IBM Watson and the Future of Work

October 15, 2011

IBM Watson™ made its debut in February 2011 when the Deep Question & Answer software system defeated two previous reigning human champions – Ken Jennings and Brad Rutter in a three-night special Jeopardy! showcase. The event made for carnival headlines but most news reports failed to connect Watson’s performance with its intended application- to transform the nature of human productivity in an age of information-rich, context dependent and software-mediated work environments. Watson is designed to augment (improve) our capacity to think through complex problems, ask the right questions,  judge possible solutions and make informed confident decisions based on real-world data that exists within our own memory banks and beyond.

Productivity and Life-long Learning via Personal Assistants IBM Watson™ and Apple Siri™  are early signals of what might transform work and lifelong learning around software based personal assistants that push human beings to think more deeply and broadly about questions, answers and their personal confidence levels in making decisions. IBM is leading the way in an emerging paradigm for software – based on improving human cognitive performance in an era of endless streams of data and changing contexts around the marketplace and collective industry knowledge base. The next step for IBM’s Watson is to enter the workplace and help to transform the capacity of human work.  IBM’s public roadmap for Watson begins in three main industries: Healthcare, Finance and Customer Service.  But first, let’s explore why Watson matters….

Why Watson Matters… Natural Language, Box in a Cloud, Focus on Answers & Honesty about Confidence Levels 

There are four things that matter with IBM Watson that are relevant to understanding the future of work and learning (machine and human):

1) Natural Language Matters Watson is not alive.  It is not artificial intelligence.  But it can (better than any other system on Earth today) understand the nuanced elements of meaning created by natural language. Forget about display screens, clicking with your mouse and typing on a keyboard.  Watson (and to a lesser degree – Siri) allow us  to engage us in conversation and overcome the ambiguity associated with human language.  Watson is an ideal ‘post screen’ interface that helps to lower the barrier to workplace applications where screens might not be conducive to workflow.

2) Knowledge in a Box Matters The web revolutionized access to information, but has also led to a world with too much information — and at times –  too much inaccurate information.  The web is also limited in its capture of unstructured and private data.   So we can recognize a fundamental limitation to a purely ‘open’ web platform for advanced human augmentation systems. Knowledge requires filters for transparency, authentication and accountability.  There is benefit to controlling information in a silo that is constantly updated. Watson is a self contained storage, retrieval, analysis systems.  Watson is a ‘box’ with a 15 trillion-byte memory capacity which allows IBM to be sure that the information output will be shaped by the input rather than extracting data from the open web. The next step is putting the Box in the Cloud — and opening up a portal to non-supercomputing devices!!

3) Answers Matter Today we search the web and receive a list of websites which we must read to find the answer.  Again, there are no filters to guide that process.  The website we choose was likely placed on the first page of a Google Search bar – or shared with us via social networks. Watson does not give you a list of websites, it gives you the answer(s). It might take years to change behavior– but in a not-so-distant future, our ‘search’ expectations are likely to shift to more answer oriented results.  Obviously there is tremendous potential for an upside and downside (e.g. narrow casting; critical thinking issues) to an answer engine web culture– but we can see other players such as Wolfram Alpha and Google validating the industry’s new direction towards ‘answers’ as the next step for search. One wonders if Google and IBM become learning management systems in the end…

4) Confidence Matters Watson knows that it is not perfect.  IBM recognizes that technology cannot deliver certainty on demand. So Watson embraces uncertainty and is honest about its confidence level with each response.  For each question it receives, Watson assigns a ‘Confidence’ level (%) and chooses to respond – or not – based on the situation. One can only dream of a world where humans approach real-world challenges filled with uncertainty – with answers that reflect our recognition that the answers – solutions – might not be perfectly clear. Imagine a work environment where people are honest and transparent in their knowledge level – and confidence level to respond to a particular question!  Instead of giving answers to please our colleagues and customers we can envision a future world of work where uncertainty is dealt with using a range of possible answers given our best set of inputs. …. Apple’s Siri is cute— but Watson is the real potential game-changer in these early days of software-mediated human performance!

Real World Applications for ‘Watson’-like Software Programs 

Healthcare Why healthcare?  It is impossible to know all new information in the world of life sciences.

IBM Watson – Commercial for Healthcare   IBM Watson – Future of Financial Services IBM Watson – Call Centers 

Finance – Assist in providing humans with real-time market information.

 

Watson is not Alone

  • Google is not shy about hiding its ‘real-time’ and ‘voice’ based interface innovations.  Google’s vision is to have the answer ‘as fast as you think’.
  • Apple has recently launched Siri as a consumer grade personal assistant.
  • Microsoft’s promotion of Bing is as a ‘decision engine’ – not a lowly search engine.
  • Wolfram Alpha (and a number of other startups) are involved in this ‘answer engine’

Key Concepts:

  • Natural Language Interface; Conversational Interface
  • Analytics; Deep Analytics; Deep Question & Answer;
  • Algorithms; Data;  Unstructured Data; Data-Information-Knowledge-Wisdom (DIKW)
  • Parallel Processing; Distributed Computing; Multi-core; Scaling; Power 7 Platform;
  • Augmentation; Intelligence Augmentation;  information processing augmentation

People

Video Resources IBM Watson: Overview (20 min) IBM – The Next Grand Challenge: Natural Language   IBM Watson on Ambiguity  Science Behind IBM Watson

Manoj Saxena IOD Keynote – Putting IBM Watson to work

IBM Watson – Future of Financial Services Work IBM Watson – Future of Customer Service 

Leave a Comment

Previous post:

Next post: