Wednesday, October 17, 2018

Reading :: Mind over Machine

Mind Over Machine
By Hubert L. Dreyfus and Stewart E. Dreyfus

I've been rereading some of the mid-1980s work that introduced activity theory to the West, specifically paying attention to how it was positioned as an alternative to information-processing cognitive psychology (IPCP). Among others that were heavily cited at the time is this 1986 book, which asks the question: why hasn't artificial intelligence (AI) yielded the results that had been predicted in the 1950s and 1960s?

The 1950s and 1960s were a fertile time for beginning AI research, with Newell and Simon working on cognitive simulation at RAND (pp.6-7). Also at RAND was Stewart Dreyfus (henceforth SD), who was programming JOHNNIAC. Hubert Dreyfus (HD), a philosopher at MIT, expressed misgivings about AI to his brother and was in short order hired as a consultant for RAND in 1964 (p.5).

HD recognized that AI researchers were animated by the "continuum hypothesis": they believed that they were making the first steps, and if they continued, they would make steady progress. In contrast, HD saw a pattern in which AI researchers would solve a simple problem, consider it a first step to more complex problems, then encounter "failure when confronted with the more intuitive forms of intelligence" (p.7). His observations were not greeted with enthusiasm (p.8), but constituted the first detailed criticism of AI. As SD says later on, "Current claims and hopes for making progress in models for making computers intelligent are like the belief that someone climbing a tree is making progress toward reaching the moon" (qtd. on p.10).

To explore the contrast between AI approaches and human expertise, the authors distinguish between "know-how" and "know-that" knowledge—i.e., tacit and embodied knowledge vs. explicit knowledge (p.16). They propose a model with five steps to expertise:

  1. Novice
  2. Advanced beginner
  3. Competence
  4. Proficiency
  5. Expertise (Ch.1).
In the early stages, formalized or explicit knowledge is critical. These also represent the areas in which AI is most suited to assist, since AI excels at processing formal knowledge. But in the later stages, what is required is intuition: "Intuition or know-how, as we understand it, is neither wild guessing nor supernatural inspiration, but the sort of ability we all use all the time as we go about our everyday tasks" (p.29, their emphasis). See Klein for more on intuition in this vein. And like Klein, the authors state that "When things are proceeding normally, experts don't solve problems and don't make decisions; they just do what normally works" (pp.30-31, their emphasis). Put another way: "Competent performance is rational; experts act arationally" (p.36). In this sense, computers are "ideal beginners" (p.63) 

There's more to the book, but let's stop here, because this criticism and this model constitute the enduring legacy of the book. If you're interested in the historical development of AI and understandings of expertise, or in a model of expertise, definitely pick up this book.

Reading :: The Extended Mind

The Extended Mind
Edited by Richary Menary

Don't confuse this book with the one I recently reviewed with the same title by Robert K. Logan. This one is a collection of essays about the extended mind hypothesis, famously discussed in an article by Clark and Chalmers and more recently elaborated in Andy Clark's book Supersizing the Mind.

This collection includes an introduction by the editor; Clark and Chalmers' original essay; and discussions by various respondents interspersed with Clark's own rejoinders. Note that the original essay was also reprinted in Supersizing the Mind and Clark also repurposed some of his rejoinders there, so if you read both books, you'll find a lot of overlapping material.

In the interest of time, I'll just highlight a few things that do not overlap.

In a massive Ch.6, "The varieties of externalism," Susan Hurley situates the extended mind hypothesis among other externalist thought. You can read the entire chapter, but see Table 6.3 on p.144, which summarizes the different strands of externalism in a four-field whose axes are "content/quality" and "what/how."

In Ch.8, "Meaning making and the mind of the externalist," Robert A. Wilson also discusses variants of externalism. Wilson notes that "active cognition arguments" have been with us a while; he categorizes Vygotsky and Luria's mediational approach as such an argument (p.172). "These arguments all focus on determinate forms of a particular cognitive ability (e.g., memory, attention, problem solving) as they are exercised by individual agents. They view the integration of individuals with both their biological and artificial environments as critical to their status as cognitive agents with these particular capabilities" (p.172). In contrast, he characterizes Clark and Chalmers' argument as a "cyborg fantasy" argument (p.173, his emphasis). Such arguments "proceed by introducing an imaginative example in which an individual's performance is mediated by external forms of technology, typically arguing, through a comparison with cases in which the same kind of activity is performed without such mediation, to the conclusion that that the boundary between what's inside the head and what is in the environment is irrelevant to whether a given agent has some particular cognitive capacity" (p.173). And "the chief aim of cyborg fantasy arguments has been to establish the extended mind as a conceptual default; they do so by shifting the burden of proof to internalists, challenging them to identify why the skin should be a relevant boundary for cognition at all" (p.173).

Overall, this collection was an interesting shakedown of the extended mind concept. Unlike Supersizing the Mind, it included different perspectives and criticisms of the concept, and I especially appreciated Wilson's contrast with mediation. At the same time, many of the authors are not as engaging writers as Clark. Nevertheless, if you've become interested in the extended mind concept, I suggest you start here first.