。我們在那裏發表了一篇關於我們史詩般的計算機遊戲的小論文，基本上闡述了現代計算機遊戲的結構，以及如何使用人工智能設施和謎題等來限制行動和解決方案的流程，使其更加刺激。今天讀起來仍然很不錯。這是一篇很出色的論文。當然也有一些糟糕的論文。然而，計算器行業的影響相當大，並且仍然非常重要。現在的問題是人工智能世界是否會像計算器行業一樣重要，還是隻會消除我們目前應用於解決問題的思考層次。我還不知道這個問題的解決方案是什麼，但我希望我們能找到答案並有機會思考它。

We’ve been now watching the the rise of artificial intelligence. And it’s not computation in the sense that we do it and it doesn’t require the level of analysis that computation requires. Computer science requires to understand problems. It also has sort of opened up the options of using artificial intelligence or at least the artificial intelligence techniques for other purposes. And we certainly see the use of some of the algorithmic artificial intelligence models to do things which are perhaps contrary to our social worries, which are release to the counterculture. People like to be sure that people tell the truth. Thanks they don’t makeup strange stories, they don’t align people, they don’t say bad things about one racial minority or sexual minority or majority or whatever. And that’s a ongoing problem right now. And some parts of the world are not doing terribly well in terms of being the desires of the other parts of the world. We have China in its great firewall and we have people in the Europe being unhappy about privacy. And we have people in the US being unhappy about almost anything and things are very fluid at the moment. And in the long run the people my age tend to want to see the countercultural ideologies effectively dominated some point. And we want people to be fair and honest who take care of business in a reasonable way and not force people to, behave in ways that they don’t not want to behave and things like that. But we’re just part of the crowd, not necessarily driving people’s choices. That’s an ongoing problem. And when you see some of the scams that happen, it’s really you deserving. Right now, it’s a war out there on the internet.

我們現在一直在觀察人工智能的崛起。這不是我們所理解的計算，也不需要計算所需的分析水平。計算器科學需要理解問題。它也開啓了使用人工智能或至少人工智能技術的其他選擇。我們確實看到一些算法人工智能模型被用於做一些可能與我們的社會擔憂相反的事情，這些事情與反主流文化有關。人們喜歡確保別人講真話。感謝他們不會編造奇怪的故事，不會對某個種族少數或性少數或多數說壞話。這是一個當前持續存在的問題。世界上的某些地方在滿足其他地方的期望方面表現得不太好。中國有其偉大的防火牆，歐洲的人們對隱私感到不滿，美國的人們對幾乎任何事情都不滿意，目前情況非常不穩定。 在長遠來看，我們這個年代的人傾向於希望看到反文化意識形態在某個時刻有效地占主導地位。我們希望那些以合理方式處理業務並且公平誠實的人，而不是強迫別人以他們不想要的方式行事等等。但我們只是人羣中的一部分，並不一定能左右他人的選擇。這是一個持續存在的問題。當你看到一些詐騙事件發生時，你會覺得真的很應該。現在，網絡上正進行着一場戰爭。

Cuiyu: Actually, we do curious about your opinions about the shifting of the environment progress since you’re keep working in this area for decades. What do you think is the most obviously shifting of the environment? What kind of the difficulties you’re faced before and later?

翠玉：其實，我們對於您在這個領域工作了幾十年，對環境進展的轉變有什麼看法感到好奇。您認爲環境最明顯的轉變是什麼？您之前和之後面臨了哪些困難？

Dennis Allison: Well, I’m a bad person to ask that question because I’ve always thought that people need education or they need to be curious and intellectually honest to themselves and so forth. I like people who think for themselves. I don’t necessarily feel comfortable with just accepting what’s there. You wanna ask questions about it and so forth. So, I think I’m an outlier there. I think the vast majority of people how work within the context of what they see today. And they don’t, they don’t deal with flaxity. Do you know the story about Black Swans? You can tell, you can agree with that. I can just repeat. Black swans are well, up until they were discovered in Australia, it was believed that all swans were white. Once the black swans were found, of course, that was significant and that just turned everything over. The same thing is true of a lot of things, that there’s a commonly agreed upon state of the world. And once that state of the world is toppled and shown to the not the only state, things get to be very complicated. And maybe that would make you worse. I think that’s a situation that we’re in today and I think that’s also the situation we were in the 60s and 70s, in the early days when counterculture was trying to create a new environment and that new environment was different, very different from the older environment and was adapted to a certain extent and not adopted otherwise. And that’s sort of what’s happening now in terms of things like AI, something’s occasionally yourself losing some of the AI today, for example, I have the problem that the training I know probably not the right place to go into how this works. Training the AI with things that came that were generated by AI seems to have some problems because it introduces another set of biases, which have a different set of statistics. And the inferences that are made from the data from that came from AIs, through AIs may or may not reflect the reality that you want to have reflected. And of course, it’s hard to interpret how the AI works. We don’t know. it’s what it learns and how it selects what it learns is still a matter of prime question in progress. So it’s a fascinating thing to watch how this goes. One of the sad things is that, the current best model for artificial intelligence, the larger language models, do require substantial training and the training needs to be over a very large sec. You can make the set somewhat smaller by making certain choices, but that makes some parts of the process less effective. And it’s hard for an individual or a small group to to do anything which is very different from what everybody else is doing in terms of artificial intelligence. It’s a small deviation from a trained you can take a trained set of weights and you can move it a little bit by adding some additional information. You can move the pit though. It ends up introducing any instabilities of various sorts. And so everybody is kind of stuck with what a large group has done and released into the public.

