MIT Project MAC and AI Group

By issack
  • AI Lab is Founded

    The AI Lab (also referred to as AI Group) was founded to research the development of artificially intelligent programs, and was eventually merged with Project MAC due to the heavy usage of resources and initially similar goals the two projects shared (Chiou 3).
  • Period: to

    MIT Computing Demands Spark Timesharing Development

    In 1957, IBM donated a 704 system to a MIT lab, with the condition that MIT would receive an eight hour shift on the computer, but was frequently undermined by the computations hackers ran on the system. Even though MIT eventually received upgraded systems, the eight hour restriction constantly blocked progress. This led to the idea of batch, or grouped, processing for multiple developer/hackers' programs at a time, and eventually the idea of timesharing (Chiou 4-6).
  • Time Sharing Concept

    In 1960, researchers at MIT started developing a multi-user computer usage concept called timesharing, as usage of the early computers at MIT became heavy (Chiou 5).
  • Time Share System Development Begins

    Time Share System Development Begins
    Two timesharing development projects started at MIT: one by Jack Dennis, a solo project working with the PDP-1, and by F.J. Corbató, writing for the IBM 7090 (along with assistance from IBM) (Levy 58-59). (Exact date unknown, but is sometime after the initial timesharing idea) Source for both images:
    https://upload.wikimedia.org/
  • Period: to

    AI Group Split Troubles

    Unlike the timesharing systems, AI development was mostly handled by hackers and very few graduate students. Minsky's intervention within the group was minimal, leading to this loose conglomeration of development. However, after the split, the AI Group continued to receive a third of the ARPA funding, as well as independent funding, now that they were independent, and the final blow to Project MAC was the development of the ITS, which negated the use of CTSS or MULTICS (Chiou 33).
  • Richard Greenblatt Joins MIT Lab

    After devoting large amounts of time to programming on the PDP, and time spent with the TMRC, Greenblatt failed out of his classes from MIT, but was eventually hired by Project MAC, before it was officially known as MAC, shortly before receiving DARPA funding (Levy 66). (Date is approximate)
  • Project MAC is Born

    Project MAC is Born
    MIT received a long-term grant from the US Department of Defense's ARPA project to work on the timesharing project, which was now known as Project MAC, MAC standing both for Multiple Access Computing, and Machine Aided Cognition. With $3 million a year in funding, the researchers were enthused (Levy 58-59).
    This new undertaking was lead by Dennis and supported largely by Marvin Minsky (image included). Image Source:
    http://eet.com
  • MULTICS Development Begins

    MULTICS was developed as part of Project MAC in response to Corbató's proof that his CTSS was, in practice, a viable timesharing system, but had limitations. However, with the funding from DARPA, Corbató was able to begin development on MULTICS, which was unique in that it was written in a high-level language (unlike CTSS, written in low-level assembly), and that MULTICS emphasized security (Chiou 16-17)
  • ITS Development Begins

    As Project MAC realized that development on MULTICS would become a long, drawn out process, the "bootleg" side development of the Incompatible Timesharing System began, for the PDP-6 (Chiou 26).
  • Development of the Greenblatt Chess Program

    Development of the Greenblatt Chess Program
    Grenblatt, along with several other members of the AI Group developed an artificial intelligence program, capable of playing chess against a human opponent. This was known as "Mac Hack", and was drawn to a screen as displayed in the picture (Greenblatt 801-802). Levy cited this as a historically significant hack, as the development nature of the program was not official, and made use of low level tinkering of the PDP-6 computer (82). Image Source:
    https://chessprogramming.wikispaces.com
  • Period: to

    Project MAC Dissolving

    The once proliferate Project MAC was weakened by the split with the AI Group, and most members left with Minsky for the AI Group. As a result, the remaining members formed the Laboratory for Computer Science, or LCS (MIT).
    (Dates are approximate as this began right before the split with the AI Group)
  • AI Group Splits from Project MAC

    After several years of internal conflict, the mostly unstructured AI Group split from the strictly organized Project MAC, citing differences in philosophy and the loose nature of the hackers associated with the AI Group (Chiou 29-31).
  • Richard Stallman Joins AI Group

    Richard Stallman Joins AI Group
    Stallman joined the AI Group as a hacker, sharing much of the Hacker Ethic of the earlier hackers. He was tasked with improving on the ITS with the PDP-10 computer (Stallman, "About the GNU Project"). Image Source:
    https://upload.wikimedia.org
  • Period: to

    Lisp Machine Development

    During this period, Greenblatt and Thomas Knight, as part of the AI Group, began work on one of the first Lisp machines; a computer designed to mostly run the Lisp programming language as its main software (Chiou 7-8).
    This development is what eventually led to the slow demise of the hacker population in the AI Group, and the events leading up to Stallman starting the GNU project (Stallman, "About the GNU Project").
  • Greenblatt and Knight with the CADR Lisp Machine

    Greenblatt and Knight with the CADR Lisp Machine
  • Period: to

    AI Group Slowly Collapses

    As DEC, the manufacturer/developer of the PDP systems, began to phase out their now aging PDP systems, the AI Group likewise began to slowly dissolve. The then new company Symbolics had hired many of the former hackers from the AI Lab, and as a result, the Lab began to regress into using proprietary formats, which began with shifting away from the ITS and using the proprietary, built-in timesharing system included with the new PDP-10 (Stallman, "About the GNU Project").
  • Stallman Leaves MIT

    After experiencing a shift in the development paradigm within the AI Group, Stallman left, as he began to despise the proprietary software and NDAs that began to permeate the Lab, and did not want MIT to take ownership of his newly formed GNU software and philosophy (Stallman, "About the GNU Project").
  • World Wide Web Consortium Founded

    World Wide Web Consortium Founded
    Tim Berners-Lee, who invented the World Wide Web (in concept and in execution) set up the Consortium's (aka W3C) HQ within Tech Square, heavily associating with the AI Lab projects and Project MAC (MIT News). Image Source:
    https://upload.wikimedia.org/
  • Tech Square Changes

    The Tech Square, which still housed the AI Lab and LCS, was bought by a development firm, and the building was renovated, bringing in many technology companies, forcing the projects to move to a new MIT building.
  • AI Lab and LCS Merge to Form CSAIL

    AI Lab and LCS Merge to Form CSAIL
    Putting their troubled histories aside, the once joined and split AI Lab and LCS (formerly Project MAC), were joined again to form the CSAIL, or the Computer Science and Artificial Intelligence Laboratory (multiple MIT pages).
  • Works Cited

    “About MAC 50 and Project MAC.” About Mac 50 And Project Mac, MIT CSAIL, http://mac50.csail.mit.edu/about.
    Chiou, Stefanie. “A Marriage Of Convenience: The Founding Of The Mit Artificial Intelligence Laboratory.”
    Greenblatt, Richard D. et al. “The Greenblatt Chess Program.” Computer Chess Compendium, 1988, pp. 56–66.
    Levy, Steven. Hackers: Heroes of the Computer Revolution. Sebastopol, CA, O'Reilly, 2010.
  • Works Cited Part 2

    “About CSAIL.” MIT CSAIL, https://www.csail.mit.edu/about.
    “CSAIL History.” CSAIL Alliance Program, http://cap.csail.mit.edu/about/history.
    “MIT Leaves behind a Rich History in Tech Square.” MIT News, 17 Mar. 2004, http://news.mit.edu/2004/techsquare-0317.
    Stallman, Richard. “About the GNU Project.” GNU Project, GNU Project, https://www.gnu.org/gnu/thegnuproject.en.html.