IBM – Watson the wanderer.

Watson goes nomadic.

• IBM has sensibly decided to allow its AI offering (Watson) to run on the cloud offerings of its rivals as well as enabling an ability to run effectively over multiple cloud instances at the same time.
• This is a good move which some may see as detrimental to its efforts to win share in the enterprise cloud, but I think helps underpin Watson as a leader in enterprise AI.
• Contrary to general opinion, Watson is not a digital assistant but a tool kit for companies to use to get up the AI learning curve much more quickly than they can on their own.
• It offers tools and pre-trained algorithms on standard data that can then be tweaked to the requirements of the customer helping them get up and running much more quickly.
• To date, Watson has only been available on IBM servers which is hideously limiting when one considers how weak IBM’s position really is.
• In hardware, IBM has a strong position but in servers these days, the hardware is incidental.
• What really matters is software and here it is effectively AWS way out front with around 32% share with Microsoft a distant second on 14%.
• IBM and Google are in 3^rd and 4^th position respectively with less than 10% share each.
• Both are struggling to make any real gains which is what lies behind IBM’s acquisition of Red Hat (see here).
• Consequently, only allowing Watson to run on IBM’s cloud is extremely limiting and precludes IBM supplying Watson to over 90% of the market.
• Furthermore, many companies are running hybrid cloud architectures where public and private cloud instances are stitched together to meet the company’s requirements.
• IBM Watson will now also support these architectures making it fully agnostic to the cloud supplier.
• This is a good move by IBM, as its cloud business appears to be going nowhere while the outlook for Watson remains quite good.
• Now that Watson can be used by enterprises without having to use IBM’s cloud, its appeal should take a big step forward.
• Although IBM is not often to be found at the cutting edge of academic AI, it has a very good offering for the enterprise and a slew of deals for Watson should follow this announcement.
• This means that Watson’s real edge is in its long experience in this area meaning that it has been working on AI long before Google, Baidu, Yandex, Facebook etc were even founded.
• RFM has found that time spend crunching data is the best indicator of the quality of one’s AI which underpins the view that IBM is in the top tier when it comes to the enterprise.
• Consequently, this is a good move and should help IBM cement its position as a leader in enterprise AI.
• Google, despite its prowess in AI, has really not made any meaningful impact in AI for the enterprise, leaving IBM’s biggest threat as Microsoft.
• Microsoft is now pretty much focused on the enterprise and it has a reasonable AI effort in place which it will obviously be pushing Azure customers to adopt rather than Watson.
• RFM rates Amazon’s AI as weak despite its position in smart speakers, meaning that IBM’s biggest opportunity is probably to target AWS customers.
• This is a net positive for IBM and a negative for Google which really needs to up its game in the enterprise cloud.