The Mother of All Demos

On December 9, 1968, in a packed auditorium at a computer conference in San Francisco, a soft-spoken engineer named Douglas Engelbart sat down at a custom console, put on a headset, and proceeded to demonstrate, in a single live session, very nearly the entire future of interactive computing.

In about ninety minutes he showed: a mouse (its public debut), moving a cursor on a screen; interactive text editing, cutting and pasting and restructuring a document in real time; hypertext links you could follow; windows dividing the screen; outlining that let you collapse and expand structure; version control of his files; and — most astonishingly — live video collaboration, working on a shared document with a colleague thirty miles away in Menlo Park, their faces and cursors on the same screen.

This was 1968. The room had likely never seen a computer respond instantly to a human at all; computing meant punch cards and printouts and waiting. Engelbart showed them a person thinking with a machine, fluidly, and the recording of it has been known ever since as the Mother of All Demos.

Not a demo of products#

It's tempting to watch the demo as a highlight reel of inventions — look, the mouse! — but that misses what Engelbart thought he was doing. He wasn't showing gadgets. He was showing a working piece of a research program he'd been pursuing for years: augmenting the human intellect.

In 1962, Engelbart had written a dense, visionary report by that name, arguing that humanity's problems were growing more complex faster than our ability to deal with them, and that the way to close the gap was to systematically raise our collective capability to think and solve problems together. He saw intellect not as a thing inside one skull but as a system — humans plus their language, tools, methods, and training, all co-evolving. Improve the tools and methods and you raise the whole system.

The system he built to prove it was called NLS, the oN-Line System, and everything in the demo — the mouse, the editing, the linking, the collaboration — was a component of NLS in service of that single goal: to help groups of people work over hard problems more powerfully than they ever could unaided.

The mouse was the least of it#

The mouse became famous because it was tangible and it shipped. But Engelbart cared far more about the parts that didn't catch on as cleanly.

He cared about structure — the ability to organize knowledge as linked, addressable, restructurable pieces rather than flat pages, so that a group's understanding could accumulate and be navigated. He cared about collaboration — multiple people working in the same shared knowledge space, which the demo showed and which most software still does clumsily half a century later. And above all he cared about bootstrapping: his lab used NLS to build NLS, deliberately turning their tools on the work of improving their tools, in an accelerating loop. The demo itself was a product of bootstrapping — they used the system to prepare the presentation of the system.

These ideas — structure, collaboration, bootstrapping — are subtler than a pointing device, and they've diffused into the world far more unevenly. We got the mouse everywhere and the deep collaborative knowledge-structuring almost nowhere.

Augmentation, not automation#

There's a fork in the road that Engelbart's work marks clearly, and it's worth naming because we keep arriving back at it.

One path is automation: build machines that do the thinking instead of people. The other is augmentation: build machines that make people, working together, better at thinking. Engelbart, like Licklider before him, was firmly on the augmentation side. His goal was never a computer that solves your problem for you; it was a computer that makes you and your colleagues dramatically more capable of solving it yourselves.

That distinction has only grown more pointed. Every era of computing renews the choice — and tends to find automation easier to fund and easier to sell than augmentation, because automation promises to remove the messy human, while augmentation requires investing in them. Engelbart spent the back half of his career underappreciated partly for refusing that trade.

Why watch it now#

The demo is on the web; you should watch at least the opening. Two things tend to strike people who do.

The first is how modern it feels — how many things you use every day were sitting there, working, in 1968, in a world that otherwise looks impossibly distant. The second is the quieter realization that comes after: that some of what he showed, we still haven't really built. The collaborative, structured, knowledge-augmenting system Engelbart demonstrated was, in important ways, ahead of the tools on your screen right now.

That gap is the standing provocation of the whole tools-for-thought tradition. Engelbart didn't show us a future we surpassed. He showed us one we partially picked up, partially dropped, and are still — slowly, unevenly — learning to deserve.