Modern linguistics and the early application of AI and Machine Learning have much in common. Designers of AI solutions can learn how to make a meaningful, measurable, and profitable impact through understanding language theory of the late 20th century:
B.F. Skinner‘s (1) behaviorist application of operant conditioning posits language acquisition as a learning process of stimuli and responses.
Noam Chomsky‘s (2) rebuttal of behaviorism (3) proposes language acquisition as an innate, uniquely human cognitive capability, founded on rules-based principles, called “syntax”.
J.L. Austin (4) refutes “grammarians” and argues that human language is a sequence of “performatives” representing actionable intentions.
Concepts such as “learning process”, “cognitive capability” and “actionable intentions” ring familiar to the current discourse of natural language processing (NLP) and our fledgling AI-enabled bot society. All elements represented in linguistic theory (learning through inferencing, rules-based engines, and intent) are foundational platforms manifest in Cortana, Alexa, Google Assistant, et al.
I have argued in favor of anticipating a bifurcation in the VPA market, separating omniscient, knowledge-based big-data applications from intention-driven, bot-enabled and action-oriented utilitarian VPAs (5).
AI application and service designers should apply 20th century linguists’ theories: all are applicable to NLP, but your commercial outcome will greatly depend on which you implement:
- Behaviorism will lend itself to navigating your customer to the desired outcome, by applying decision trees in interactive voice response systems (IVRs).
- Cognitive science and the study of syntax will help your AI scale from a finite set of utterances to a practically infinite set of interactions.
- Intent-driven pragmatist approaches will lead to purpose-driven commercial outcomes.
For many b2c applications, the latter “AI Pragmatics” approach offers the easiest time-to-ROI. It requires a clear understanding of the user intention (“Purchase movie ticket”), a set of skills specific to the intentional domain (local movie titles, showtimes, theaters), and a bot that applies the AI to the specific, pragmatic outcome. You can try it out today using the Fandango Alexa skill.
(1) B.F. Skinner. Verbal Behavior, 1957
(2) Noam Chomsky. Syntactic Structures, 1957
(3) Noam Chomsky. “A Review of BF Skinner’s Verbal Behavior.” 1957
(4) J.L. Austin. How to Do Things with Words, 2nd Edition, 1962
Comments or opinions expressed on this blog are those of the individual contributors only, and do not necessarily represent the views of Gartner, Inc. or its management. Readers may copy and redistribute blog postings on other blogs, or otherwise for private, non-commercial or journalistic purposes, with attribution to Gartner. This content may not be used for any other purposes in any other formats or media. The content on this blog is provided on an "as-is" basis. Gartner shall not be liable for any damages whatsoever arising out of the content or use of this blog.