[Humanist] 30.824 history of AI

Humanist Discussion Group willard.mccarty at mccarty.org.uk
Thu Mar 16 09:56:09 CET 2017


                 Humanist Discussion Group, Vol. 30, No. 824.
            Department of Digital Humanities, King's College London
                       www.digitalhumanities.org/humanist
                Submit to: humanist at lists.digitalhumanities.org



        Date: Wed, 15 Mar 2017 09:23:35 +0100
        From: Tim Smithers <tim.smithers at cantab.net>
        Subject: Re:  30.823 events: DayofDH; history of AI; epigraphy
        In-Reply-To: <20170315063347.361CA8C81 at digitalhumanities.org>


Dear Colin,

In your proposal for a SHOT Workshop, on The History of
Artificial Intelligence and its Discontents, you cite
McCorduck 1979, 2004; Crevier 1992; Brooks 1999; Boden 2006;
Nilsson 2010.

May we have full references for these, please.

In the full proposal you point us to [1], you present a list
of questions as "Some possible directions for inquiry ..."

These all seem rather loaded to me.  Does this make for good
historical inquiry, I wonder?

The fourth one, for examples, asks

  What does "failure" mean in a discipline that has never
  enjoyed "success" until recent decades?  (e.g. 1997 Deep
  Blue, Watson, AlphaGo)

Historically, AI as a discipline got started, I think most
historians would agree, with the Dartmouth Conference of 1956
[2].  Computer Chess was a topic discusses, but no working
Chess program was presented.  (Newell and Simon were the only
participants to present a working program, of their Logic
Theorist [3]).

To go from not much computer chess in 1956 (see [4]), to a
world champion beating Chess machine in 1997, little more than
four decades, looks like remarkable progress to me, and to
some chess players too, I think.

To talk of no "success" until recent decades is to distort
history, I would say. AI has only existed in recent decades.
During the four decades from Dartmouth to the Deep Blue II
success over Garry Kasparov, there were many relative, but
real, successes, as computer chess programs got better.

A perceived failure of AI to fulfill publically stated overly
ambitious predictions does not mean no success.  The real
history of AI is a lot more complicated than your "possible
directions of inquiry" seem intent upon making out.

But, I'm not the historian here, so perhaps it's not for me to
complain about how good history is to be done.

Best regards,

Tim

References

[1] SHOT 2017 Open Panel Proposal: Title: The History of
    Artificial Intelligence and its Discontents
    <http://www.historyoftechnology.org/media/Philadelphia_2017_open_panel/open_panels/SHOT_Open_Session_History_Artificial_Intelligence_and_its_Discontents.pdf>

[2] Dartmouth Artificial Intelligence (AI) Conference
     http://www.livinginternet.com/i/ii_ai.htm

[3] Gualtiero Piccinini: Allen Newell,
     http://www.umsl.edu/~piccininig/Newell%205.htm

[4] Dietrich Prinz, a colleague of Alan Turing, is
    recorded as having developed the first automated chess
    playing program, in 1951.  It ran on the then new Ferranti
    Mark I computer at Manchester University.  It couldn't
    play a complete game of chess, however, due to memory and
    computational limitations of the Ferranti machine.

> On 15 Mar 2017, at 07:33, Humanist Discussion Group <willard.mccarty at mccarty.org.uk> wrote:
> 
>                 Humanist Discussion Group, Vol. 30, No. 823.
>            Department of Digital Humanities, King's College London
>                       www.digitalhumanities.org/humanist
>                Submit to: humanist at lists.digitalhumanities.org
> 
>  [3]   From:    Colin K Garvey <garvec at rpi.edu>                            (7)
>        Subject: CfP: SHOT 2017 Open Panel on "AI & it's Discontents"
> 
> --[3]------------------------------------------------------------------------
>        Date: Tue, 14 Mar 2017 17:41:04 -0400
>        From: Colin K Garvey <garvec at rpi.edu>
>        Subject: CfP: SHOT 2017 Open Panel on "AI & it's Discontents"
> 
> 
> [forwarded from the Society for the History of Technology (SHOT)]
> 
> Dear all,
> 
> In preparation for SHOT 2017, we are soliciting papers for an open panel on:
> 
> The History of Artificial Intelligence and its Discontents
> 
> With origin myths involving many of the 20th century’s Great Men of Science and their intimate relationships with that most revolutionary and profound of transformations—the emergence of computing—the history of artificial intelligence (AI) has long been popular fare. Indeed, the field itself is remarkable for its studious cultivation of its own history, with much of the canon written primarily by AI insiders and developers themselves (e.g. McCorduck 1979, 2004; Crevier 1992; Brooks 1999; Boden 2006; Nilsson 2010). Perhaps unsurprisingly, the story they tell is more often than not a triumphant march of progress towards the current pinnacle we are standing on today—with a few bumps along the road, thrown in for good measure. Now that AI is back and booming again today, offering promises and threats of massive social transformation, this progress narrative risks becoming a hegemon, legitimating the technological status quo and closing down opportunities for democratic deliberation about the direction of change. What possibilities exist for disrupting, destabilizing, and otherwise challenging the insider’s Whig history of AI? Who and what are AI’s discontents? How have they challenged, resisted, disrupted, or destabilized AI?
> 
> Submissions are due before March 31st. Please feel free to contact me, Colin Garvey <garvec at rpi.edu> for more information.
> 
> For more details please see http://www.historyoftechnology.org/call_for_papers/open_panels.html  






More information about the Humanist mailing list