丹尼斯·艾里森：嗯，我不是一個合適的人來回答這個問題，因爲我一直認爲人們需要教育，或者他們需要對自己保持好奇心和知識誠實。我喜歡那些能夠獨立思考的人。我並不一定對接受現狀感到舒適。你想要對此提出問題等等。所以，我認爲我在這方面是個例外。我認爲絕大多數人只是按照他們所見到的現狀工作，並不會去處理複雜性。你知道關於黑天鵝的故事嗎？你可以告訴，看是否同意這個觀點。我可以重複一下。在發現澳大利亞的黑天鵝之前，人們一直認爲天鵝都是白色的。一旦發現了黑天鵝，當然，這是非常重要的，它徹底改變了一切。同樣的事情也發生在很多其他事物上，即世界的狀態是被普遍認可的。一旦這種狀態被推翻，並且顯示出不是唯一的狀態，事情就變得非常複雜。也許這會讓你變得更糟糕。 我認爲這是我們今天所處的情況，也是60年代和70年代我們所處的情況，當時反文化正在試圖創造一個新的環境，這個新環境與舊環境非常不同，在某種程度上進行了適應但沒有完全接受。這就是現在在人工智能等方面正在發生的情況，例如，今天有時候你自己會失去一些人工智能，我有一個問題，我知道這可能不是討論它如何工作的正確場所。用由人工智能生成的東西來訓練人工智能似乎存在一些問題，因爲它引入了另一組偏見，這些偏見具有不同的統計數據。從來自人工智能的數據中進行的推論，可能或可能不反映您希望反映的現實。當然，解釋人工智能的工作方式很困難。我們不知道它學到了什麼，以及它如何選擇學到的東西仍然是一個重要的問題。所以看着這個過程是一件迷人的事情。 其中一個令人傷感的事情是，目前人工智能的最佳模型，即更大的語言模型，確實需要大量的訓練，而且訓練需要在非常大的範圍內進行。你可以通過做出某些選擇來使訓練集變得稍微小一些，但這會使過程的某些部分變得不那麼有效。對於個人或小團體來說，很難在人工智能方面做出與其他人非常不同的事情。這只是對訓練過的一個小偏差，你可以通過添加一些額外的信息來稍微調整一下訓練集的權重。你可以稍微移動小圈，但這最終會引入各種不穩定性。所以每個人都被困在一個大團體所做並公開發布的東西中。

Well, you think that would be okay. Sorry, I’ll use the Google or the Facebook (unclear) and I can modify it a little bit. There are places I can plug some things in. It’ll be good. Except for the fact that some person has a radically new idea, much better than this currently around.

嗯，你覺得那樣應該沒問題。抱歉，我會使用谷歌或臉書（不確定）並稍作修改。有些地方我可以加入一些東西。這樣就好了。除非有人有一個完全新的想法，比現在這個好得多。

Well, what do they do? You don’t have the resources to take the very large problem with half a trillion or billions or tens of billions or whatever it is, variables and do a set of weights based on it. It makes being large as an organization and being rich as an organization are much more powerful place to be than being small and smart. And I think in the long run that’s a bad choice. A large organizations which are please say only ideas, you know, concepts, are not necessarily the right ones to go with for the future. He wants something that’s gonna behave in a way which is substantially better then you currently have. And substantially better is measured in terms of your scale of what’s important and what’s not, and not what’s the large organization. So we see political organizations, for example, doing things which are irrational invest and they think it’s perfectly good for them.

他們做什麼呢？你沒有資源來解決一個非常龐大的問題，涉及數萬億、數十億或數百億個變量，並基於此進行權重設定。這使得作爲一個龐大的組織和富有的組織比作爲一個小而聰明的組織更具有優勢。但我認爲從長遠來看，這是一個錯誤的選擇。那些只提出想法、概念的大型組織未必是未來的正確選擇。他希望找到一種行爲方式，比你目前擁有的要好得多。而"好得多"是根據你所認爲重要與不重要的尺度來衡量的，而不是根據組織的大小。所以我們看到政治組織做出了一些非理性的投資，他們卻認爲這對他們來說是完全合適的。

But I tend to not. No, I’m worried about the other. But I expect this not good for everybody. And that’s, that was true in the personal computing business as well. Having a personal computer was important. And access to a personal computer is important because you could check and see, and you could experiment, and you can look at, find out how world works and so forth. Of course, the personal computers we had access to were all small, had too little memory, no disk and forth and so on. But now you can get a personal computer that’s well, my phone is, has more computing power than all of the university in 1960. That’s quite impressive. I said this, I’m still impressed. We didn’t talk about some of the other things that I’ve done by the way. I design computers, have designed prices for chips. design instruction, I’ve design compilers, design semiconductor processes, conduct group tools for making semiconductors. Well, things of that deal.

但我不太這樣做。不，我擔心其他的事情。但我預計這對每個人都不好。這在個人計算機業務中也是如此。擁有一臺個人計算機很重要。而且擁有一臺個人計算機很重要，因爲你可以檢查和觀察，你可以進行實驗，你可以查看，找出世界是如何運作的等等。當然，我們能夠接觸到的個人計算機都很小，記憶容量太小，沒有磁盤等等。但現在你可以獲得一臺個人計算機，我的手機的計算能力超過了1960年的整個大學。這相當令人印象深刻。我說過這句話，我仍然感到印象深刻。順便說一下，我們沒有談論我做過的其他事情。我設計計算機，設計芯片價格，設計指令，設計半導體工藝，開展製造半導體的工具等等。那些都是我所從事的事情。

So, one of the things that the personal computer gave to me was the opportunity to do all of these other things, which are important and significant and exciting. I’m still doing today, even though I’m at past retirement age. Well, my pull dates haven’t reached. And so I’m still (unclear).

所以，個人計算機給了我一個機會去做其他重要、有意義和令人興奮的事情。即使我已經超過退休年齡，我現在仍然在做這些事情。嗯，我的過期日期還沒到。所以我還在(不清楚)。

Cuiyu: Okay, I have a question about the computing power like you mentioned, that today the phone we have has more computing power than all the stuffs in 1970s. And it is also very interesting to see that people most people in today we have phones, but we don’t use it to computing. I was wondering what is the situation in 1973, is the user more more willing to compute with their machine? Maybe it’s the difference between generations. I don’t know.

翠玉：好的，我有個關於你提到的計算能力的問題，就是現在我們手機的計算能力比1970年代的所有東西加起來還要強大。而且有趣的是，今天大多數人都有手機，但我們並不用它來進行計算。我想知道1973年的情況如何，當時的使用者是否更願意用他們的機器進行計算？也許這是世代之間的差異。我不知道。

Dennis Allison: There are big differences. In the early 70s, the machines didn’t do much at all. The ad to track multiple and divide. They ran basic BASIC. Well, Tiny BASIC had 26 variables only. And so you could have an a, you could a, B, C, d could have a value and EF and G could have a string value. And that was it and that was it. Today the see phone we have has huge amount of memory. And yes, it’s not. It’s so easy to do things today that use memory and computation to do what seems really trivial, like, I don’t know, we do speech recognition. It’s a surprise when it’s not right these days. Whereas in 1973, if you knew there was speech there, it was a great success. It wasn’t. It’s just really very different. Lots more computation, lots more understanding of how the relationships of pieces of computations are with other pieces. Their ways of getting writing programs that don’t really write any programs at all that use components that are bigger than program module so that you can do computations simply. It’s a totally different business.

