Stanford Artificial Intelligence Laboratory (SAIL)

What it is

The Stanford Artificial Intelligence Laboratory (SAIL) is an AI laboratory at Stanford University.^[1] Christopher D. Manning is listed as its current director.^[1]

Origins

SAIL was started by John McCarthy in 1963, after he moved from MIT to Stanford. The SAIL archive dates the start of the Stanford AI Project to 15 June 1963.^[1][2] The Kyoto Prize biography credits McCarthy with founding the AI laboratory at Stanford.^[4] The SAIL archive, which records the lab's own history, is one of the sources for its 1963 start date.^[2]

Lester Earnest was its executive officer from 1965 to 1980. From 1966 to 1979 the lab was housed in the D.C. Power building, a remote hilltop site about 5 miles from campus.^[1]

Known for

The Stanford arm, designed at SAIL in 1969, is described as the forerunner of most industrial robots in use today.^[1]

Lab members created the SAIL programming language, with Dan Swinehart and Bob Sproull in 1970, and the WAITS operating system for DEC PDP computers. The SAIL computer system was shut down in 1991.^[1]

The Stanley self-driving car, associated with the lab, won the 2005 DARPA Grand Challenge.^[1] The lab's robotics and vehicle work, from the Stanford arm to the Stanley car, is one of the threads that runs through its history.^[1]

People from the original SAIL include Raj Reddy, Hans Moravec, Alan Kay, Victor Scheinman, Larry Tesler, Don Knuth and Edward Feigenbaum.^[1]

Place in AI history

In 1979 SAIL merged into Stanford's Computer Science Department. It reopened in 2004 in the Gates Computer Science Building, with Sebastian Thrun as director.^[1]

Fei-Fei Li directed SAIL from 2013 to 2018.^[3]

For researchers today, the lab's name is attached to projects in which robotics, autonomous vehicles and the early symbolic AI tradition appear side by side. The lab's history connects the early years of AI, when a small group in one building pursued the field's basic questions, with later work in robotics, self-driving vehicles and computer vision.

For readers, SAIL shows how one laboratory can carry a field's ideas across decades, from the hardware and software of its early years to the robotics and vehicle projects of later decades.