IBM Watson – The rise of the (thinking) machine

I remember a few years ago, speaking on topics of AI and thinking machines, only to be critiqued (by some, not all) of my analyst colleagues for having expectations well beyond accepted realms of timescales and possibility. Visiting IBM’s World of Watson event in November proved that not only was I right, but even (dare I say it) not aggressive enough on my expectations.

To me, what the World of Watson, with its 17,000 attendees, quite eloquently reflected was that the era of the ‘thinking machine’ is here, today. From its first public appearance on Jeopardy (2011, Watson has clearly matured from a fascinating proof of concept to practical AI. In IBM speak, it’s hailed as their ‘Cognitive Platform’. It’s versatile too; made abundantly clear through the many customer cases highlighted at the event, so too many demonstrations and presentations on its application. From email to whiteboards, consumer products to manufacturing, connected cars, finance to insurance, buildings to cities, health to IT support, the list goes on and on….

So what is it about IBM and Watson that’s different? For one thing IBM has built a platform that allows anyone, from the individual to the large enterprise, to use and adapt Watson’ capabilities. Watson brings with it analytics, reasoning and interactive methods (data, speech, text, vision etc.); enabling users and developers alike to execute, understand, gain insight and act on latent and active situations and associated data sets.

Of course, there are limitations. Watson is not a silver bullet to all problems, but IBM are doing a good job of extending its performance, reach and function. For example, they’re working with chip maker NVIDIA to accelerate compute performance, collaborating with companies such as Siemens to offer industrially cognitive (Buildings and Industrial IoT) analytics, and working hard to expose new APIs for partners (and other IBM entities) to build on.

Of course IBM isn’t the only company playing in the AI/Machine learning arena. Some cite estimates of over 3000 start-ups in the AI domain. Naturally one must add to these the likes of existing developers; the likes of IBM, Apple, Amazon, Microsoft, Google, Facebook and a multitude of other developers to complete the picture. In the world of engineering software (for example); Autodesk has recently partnered with Nutonian to embed the Eureqa intelligent modeling engine into their IoT platform Fusion Connect, and PTC has purchased ColdLight which contains elements of AI to augment their ThingWorx IoT offering.

In areas of industry and AEC markets, IBM’s comprehensive Watson IoT platform (and built on Watson) is ranked by many (including the likes of Forrester) as the best in the business. Watson IoT builds on IBM’s strengths, not just in areas of AI and analytics, but so too from IBM’s vast solution portfolio which include elements of (f/e) software development (Bluemix for apps and analytics, and Continuous Engineering solutions, namely Rational), security, communications and billing solutions.

On a final note, historically when one hears of IBM, some might consider that Watson is focused on the needs of the enterprise, but no. IBM’s clear goal is to make it ubiquitous. Developing solutions and traction in the ‘smaller’ end of the market is now down to the recently formed Watson IoT Consumer and Volume Offerings team. IBM, of recent times hasn’t been particularly visible at this end of the market so I, for one, will be keen to keep track of their developments in this space.