丹尼斯·艾里森：有很大的差異。在70年代初，機器幾乎什麼都做不了。廣告追蹤多個並進行分割。它們運行基本的BASIC。嗯，簡化版BASIC只有26個變量。所以你可以有一個a，你可以有一個b，c，d可以有一個值，EF和G可以有一個字符串值。就是這樣，就是這樣。而今天我們所擁有的手機擁有巨大的內存。是的，這並不容易。現在做一些看似微不足道的事情，如語音識別，使用內存和計算來完成，這是非常容易的。如果現在語音識別出錯，那纔是驚訝的。而在1973年，如果你知道那裏有語音，那就是一個巨大的成功。這真的非常不同。有更多的計算，更多對計算組成部分之間關係的理解。有一種編寫程序的方式，實際上根本不需要編寫任何程序，而是使用比程序模塊更大的組件，以便簡單地進行計算。這是完全不同的業務。

Now, the level at which I tend, have personally tended to work is very small. I like to work at a very low hardware level and make programs that do very difficult things with pieces of physical hardware. And I think that’s sort of exciting. Most people don’t get to that level at all. They’re quite happy to use their phone and talking to the phone and say, call Dennis Allison and the phone says. And then I’m on the other end of the phone and calls me may. That involves an enormous amount of software and very complex software, most of which is pretty well understood these days, some of it is not. And part of it is the metric of translation of ideas into text in some fashion, for sound and text. And that’s what drives a lot of the problems with AIs that you, you have to translate from the real world, which is sound and pressures and so forth, to some sort of internal representation, which then gets translated, to translated until you understand what is really given. And then you back it out. And that’s it.

現在，我個人傾向於從事非常細微的工作。我喜歡在非常低的硬件層面上工作，並使用實體硬件的部分來完成非常困難的任務。我認爲這種工作方式相當令人興奮。大多數人根本無法達到這個層次。他們很樂意使用手機，對着手機說：“打給Dennis Allison”，然後手機就會打給我。這涉及大量的軟件和非常複雜的軟件，其中大部分現在已經被很好地理解，但也有一些尚未被理解。其中一部分是將思想轉化爲文字或聲音的方式，這是解決人工智能問題的一個重要指標。你必須將現實世界中的聲音、壓力等轉化爲某種內部表示，然後再進行翻譯，直到你真正理解所給出的內容。然後再反向操作。就是這樣。

I kind of like both. I have always been fascinated by the the ways that you can do a computational in hardware at some level. And so I found doing computations with synergy arithmetic scaled in a signal processing context to be kind of need(unclear). It’s very surprising. It works. And when it works, it works beautifully. And that’s it.

我有點喜歡兩者。我一直對在某種程度上在硬件中進行計算的方式感到着迷。所以我發現在信號處理的背景下使用協同算術進行計算有點有趣。這非常令人驚訝。它有效。而且當它有效時，它的效果非常出色。就是這樣。

Most people don’t get into that level of detail. Some people like to watch show programs work and they look only at the very high level program. So you have a program which will resolve one question and it doesn’t by breaking it up into four other big questions and so forth. And that’s about as far as it has to go. You don’t go down to the very detailed level in order to resolve the question. And that is what more and more programming is to becoming. The concepts are bigger. The programs that you have available to do them are bigger. They operate in a very limited way to kind of cumbersome and dumb and they don’t really help us very much and that’s fine. They’re free and they work, sort of. Really elegant little programs that are very fast and do really interesting things. Well, that’s a harder thing.

大多數人不會深入細節。有些人喜歡觀看節目，只關注非常高層次的節目。所以你有一個能解決一個問題的節目，它不會將問題分解成其他四個大問題等等。這就是它需要達到的程度。你不需要深入細節來解決問題。這就是越來越多的編程所要達到的。概念更大了。可用於執行它們的程序也更大了。它們以一種非常有限的方式運作，有些笨拙，並且並不真正幫助我們很多，但這沒關係。它們是免費的，而且它們能夠運作，某種程度上。真正優雅且非常快速且能夠執行有趣事物的小程序，那就更難了。

I was these days with some people at a company called Maxler Technologies. Maxler is a subsidiary of Rock, which is a fabulous semiconductor house but Maxler is the advanced technology and Cassie and what Maxwell does is find problems that are hard to solve and then solve them.

It’s one of the things that Maxler has done remarkably well. It find ways to get data onto a chip in a way that’s very fast and that’s hard. It’s a lot harder than people think. And as a result, Maxler has built some of the world’s fastest machines. Fast machines are not all that useful sometimes because you don’t need a lot of them. You just need one of them. And when you only make one of something in the semiconductor business, you’re in trouble. You wanna make lots of them. But anyhow, it’s a fascinating thing because the idea is that you want to find isolated new blocks of functionality, which you can involve. And it’s not you can construct that out of normal software or you construct it out of some other way of doing it, but you can get the result.

And that’s often much faster and much more efficient in terms of cost and that sort of thing itself. There there’s still a range of things, so just because we have a way of solving a problem does not necessarily mean the way we’ve to solve it is the right thing to use.

我最近和一些人在一家叫做Maxler Technologies的公司待了幾天。Maxler是Rock的子公司，Rock是一家非常出色的半導體公司，而Maxler則是專注於先進技術和Cassie。Maxwell的任務是找到難以解決的問題並解決它們，這是Maxler非常擅長的事情之一。它找到了一種非常快速且困難的方式將數據傳輸到芯片上。這比人們想象的要困難得多。因此，Maxler已經建造了一些世界上最快的機器。快速的機器有時並不是那麼有用，因爲你不需要很多臺，只需要一臺就夠了。而當你在半導體行業只生產一臺產品時，你就會陷入困境。你希望能生產很多臺。但無論如何，這是一件非常有趣的事情，因爲你想要找到孤立的新功能塊，然後進行開發。這不是通過常規軟件構建，也不是通過其他方式構建，但你可以得到結果。從成本和效率的角度來看，這通常更快速和更高效。 雖然有很多事情，但僅僅因爲我們有解決問題的方法，並不意味着我們必須使用的方法就是正確的。

Cuiyu: I actually wanna ask a question. We’ve talked how about your recently publishing, I mean PCC.

