digitalwellbeing.orgHow to thrive in our hyper-connected world

Artificial Intelligence Defined: Useful list of popular definitions from business and science

The world (including marketing) seems to have hit peak-A.I.  Marketers are either marketing to us with A.I. (with a view automating and improving their marketing), or they are actually marketing A.I. to us with products and services infused with A.I.

But what exactly is A.I.? In a world of marketing hype and spin, it can be difficult to distinguish hype from reality.  So here’s a list of simple, basic definitions of A.I. from industry and science – with sources – to help you see beyond the buzzword.

My definition and contribution, borrowing and adapting from these definitions is

Artificial Intelligence is technology that behaves intelligently using skills associated with human intelligence, including the ability to perceive, learn, reason and act autonomously.

Put very simply, A.I. is a collection of technologies that allow computers to think for themselves – i.e. work out what to do – usually in the context of achieving a particular task. This is different from traditional data processing and analytics that run on fixed algorithms. Instead, A.I. technology gives computers cognitive powers such as the ability to perceive, learn, abstract (classify, conceptualize and generate rules) and reason. In other words, A.I. makes ‘dumb’ computers intelligent, where intelligence is the computational part of the ability to achieve goals in the world. In the context of marketing, this means giving computers cognitive smarts (mental powers) to make marketing automation and marketers smarter. Put simply, A.I. technology in marketing is an area of Applied A.I. that helps marketers define and execute marketing strategies. Just think of A.I. as smart-drugs for computers (or as DARPA call it spreadsheets on steroids).

But onto the list of more established definitions…

Artificial Intelligence is...

  • “… the science and engineering of making intelligent machines” … “[where] intelligence is the computational part of the ability to achieve goals in the world” (original definition by John McCarthy who coined the term ‘Artificial Intelligence’ in 1955)
  • “… making a machine behave in ways that would be called intelligent if a human were so behaving” (alternative definition by John McCarthy who coined the term ‘Artificial Intelligence’ in 1955)
  • “… the science of making machines do things that would require intelligence if done by men” (definition offered by A.I. pioneer Marvin Minsky in 1968)
  • “… the science of making machines smart” (Demis Hassabis, CEO and founder of DeepMind, now part of Google)
  • “… is an intelligent machine” (Google’s Avinash Kaushik)
  • “… is the next, logical step in computing: a program that can figure out things for itself. It’s a program that can reprogram itself” (Jim Sterne, author of Artificial Intelligence for Marketing)
  • “… anything a machine does to respond to its environment to maximize its chances of success” (Steven Struhl, author of Artificial Intelligence Marketing and Predicting Consumer Choice)
  • “… technologies emerging today that can understand, learn, and then act based on that information” (PwC‘s definition)
  • “… anything that makes machines act more intelligently” (IBM‘s definition)
  • “… a constellation of technologies that extend human capabilities by sensing, comprehending, acting and learning – allowing people to do much more” (Accenture‘s definition)
  • “… getting computers to do tasks that would normally require human intelligence” (Deloitte’s definition)
  • “… the ability of machines to exhibit human-like intelligence”  (McKinsey‘s definition)
  • “… a field of computer science that focuses on creating machines that can learn, recognize, predict, plan, and recommend — plus understand and respond to images and language (Salesforce‘s definition).
  • “… technology that thinks and acts like humans” (popular business use of the term as reported in a 2015 Narrative Science survey)
  • “… a subfield of computer science aimed at the development of computers capable of doing things that are normally done by people — in particular, things associated with people acting intelligently” (Definition in Practical A.I. for Dummies)
  • “… the replication of human analytical and/or decision-making capabilities Steven Finlay (Author of Artificial Intelligence and Machine Learning for Business, 2017)
  • “… the ability of a digital computer or computer-controlled robot to perform tasks commonly associated with intelligent beings” (definition in Encyclopedia Britannica by Prof. B.J. Copeland)
  • “… a set of computer science techniques that enable systems to perform tasks normally requiring human intelligenceEconomist Intelligence Unit‘s definition
  • “… a computerized system that exhibits behavior that is commonly thought of as requiring intelligenceUS Government definition (NSTC)
  • “… intelligence demonstrated by a machine or by software…[where] intelligence measures an agent’s general ability to achieve goals in a wide range of environments” (Calum Chase, author of Surviving A.I.)
  • “… intelligence exhibited by machines, rather than humans or other animals (natural intelligence, NI)” (Wikipedia definition 1)
  • “… intelligence exhibited by machines or software” (Wikipedia definition 2)
  • “… the study of agents that receive percepts from the environment and perform actions” (the classic textbook definition from Peter Norvig and Stuart J. Russell that classifies A.I. into four areas (and focuses on the fourth), agents that 1) think like humans, 2) act like humans, 3) think rationally, 4) act rationally – “a rational agent is one that acts so as to achieve the best outcome or, when there is uncertainty, the best-expected outcome”)

Towards a simple summary

Although these definitions are varied, there are some common themes. The first is that A.I. typically takes actions (it acts on information rather than simply processes it), and does so with a degree of autonomy (i.e. A.I. automates intelligent actions typically taken by a human). Secondly, A.I. is task or outcome-focused and adapts its behavior to achieve its goals. Thirdly, A.I. can re-program itself, based on what it learns. In other words, A.I. is active, agentic, automatic and adaptive.

More simply, these definitions can be synthesized by summarising A.I. as any technology that behaves intelligently” [insofar as it responds adaptively to change] (the capacity to respond adaptively to change through the acquisition and application of knowledge is a hallmark of intelligence – i.e. the ability to cope with novel situations).

In humans, this information processing ability requires intellect and intuition, made up of basic cognitive (mental) processes including attention, perception, reasoning, problem-solving, judgment, learning and memory. It follows that one useful synonym for A.I. (along with machine intelligence, or computational intelligence) is cognitive computing.

But it is the capacity to respond adaptively to change and circumstance – as a defining hallmark of intelligence – that might be particularly useful for the practical understanding and use of A.I. in marketing and beyond. From this perspective, A.I. is adaptive technology that can pursue goals (e.g. improving sales) by adapting to changing inputs and contexts. For example, a website, ad or promotion that adapts to a user profile or situation in order to achieve a specified marketing goal, could be said to use A.I.

So beyond that idea that A.I. is an umbrella term referring to technology that behaves intelligently, or what smart computers use when they don’t know what to do, A.I. is adaptive technology that allows machines to complete tasks successfully in situations of change or ambiguity.

Written by
Dr Paul Marsden
Join the discussion

15 comments
digitalwellbeing.org

digitalwellbeing.org

Digital wellbeing covers the latest scientific research on the impact of digital technology on human wellbeing. Curated by psychologist Dr. Paul Marsden (@marsattacks). Sponsored by WPP agency SYZYGY.