Bigger, better, stronger, faster: How IBM’s Watson upends Moore’s Law
This article was first posted on 27 June 2013.
This InfoWorld article on IBM's Watson is absolutely worth a read. If you haven’t already heard about IBM’s Watson you should take a few moments to familiarise yourself with what it’s all about (you can see more at this link – The science behind Watson). Watson is at the very cutting edge of what IBM calls “cognitive computing” and, like Google’s work on artificial intelligence, it will almost certainly be a massive game-changer. My personal view is that, for the most part, the magnitude of its potential impact is still not widely understood. In that context, this article is worth a read because it provides a couple of hints as to where Watson is headed.
Perhaps the most significant pointer to where IBM’s Watson is headed is the announcement of “Watson as a service.” Rather than forcing organisations to buy their own individual installations of Watson, IBM will be hosting Watson “in the cloud”. This is a major development in the evolution of Watson because it is likely to make it easier for organisations to access Watson and, as a result, accelerate take up rates and market penetration.
The second pointer is the clear intention to give end-user customers direct access to Watson via things like web chats, email, smartphone apps, and even SMS. Not only is it highly likely that Watson’s cognitive computing power will be available to consumers via their smartphone, tablet, laptop etc., it’s also likely that we’ll start to see large organisations providing that access within the next six to twelve months.
The third pointer actually comes from a link within the article back to IBM’s commentary on Watson at your service. Consistent with the theme of exponential improvement in computing, IBM have announced that, since it won the Jeopardy Challenge in 2011, Watson has enjoyed a 240% improvement in processing power and a massive reduction in physical size. The progress made in reducing Watson’s size is staggering. Where the 2011 version of Watson was built on 90 IBM 750 servers and was the size of a bedroom, the 2013 version is built on a single IBM 750 server the size of four pizza boxes.
The final, and perhaps most significant, pointer also comes from the link to the IBM website. The pointer is that IBM sees Watson as being a cognitive computing assistant capable of delivering insights based on unstructured data – just like people do. You need to think about this in the context of the hierarchy of knowledge shown in the graphic below.
You can argue about where traditional internet search fits on this hierarchy but there’s little argument that it sits towards the bottom half of the hierarchy. Importantly, cognitive computing starts to shift computing capability into the upper half of the hierarchy. When combined with 24/7 access and the ability to interact in natural language that becomes a powerful proposition and a real threat to the core business of every organisation whose core value proposition is the delivery of knowledge or insight. IBM is already well-progressed in the application of Watson to medical services and it’s not a big stretch to see it being applied to a very wide range of industries built upon the delivery of professional advice.
There are few industries that won’t be touched by this technology. It’s coming fast, it will shake the foundations of all knowledge-based industries and it will be within reach of ordinary people all over the world. If that isn’t disruptive then I don’t know what is.