翠玉：其實我有個問題想問。我們之前談過你最近的出版物，我指的是PCC。

Dennis Allison: PCC is pretty much not functional at the moment. It was (unclear). PCC sold off various pieces of its operation to commercial publishers for various reasons, usually financial. I mean PCC at one point had a a reasonable financial base at the point at which you’d have the financial base that has already been sold off to a commercial publisher. And the the BASIC journal and its staffing and so forth continued on with the same BASIC philosophies all that time. The range of editors that PCC found for Dr Drobbs for example, we’re all pretty good. I think it went through several iterations where it was sold off from one place to another. Magazine publishing is not a path to richness, but it could be quite stable. So I think the over time, particularly with the rise of the internet, the PCC Journals became less significant because everyone read it online. And advertising and so forth for this particular areas is pretty limited. The number of people who read code and think that they’re interested in how code works in the internals, it’s probably not this large as you would like to have (1:39 unclear) one organization. At one point, the things that were moderately valuable and had a value were sold off to commercial publishers and PCC went on trying to teach people about BASIC and other languages and about computers and so forth and doing new projects. And then people retired and got older. They decided that they didn’t want to be involved. And so new people came in who had new ideas. And again, keeping sort of the same basic thrust for quite different.

Eventually the PCC closed down became that terminal, suspended itself. Under California law not-for-profit needs to divest itself of its assets only by giving it to another not-for-profit. And currently PCC is in a state of suspension waiting to decide whether it wants to give away its right or to stay in a state of suspension.

Dennis Allison：PCC目前基本上無法正常運作。它（不清楚）了。PCC出售了其運營的各個部分給商業出版商，原因各不相同，通常是出於經濟考慮。我的意思是，PCC曾經有一個相對合理的財務基礎，但這個財務基礎已經被出售給商業出版商。而BASIC期刊及其人員等一直秉持着相同的BASIC理念。PCC爲Dr Drobbs找到的編輯範圍，都相當不錯。我認爲它經歷了幾次從一個地方出售到另一個地方的迭代。雜誌出版並不是致富之路，但它可能相當穩定。所以我認爲隨着互聯網的興起，隨着時間的推移，PCC期刊變得不那麼重要，因爲每個人都在在線閱讀。而且這個特定領域的廣告等相當有限。閱讀代碼並對代碼的內部運作感興趣的人數，可能不像你希望的那麼多，以至於只有一個組織。 在某個時候，那些有一定價值的東西被賣給了商業出版商，PCC繼續努力教授人們BASIC和其他編程語言以及計算機等方面的知識，並進行新的項目。然後人們退休了，變老了。他們決定不再參與其中。於是新的人進來了，他們有新的想法。儘管如此，PCC仍然保持着相同的基本動力，只是方向有所不同。

最終，PCC關閉了，變成了一個終端，暫時停止了運作。根據加利福尼亞州的法律，非營利組織只能通過將其資產轉讓給另一個非營利組織來剝離自身的資產。目前，PCC處於暫停狀態，等待決定是要放棄其權利還是繼續保持暫停狀態。

And I don’t know exactly where it stands now. We have had plans to to restart it because we thought there was a place where PCC could be useful for this problem of the digital chasm into which our youth are falling, and perhaps we could find ways to improve that. Now with the rise of AI, rise starting probably not too long ago.

我不確切知道它現在處於什麼狀態。我們曾計劃重新啓動它，因爲我們認爲PCC在解決我們的青年陷入數字鴻溝這個問題上可能有用，也許我們可以找到改善的方法。現在隨着人工智能的崛起，這個崛起可能不久前就開始了。

There are other reasons for wanting to restart PCC and see if there’s a good way to do that. There’s a lot of interest in AI as a computational search tool. It would be nice if there were some center focal place where people interested in that particular aspect of computation could find information. It’s interesting and to France (unclear) There’s nothing really that is entirely outside of the proprietary secret key thing of governance and companies.

And so people who want to learn about it either have to be involved with the government or involved with the companies. And then if they’re involved companies they can’t talk to other people and other companies and so forth. Except that they publish journal articles, but that’s not some pattern things to the record. It’s not an opened intellectual climate. I personally favor in opening intellectual climate so like a personal computer area. I think this is a time when it would be suitable to have people out there doing interesting things. And the only way that happens is if they get the information they need in their hands and access to the tools that are needed in order to use that information, which sort of says that there’s a need for a large scale computation.

有其他原因想要重新啓動PCC並看看是否有好的方法可以做到這一點。人們對AI作爲計算搜索工具的興趣很大。如果有一箇中心焦點的地方，讓對計算這個特定方面感興趣的人可以找到相關信息，那將是很好的。這對法國來說很有趣（不清楚）。除了專有的祕密密鑰治理和公司之外，沒有真正完全獨立的東西。所以想要了解它的人要麼必須與政府有關，要麼必須與公司有關。而且如果他們與公司有關，他們不能與其他人和其他公司交流。除了發表期刊文章，但這並不是一個開放的知識氛圍。我個人偏好開放的知識氛圍，就像個人計算機領域一樣。我認爲現在是一個適合讓人們做有趣事情的時候。 唯一的方式就是讓他們手上拿到所需的信息，並獲得使用該信息所需的工具，這也意味着需要進行大規模的計算。

I’m like most computations. Most computations have a problem and a solution and you can make the solution by passing the information and situation and how the thing that you’re examining works to a program which then produces an answer. That’s not true of some of the AIs because in fact you’re solving a very large, broad class of problems at the same time and then having a way of selecting out the answer. It’s not the same. This is different. And to do that you need a large machine and so there’s no market. Actually, there is a market I suspect but most people are not going to have the money necessary to do a full web scan and examination of information.

And so what they do is to find ways to perturb what’s happening now is the new. There are several public domain. Sets of weights for AI, artificial intelligence commune depends on a large set of weights to encode things and some of them are available and you can perturb them within certain range and maintain things to be still stable. But if they are still part of proprietary and you just need to stay with the proprietary stuff and you have to pay the price and do it. Whereas we currently in a computing area, people just write programs, because they can go down to the local computer store and buy computer that has a lot of power, not enough power to do AI, but power to execute AI.

