Consilient Network· The graph of everyone who built this · Add your entry ↗
Story path

Punch cards at eight, a PDP-8 at fifteen, and what each machine taught

Every layer of this architecture was preceded by a smaller version of the same problem, encountered on real hardware, twenty to forty years earlier.

7 stopsFor: Engineers, and anyone who wants the technical formation rather than the biographyAll paths
His first real scientific experiment rested on Einstein's fourth 1905 paper. He would attend the centenary of that year twenty-five years later, and be walked into the keynote by Murray Gell-Mann.
The path

Follow it in order

Each stop is a page in the graph. The order is an argument — but every one of them is also a door into the rest of it, so you can leave the path wherever it stops being useful.

01My first real experiment was Einstein's, run on a PDP-8 at fifteenA rack-mounted minicomputer on wheels, assigned to sweep a laser across a water flume so particle transport in fluid boundary layers could be visualized. Brownian motion — Einstein, 1905.→02Recruited by USC at fifteen, and shown for the first time what real wealth looked likeThe program it happened on: recruitment dorms in South Central, a donated cinema he read in and never saw a film in, a boat to Catalina, and a pumped-storage plant at Tehachapi Pass.→03Impell: nuclear safety software at sixteenNuclear safety systems, taken deliberately on the view that nuclear was the most sustainable energy available — outside orthodoxy then, largely adopted since. The grid knowledge from Tehachapi was directly useful.→04UC Berkeley: a 2.7 forced curve and twenty-nine Nobel laureatesVAX and VMS in the engineering departments, PDP for real-time control, and simulation of dynamic systems as the formal discipline.→05Blyth Software: Omnis, DEC, and the Immaculate RevolutionOmnis — the only client-server tool running across Mac and Windows and connecting to Oracle, Sybase, SQL Server, DB2 and DEC's Rdb. Cross-platform coordination, in product form.→06TGV Software: the class project that opened every door afterThe class project: TCP/IP bridging Macs, PCs, Unix and VMS. The same coordination problem one layer down — and the study that put him inside Microsoft.→07I named the missing layer at Microsoft in 1991. Thirty-five years later it is still missingWhere it was finally named. A coordination layer above applications and below business practice, identified in 1991 and still the claim.→
Where it lands

What the path is for

The point

Nobody arrives at a coordination architecture from theory. They arrive at it by hitting the same wall at every layer — a laser controller, a grid, a database connector, a network bridge, an enterprise application — until the wall itself becomes the subject. The power constraint on AI today is the Tehachapi lesson returning: Moore's law is far weaker in energy than in silicon.

Add to this record → Other paths →

Share this
← PreviousAll paths
+ Add to the record