Technology · Tech History

The Mother of All Demos

One afternoon in 1968 held almost every idea you now use on a screen, and one idea nobody has managed to build since.

A black and white photograph of a man in his forties in a dark suit and narrow tie, seated at a plain wooden desk. He wears a thin headset with the microphone wire running down his lapel. His right hand is raised, open, mid-explanation. On the desk beside him sits a small pale box trailing a cord. A chalkboard behind him carries the words PLOT BLANK and PLOT SOLID and an arrow.
Douglas Engelbart at the Stanford Research Institute, 1968, wearing the headset and working beside the three-button mouse his lab had built. Source: Wikimedia Commons, where the author and licence terms for this file are recorded.

On the afternoon of 9 December 1968, Douglas Engelbart came out onto the stage of San Francisco’s Civic Auditorium, sat down at a console, and apologised to the room for staying in his chair. He was wearing a headset. Above and behind him a screen twenty-two feet high carried his face on one side and, on the other, the live contents of a display he was driving from a computer thirty miles down the peninsula. The programme listed his session as A Research Center for Augmenting Human Intellect. For the next ninety minutes he ran the research in front of a hall that had run out of seats.

In one continuous session, on one system, the audience watched a cursor steered by a small box under his hand; text edited on the screen and folded into an outline that collapsed and opened on command; windows; a file that reached another file when he clicked a word inside it; a record of who had changed what; and two colleagues in Menlo Park appearing in a corner of the same display, talking to him and editing the same document while he worked. Every one of those has a history of its own. December 1968 is the only date on which they were all shown running together.

The rigging deserves as much credit as the software. The computer was an SDS 940 at the Augmentation Research Center, Engelbart’s lab inside the Stanford Research Institute in Menlo Park. Two homemade modems ran at 1200 baud, fast for the year and each capable of carrying traffic in only one direction, sending keyboard and mouse signals down a leased telephone line while two microwave links brought video back. The picture reached the hall through a borrowed Eidophor, then the only sort of projector that could throw a live computer display large enough for an auditorium. Bill English, the lab’s lead engineer, sat at a switcher deciding what the audience saw, and the camera at the Menlo Park end was worked by Stewart Brand, who knew nothing about computers and was at the time editing the Whole Earth Catalog.

English had built the box under Engelbart’s hand four years earlier, from a sketch: a hollow block of wood, two wheels set at right angles so that one read movement across and the other movement up, and room on the shell for a single button. The patent, filed in June 1967 and granted in November 1970, calls it an X-Y Position Indicator for a Display System. The lab called it a mouse on account of the cord.

A museum display case. A small rectangular block of dark polished wood with a single red button on its upper left corner sits on a stone plinth, trailing a braided silver cable that coils to a green connector. A printed card behind it describes the object.
The 1964 prototype, in a replica held by the Computer History Museum. One button, and a cord that leaves by the front. Source: Wikimedia Commons, where the author and licence terms for this file are recorded.

The mouse is the piece the world kept, and it is the smallest thing in the room that day. The load-bearing word sits in the title of the session. Engelbart had spent the end of the war as a Navy radar technician in the Philippines, where he read Vannevar Bush’s essay As We May Think, published in The Atlantic in July 1945, which imagined a desk that could hold a man’s whole library and remember the trails he cut through it. What Engelbart drew from the years after was a proportion. The problems in front of the species were getting harder faster than the species was getting better at handling them, and closing that gap struck him as work worth a whole career. He set the case out in October 1962 in a report for SRI called Augmenting Human Intellect: A Conceptual Framework, and NASA and ARPA paid for the lab that followed. His team then used its own unfinished system for every hour of its own work, so that each improvement to the system improved the people improving the system. He called that bootstrapping.

He opened the session with a question, and the stumble in the middle of it is still on the tape.

ReadEngelbart, from the floor, 9 December 1968

The research program that I’m going to describe to you is quickly characterizable by saying, if in your office, you as an intellectual worker were supplied with a computer display, backed up by a computer that was alive for you all day and was instantly responsible, responsive…

[Laughter]

Instantly responsive to every action you had, how much value could you derive from that?

Douglas C. Engelbart, opening remarks, Fall Joint Computer Conference, from the Doug Engelbart Institute’s transcript of the demonstration film.

Watch Remastered from the original reels, Stanford MouseSite

Give it the first five minutes, which is all it takes. Watch his left hand, not his right: it never leaves a five-key set he plays like a chord, so that the mouse points while the other hand types commands. That half of the design nobody copied, and it is the reason he could move so fast on a machine with none of the speed of yours.

The hall stood and applauded. Then, on the testimony of the people who were in it, almost nothing happened. Andy van Dam was there, building a hypertext system of his own at Brown and stunned by how far ahead this one was, and it was van Dam who eventually hung the name on the afternoon. Speaking at a hypertext conference in 2019 he gave the flattest possible account of what followed: “Everybody was blown away and thought it was absolutely fantastic and nothing else happened. There was almost no further impact.” The field was still working on teletypes and treated the whole afternoon as a fantasy.

What travelled was the staff. Through the early 1970s most of the Augmentation Research Center went up the road to Xerox PARC, English among them, and in 1972 he threw out his own knife-edge wheels and put a ball in the mouse instead. Bob Taylor, who had funded Engelbart first from NASA and then from ARPA and had pushed him to give the demonstration at all, moved to PARC as well. The Alto appeared there in 1973 with a screen, a mouse and windows, sized for one person at one desk, and that is the line that runs to the Macintosh and to Windows. It carries the half of the idea that fits on a desk.

The other half was the half Engelbart wanted. He was arguing for a trained group living inside a shared system, working on problems none of them could hold alone, and getting measurably better at it year on year. Nobody bought that, and he spent the rest of a long life saying so. So the reputation of the afternoon is upside down. It is filed as the public debut of the mouse, the one piece of it everyone did go on to build. The rest was a proposal about what a computer is for. Introducing the thirtieth-anniversary panel in 1998, the forecaster Paul Saffo observed that the extraordinary thing about the demonstration was how much more of it remained unfinished than had been completed. Another twenty-eight years have not moved that arithmetic very far.