我和大多數計算一樣。大多數計算都有問題和解決方案，你可以通過將信息、情況和你正在檢查的事物的工作方式傳遞給一個程序來得出解決方案。但這對某些人工智能來說並不成立，因爲實際上你同時解決了一個非常廣泛的問題類別，然後找到一種選擇答案的方法。這是不同的。爲了做到這一點，你需要一臺大型機器，所以市場上並沒有這樣的需求。實際上，我懷疑市場上可能有這樣的需求，但大多數人沒有足夠的資金來進行全面的網絡掃描和信息檢查。

所以他們會尋找方法來干擾現在正在發生的事情，這是新的。有幾個公共領域的AI權重集，人工智能小區依賴於一個大型的權重集來編碼事物，其中一些是可用的，你可以在一定的範圍內干擾它們並保持穩定。但如果它們仍然是專有的，你只需要使用專有的東西並支付代價。 鑑於我們目前處於計算領域，人們只需編寫程序，因爲他們可以到當地的計算機商店購買一臺功能強大的計算機，雖然不足以執行人工智能，但卻具備執行人工智能所需的能力。

Nice thing. I’m looking out for my own new computer, right? I am looking at some of the computers. I want a big computer now. I would like, you know, is probably resolved the record, but I’m looking, I would like to build one more computer design and build one more computer and you know, I like to build the next biggest computer. And it’s a problem. Is that you. Yes, cost for money.

好東西。我正在尋找自己的新電腦，對吧？我正在看一些電腦。我現在想要一臺大電腦。我想，你知道的，可能解決了記錄，但我正在尋找，我想設計和建造另一臺電腦，你知道的，我喜歡建造下一臺更大的電腦。但問題是，這需要花錢。

Cuiyu: Do you have any and specific plan? Or the specific design. Do you have interest to draft or something?

翠玉：你有沒有具體的計劃？或者具體的設計。你對起草或其他方面有興趣嗎？

Dennis Allison: Well, we have, there are several of us who have some ideas to what we want to do. And those are all, of course, secret and proprietary. Because in theory, the ideas are very valuable. Now, in truth, maybe the ideas are not (unclear), but we’d like to think so. And I think that the need for big computers is well, it may be less than we think. In video it is doing extremely well. The question is the current approach the right one or not? And I think it is eventually dead ends but that’s my impression and I have colleagues who say that’s not the case. They’re wrong, but that’s okay.

Dennis Allison：嗯，我們中有幾個人對我們想做的事情有一些想法。當然，這些想法都是祕密和專有的。因爲理論上來說，這些想法非常有價值。現實上，也許這些想法並不是那麼重要，但我們願意這樣認爲。我認爲對於大型計算機的需求可能沒有我們想象的那麼大。在視頻領域，它表現得非常出色。問題是當前的方法是否正確？我認爲最終是死衚衕，但這只是我的印象，我有一些同事認爲不是這樣。他們錯了，但沒關係。

Cuiyu: I’m still a little bit curious about the big machine. Why do you think that machine should turn big right now? Why should we turn the machine too big? Since in the past we turn machine to smaller right? Because it’s portable.

翠玉：我對這個大機器還有一點好奇。你爲什麼認爲現在應該把機器做得更大呢？爲什麼我們要把機器做得太大呢？過去我們不是把機器做得更小嗎？因爲它是可攜帶的。

Dennis Allison: Oh, portable is the sort of problem I’m thinking about. The problem is that you want to part the machine to very small so you can put them all very close together in a small space, because you have so many of them. And the problem is that, there’s a maximum size that you can make the machine. Speed of light determines how big the machine can be and how you can construct things, because you can have two separate things that are separated by some distance, but they have to communicate, and then the communications rate then becomes dominant factor. But ignoring that, how you build a machine is usually constrained primarily by how closely you can pack things, like how well you can take the heat out, and how rapidly you can communicate point to point. And that’s good.

丹尼斯·艾里森：哦，便攜性是我正在思考的問題。問題是，你希望將機器分成非常小的部分，這樣你可以將它們都放在一個小空間裏非常靠近，因爲你有很多這樣的機器。問題是，有一個最大尺寸限制了機器的大小和構造方式，因爲光速決定了機器的尺寸，以及你如何構建它們，因爲你可以有兩個相隔一段距離的獨立物體，但它們必須進行通信，而通信速率則成爲主要因素。但是忽略這一點，你如何構建一臺機器通常主要受到你能夠多麼緊密地打包物體的限制，比如你能夠有效地散熱以及你能夠多快地進行點對點通信。這是很好的。

And of course, as you may curious compromises building this kind of machine. You lose the ability you do it by saying, well, you can only have 10 of those and this has to be within three inches and so forth, that sort of thing. Actually, we should say 6 cm, 9 cm, but that’s what constrains things. And if you get more than that, then it introduces delays, which makes the machine run slowly. And then the other thing is that the way you get machines to run quickly is you pipeline them. So you you take a long time to execute a single instructions but you do those singles instructions at each step in the instruction is very short and so it goes very quickly. That’s fine, as long as the problem works that way. But if the program is not well structured in that regard, then you do the parallel part up to a point and then you can’t go any further because you’re waiting for another part to the of the computation to be finished so you can join the computation result together and let it proceed on. So that gets to the very complicated.

當然，你可能會好奇建造這種機器的妥協。你失去了用語言來做的能力，你只能說，你只能有10個這樣的東西，而且這些東西必須在三英寸內，等等，就是這樣的事情。實際上，我們應該說6釐米、9釐米，但這就是限制的事情。如果你超過了這個限制，就會引入延遲，使機器運行緩慢。另一個方法是讓機器快速運行的方式是將它們進行流水線處理。所以你需要花很長時間來執行一個單一的指令，但你在每個步驟中執行這些單一的指令的時間很短，所以速度很快。這是沒問題的，只要問題是這樣工作的。但如果程序在這方面結構不良，那麼你只能進行到一定程度的並行處理，然後就無法再繼續下去，因爲你正在等待計算的另一部分完成，以便將計算結果合併在一起並繼續進行。所以這變得非常複雜。

Cuiyu: I think I have already went through all the question list all the questions. Personal I’m really wondering how do you prefer to describe yourself? As a educator, engineer, designer or? You do a lot of things, right?

翠玉：我想我已經看過所有的問題清單了。個人很好奇，你更喜歡如何描述自己呢？作爲教育工作者、工程師、設計師還是其他身份？你做了很多事情，對吧？

Dennis Allison: I do all those things. I think I’m a generalist. Like they do everything. Actually we all aspire to be Renaissance men or women. I don’t know. I do have a friend who says I’m a dilemma. Gotta say I do a little bit of all the things, but I don’t do anything very well. I don’t think that’s really true. I have some credentials as a futurist. So that what I’ve done in the last few years has often been to look at technology and society and the economic structure and try to see where I think things are going to go. That’s very nice because it’s a very useful piece of information if you’re right. And if you’re wrong, it’s not very useful at all. There are a few instances where I have been very right and I tell everybody about those. But there are lots of times I’ve gotten just the wrong guys. And so we don’t talk about those. One of my friends suggests that if you’re a futurist, what you do is you always wanna predict for 25 years out or 15 years out, one of the two the 15 is good because if you’re right and they remember what you said, or at least you have a piece of paper that you can show, and you could say you wrote this and you’re a hero. And of course you don’t show anybody the papers that don’t match the reality of this is happened. You’ve done some futurist report and sent several of these issues and these days it’s a little harder, but you can come pretty close just by thinking that next year is gonna be very close to last year. And the small changes rarely make huge changes and what you’re really looking for is a small change which can get larger very quickly, and has a good reason for for being important. so personal computing was clearly going to be a big change because it means that, well, I don’t think, but the personal computing guys didn’t get it right. I thought everyone is gonna sit around a program and use programs and so forth. and what what’s happen is there are a few programs that are very being incomprehensive, say WhatsApp or something like that. And everybody uses them and they facilitate other things and the program itself is not all that important. And the hardware itself, if it’s fast enough to be regional, it’s not that important either. And that’s significant point in means that a lot of things that people think are important probably are. When we were inventing personal computers and being interested in Tiny BASIC and getting everyone to program, we really were addressing the sorts of issues that we might wanted to address. We were kind of learning how to write with a pencil or to eat food with a fork and knife is supposed to chopstick and things like that. The thing which is important is how to use the computation tool has a subset of the way in which you think and you want to use the thinking process to sort out the ideas and what’s important, what’s not.

丹尼斯·艾里森：我做所有那些事情。我認爲我是一個通才。就像他們什麼都做一樣。實際上，我們都渴望成爲文藝復興時期的人。我不知道。我有一個朋友說我是個困境。不得不說，我做了一點點所有的事情，但是我沒有做好任何一件事。我不認爲這是真的。我有一些未來學家的資質。所以在過去幾年裏，我經常研究技術、社會和經濟結構，試圖看清事物的發展方向。如果你預測正確，這是非常有用的信息。如果你預測錯誤，那就沒有什麼用了。有一些情況下，我預測得非常準確，我會告訴每個人。但也有很多次我預測錯了。所以我們不談論那些情況。我的一個朋友建議，如果你是一個未來學家，你總是要預測25年或15年後的情況，15年是個好時間，因爲如果你預測正確，他們會記住你說過的話，或者至少你有一張可以展示的紙，你可以說你寫了這個，你是個英雄。 當然，你不會向任何人展示與現實不符的文件。你做了一些未來主義的報告，併發送了幾個這樣的問題，這些天變得有點困難，但你可以通過思考明年將非常接近去年來接近實際情況。小的變化很少會帶來巨大的變化，你真正在尋找的是一個可以迅速變大並且有很好的理由成爲重要的小變化。所以個人計算顯然會是一個重大變革，因爲這意味着，嗯，我不這麼認爲，但個人計算的人們沒有做對。我以爲每個人都會坐在一起編程和使用程序等等。而實際上發生的是有一些非常不可或缺的程序，比如WhatsApp之類的。每個人都在使用它們，它們促進了其他事情，而程序本身並不是那麼重要。而硬件本身，如果足夠快速以區域性的方式運行，也不是那麼重要。這是一個重要的觀點，意味着很多人認爲重要的事情可能並不重要。 當我們在發明個人計算機、對Tiny BASIC感興趣並鼓勵每個人進行編程時，我們真的在處理我們想要處理的問題。我們有點像在學習如何用鉛筆寫字，或者用刀叉喫飯，而不是用筷子之類的東西。重要的是如何使用計算工具，它是你思考方式的一個子集，你希望用思考過程來整理想法，找出重要的和不重要的。

Huang: So do you have anything to add? If there’s nothing, maybe see if you wanna draw a conclusion.

黃：那麼，你有什麼要補充的嗎？如果沒有的話，也許可以看看你是做出結論。

Dennis Allison: Well, I don’t really have to draw any conclusions. I can say that knowing what the answer or the current state is, looking backwards to personal computing, and knowing a little bit about how cultures evolve and our thinking is involved. If I had known now or what I know now, at the point at which we started, I think I would have done, things may be a little differently, but I don’t think I would make all that much change. I think that a lot of this is stuff that we don’t have a whole lot of control. You sort of do it and what I do is, what I saw as being important and significant for all these other people or you’re doing the same thing and it’s the general culture of the time, that determines what wins and what’s loses. I don’t think we have a lot of control over that. I think that there are some great flaws in some of the thinking about computation and connectivity and communications these days I guess. Lots of emails——that’s very large letters, LOTS of emails. It’s a huge amount. It is a problem. It’s hard to maintain focus communications with 1000 people a day or 500 people a day. So I think that’s a problem. I think that the rate of which organizations wish to control access to information is a problem. I think that inevitably if you try to say, well that information is okay, but this stuff over here is not, inevitably leads to problems no matter what. And so you have two choices, you don’t let anybody see any information, or you let them see everything. And not seeing any information is kind of a bad one too. So I think the sort of general open open access is extremely important to have things work.

Dennis Allison：嗯，我其實不需要得出任何結論。我可以說，瞭解答案或當前狀態，回顧個人計算的歷史，並瞭解一點文化演變和我們思維的參與。如果我現在知道的東西，或者在我們開始的時候知道的東西，我想我可能會做一些不同的事情，但我不認爲我會做出太大的改變。我認爲很多事情我們並沒有太多的控制權。你只是去做，而我所做的是，我認爲對於其他人來說重要和有意義的事情，或者你也在做同樣的事情，而那個時代的普遍文化決定了什麼會成功，什麼會失敗。我不認爲我們對此有太多的控制權。我認爲在當今關於計算、連接和通信的思考中存在一些嚴重的缺陷，我猜是因爲有太多的電子郵件——那是非常多的信件，非常多的電子郵件。這是一個巨大的數量。這是一個問題。每天與1000人或500人保持專注的溝通是困難的。所以我認爲這是一個問題。 我認爲組織希望控制信息存取的速度是一個問題。我認爲，無論如何，如果你試圖說，這些信息是可以的，但這邊的東西不行，無論如何都會引起問題。所以你有兩個選擇，要麼不讓任何人看到任何信息，要麼讓他們看到所有的東西。而不看任何信息也是一個不好的選擇。所以我認爲普遍的開放存取非常重要，這樣事情才能運作。

On the other hand, if you allow arbitrary open action and you have to sort of sort through it all, you don’t have any time to do anything else. And so that’s not a good thing either. So you’d really like people to self censor more than they do, but to do it not because of some known thing. I look at the privacy laws that are coming up in Europe, for example, CC understand why the privacy laws kind of why it would be really nice to have good privacy, except for the fact that it’s going to cause more problems than its worth. For example, Google has said well, you can’t talk about these people. Well, that’s fine except that finding out whether a particular article which mentions a name which is identical to or similar to a name that you’re not supposed to talk about, whether that’s a thing you can talk about or you can’t talk about, is a prodigious task. That’s right. It doesn’t work. And one of the laws about this are being prepared by people who are interested in the end results and have no skills figuring out the middle part. And they say, well, you can’t mention that name. Hands and names are overloaded and therefore, well, you don’t know whether this included who’s being referenced. I don’t know. It becomes increasingly complicated to have a clear and right understanding of how things are done. And the other thing is to spend a large amount of time and effort now on things like internet and even internet history is probably not important. We have other existential problems that are really significant. You know, the the planet is under siege by a number of things. And in some sense, the history of personal computing and counterculture is less important than what the temperature is going to be in five years. We’re not putting a lot of effort into areas that we perhaps should. And again, actually, that’s an interesting point to. Cuz come back to the counterculture thing. If you think of the global warming ss an issue. Well, stop that. Like I said that, I guess I don’t believe, I don’t even believe that. Let me think if I get this, my thought sorted out. Well, have to sort of decide what is important in the space of we can accomplish.nI don’t think we’re set up yet to do that this time. We need to think more about it, what we’re going to do. But we have less and less time in which to do it. We should have started and 1750 (unclear). 1900 was probably an absolute late start and waiting until 2023 or whatever is kind of disaster. You don’t have good stuff. On the other hand, we will have our personal computers and we can look at statistics and understanding. That’s actually not a bad idea. And so even the simple computations allow you to see things that you would not see without the computation. Simple models actually work pretty well for most things. The reason they’re simple is that they deal with only the things that are really important. Okay, does that come close to answering your question?

另一方面，如果你允許任意開放的行爲，而你必須對其進行排序，那麼你就沒有時間做其他事情了。這樣也不是一件好事。所以你真的希望人們能夠自我審查得更多，但不是因爲某種已知的原因。例如，我看到在歐洲出現的隱私法，我明白爲什麼隱私法很重要，但它可能會帶來比它所值得的更多問題。例如，谷歌曾經說過，你不能談論這些人。這沒問題，但是要找出一篇特定文章是否提到了一個與你不應該談論的名字相同或相似的名字，這是一項艱鉅的任務。沒錯，這行不通。而這些法律的制定者對最終結果感興趣，卻沒有能力解決其中的問題。他們說，你不能提到那個名字。 手和名字都被過度使用，所以，嗯，你不知道這包括了誰。我不知道。越來越難清楚且正確地理解事情是如何進行的。另一件事是現在花費大量時間和精力在像互聯網甚至互聯網歷史這樣的事情上可能並不重要。我們有其他真正重大的存在問題。你知道，地球正受到多種威脅。在某種程度上，個人計算和反文化的歷史比未來五年的溫度更不重要。我們並沒有在我們可能應該的領域投入太多努力。而且，實際上，這也是一個有趣的觀點。回到反文化的事情上來。如果你把全球變暖當作一個問題。好吧，停止吧。就像我說的那樣，我想我不相信，我甚至不相信那個。讓我想一想，如果我能整理出我的想法。嗯，我們必須決定在我們可以實現的範圍內什麼是重要的。我認爲我們還沒有準備好這次做到這一點。我們需要更多地思考我們將要做什麼。 但是我們越來越沒有時間去做這件事。我們應該在1750年就開始了。1900年可能是一個絕對的晚開始，等到2023年或者其他時間就有點災難了。你沒有好東西。另一方面，我們將擁有個人電腦，可以查看統計數據並理解。這實際上不是一個壞主意。因此，即使是簡單的計算也能讓你看到你在沒有計算的情況下看不到的東西。對於大多數事情來說，簡單的模型實際上效果很好。它們之所以簡單，是因爲它們只處理真正重要的事情。好吧，這是否接近回答你的問題了？

Huang: Yes, it’s very true. I think it’s quite good answer true. What should we do today or from now?

黃：是的，這是非常正確的。我認爲這是一個很好的答案。今天或從現在開始，我們應該做些什麼呢？

Dennis Allison: One of the things you do is when you’re looking into predictive is to build models of how things work and how they interact. And the models which we have, say, it was really thing important that we start a long time ago because it takes a very long time to impact what we’re doing. You know, this 6000 kilometer sphere in which we live, we’ve heated it up just a little bit, but it has major impacts on how it behaves. And we didn’t think about it when we started cutting down forests in Greece in 300 BC and that that’s caused this problem.

丹尼斯·艾里森：你所做的其中一件事情是，在進行預測時建立事物運作和互動的模型。而我們所擁有的模型，說實在的，很重要的一點是我們很久以前就應該開始了，因爲要對我們所做的事情產生影響需要很長的時間。你知道，我們生活在這個直徑6000公里的球體中，我們只是稍微加熱了一點點，但這對它的行爲有着重大的影響。而當我們在公元前300年開始砍伐希臘的森林時，我們並沒有考慮到這個問題。

And so, questions do you believe the simulations because they look at very small fluctuations and they say, well, it looks like if you continue to do this, these things are going to happen. And pretty soon you have predictions of disaster. Now sometimes people say, that’s not the case. We’ll just have to wait and see. Well, that’s not a good choice. You can say, well, let’s see if it continues. And if it continues, we’ll do something about it. And that’s probably also not a good choice because it says that you’re pushing the time. You begin to make efforts, help further. Well, what do you do? and how do you arrange it so that there’s enough political and intellectual energy expanded on solving the problem, that it will get solved–if it’s possible to solve. And by the way, there times after which you can’t solve it. And the arguments that some people make, I guess, I mean, I get it. Is the situation we have now is probably one where we’ve passed the point where we can actually fix it, and we can just simply mitigate the problem. And having a computer around to do that examination and look at the statistics and the data is a useful thing. And personal computers that are used for solving real problems are probably more important now than they were when you could get an apple to (unclear). No, that’s sort of where it is. And that’s a problem worldwide. It’s not a simple problem because things done and (unclear) do affect how things are in Bogota or whatever. I mean it’s a global problem and we’re not protected by the situation. A global problem here is limited. Ebola for example, Ebola is disastrous disease. You get infected, you die with 90% probability. On the other hand, you have an infection of Ebola here. By the way, I think the 90% number may be low. But you have a localized to the bullet infection here. The likelihood that it’s going to affect people in New York City, which is a long way forever. It’s not very high because you don’t have any communication or transfer of bodily fluids. That’s a significant factor. So isolation is important to allow things which are very dangerous to die out. And that’s true of ideas and, and people as well. But in any case, helps to have a computer to figure out.

所以，問題是你相信這些模擬嗎？因爲它們觀察到非常微小的波動，然後說，嗯，看起來如果你繼續這樣做，這些事情就會發生。很快，你就會聽到災難的預言。有時候人們會說，那不是事實。我們只需要等待並觀察。嗯，那不是一個好的選擇。你可以說，嗯，讓我們看看它是否持續下去。如果它持續下去，我們會採取措施解決它。但這也可能不是一個好的選擇，因爲這意味着你在拖延時間。你開始努力，進一步提供幫助。那麼，你該怎麼做呢？如何安排足夠的政治和智力資源來解決問題，以便問題能夠得到解決——如果問題是可以解決的話。順便說一句，有時候會有一些情況是無法解決的。有些人提出的論點，我想我明白了。現在我們面臨的情況可能是已經過了我們能夠修復的點，我們只能簡單地緩解問題。有一臺計算機來進行檢查和分析統計數據是一件有用的事情。 而用於解決實際問題的個人計算機，現在可能比你能讓蘋果（不清楚）時更重要。不，這就是問題所在。這是一個全球性的問題，並不簡單，因爲所做的事情和（不清楚）確實會影響波哥大或其他地方的情況。我的意思是這是一個全球性的問題，我們並不受情況的保護。全球性的問題在這裏是有限的。例如，埃博拉，埃博拉是一種災難性的疾病。感染了，有90%的概率死亡。另一方面，你在這裏有一個局部的感染。順便說一下，我認爲90%的數字可能低了。但是你在這裏有一個局部的感染。它會影響到紐約市的人們的可能性非常低，因爲你沒有任何液體的交流或傳播。這是一個重要的因素。因此，隔離是重要的，可以讓非常危險的事物消失。這對於想法和人也是如此。但無論如何，有一臺計算機來解決問題是有幫助的。

And modeling the world is interesting. I have one little more story to tell. My friend (unclear) is a fairly significant hacker type. He was of one of the founders of a company called Cygnus which was bought by Red Hat and the popularized Unix or Linux rather worldwide. And he’s done lots of other things too, but he’s been working on a company he calls Blue Dot Change. And it’s an interesting idea. He wants to burn the methane in the atmosphere. Sounds like a ridiculous thing, but it’s actually very smart. Methane is a very effective greenhouse gas. It’s 80 times worse than carbon dioxide. If you can convert and molecule methane is in the atmosphere to carbon plus oxygen plus water. By burning it, then it operates only as a carbon dioxide molecule and that is a capture molecule. And it’s much less effective. He does a simulation. It’s about, well, at least one degree of heat on the earth if you can do it for all is really nothing. Methane is mostly released by cargo ships. Cargo ships always have a vent at the top for getting rid of the hot air from a moderation sense and you can put a catalyst device into that vent and convert the distribute catalyst that causes the methane, and encounters to oxidize that is burn and it will never impact on global warming. so he’s happier looking for people to pay money to build little gadget that goes into a a cargo ship, and if you get all cargo ships drop the temperature. Okay of the earth by a substantial amount and it’s pretty easy. Now if he didn’t have a computer that let him look at the impacted, making that change, he wouldn’t be able to talk people into doing it. So it’s a very interesting thing. I think the way in which you use the computers can be very very important. And you can put into the hands of people who are relatively unpowerful, normally. You can give them the tools they’re need to make them at least competently argumentative. I think it’s a good thing, right?

而且對世界進行建模是有趣的。我還有一個小故事要講。我的朋友（不清楚名字）是一個相當重要的黑客類型。他是一家名爲Cygnus的公司的創始人之一，該公司被Red Hat收購併在全球推廣了Unix或Linux。他還做了很多其他的事情，但他一直在從事一家他稱之爲Blue Dot Change的公司的工作。這是一個有趣的想法。他想要燃燒大氣中的甲烷。聽起來像是一個荒謬的事情，但實際上非常聰明。甲烷是一種非常有效的溫室氣體。它比二氧化碳嚴重80倍。如果你能將大氣中的甲烷分子轉化爲碳加氧加水，通過燃燒它，那麼它只會作爲一個二氧化碳分子存在，而這是一種捕獲分子。而且它的效果要小得多。他進行了模擬。如果你能爲所有的甲烷做到這一點，對地球來說至少會增加一度的溫度，這實際上並不算什麼。甲烷主要是由貨船排放的。 貨船總是在頂部設有一個通風口，以排除熱空氣，你可以將催化劑裝置放入該通風口，將分散的催化劑轉化爲甲烷，並使其氧化，也就是燃燒，這將不會對全球暖化產生影響。所以他正在尋找願意支付費用來建造一個小裝置的人，這個裝置可以放入貨船中，如果所有貨船都能降低地球溫度，那將是一個相當大的改變，而且這是相當容易實現的。現在，如果他沒有一臺讓他能夠查看影響並進行改變的計算機，他就無法說服人們去做這件事。所以這是一件非常有趣的事情。我認爲你使用計算機的方式非常非常重要。你可以將這些工具交給相對無權無勢的人們，讓他們至少能夠有能力進行辯論。我認爲這是一件好事，對吧？

Huang: Yes. Good. Very good. Thank you. Very interesting stories and very good conclusion.

黃：是的。好的。非常好。謝謝。非常有趣的故事和很好的結論。