On the social front, things are even more shaky with executives fearful of the power of negative commentary. It’s not unusual to hear business leaders saying, “we can’t have a twitter, linked-in or facebook account because someone might say something bad about us”.
Guess what. If people want to say bad things about your company in a public forum, then they most likely already are. By denying the company an official platform, you’re not denying your customers a chance to speak about you, you’re denying your company the opportunity to provide timely and “official” responses..
Like it or not, the future is cloud, the future is social. We have to embrace it.
On Prem vs CloudUnless you're in a company with military-grade security requirements, the choice of "on-prem" versus cloud needs to be more than a vague suggestion of security. In IBM Connections for example, there’s a very compelling reason to choose the “on prem” (on Premise) version over the cloud version. It’s to do with customisation as the on-prem version can be customised much more that its cloud-based counterpart.
Of course, the on-prem solution is also much more expensive. Not just in terms of licensing but overall. If you host your own connections system, then you’re taking onboard the costs of running hardware, keeping firmware, drivers and software up to date, performing backups and failovers and tests, and of course, troubleshooting.
Unless you really can’t live without the customisations that the on-prem solution offers, you’d be far better off letting the cloud take care of things for you.
This… is the InternetThere was an interesting set of questions in one of the IBM verse sessions at IBM Inform 2016 in Sydney (and I'm sure that similar scenarios were playing out at Inform sessions globally).
The presenter asked how many people were using Connections.Cloud and Verse, then asked how many were waiting for the “on Prem” versions because their companies were too scared to put their data in the cloud.
The overwhelming majority of attendees fell into that category.
At the end of the session, where the IBMers demonstrated many of the amazing features of Verse, they opened the floor to questions.
One of the first questions was;
“Since my company is planning on staying with the on-prem versions of this software for security and privacy reasons, what is the deal with getting access to IBM Watson?”
I was somewhat flabbergasted, as was half of the room and, it seemed - and so were the IBM presenters. They handled the question far better than I would have though. In my case, scenes from the IT Crowd flashed through my head -- particularly the episode where they give their boss “the internet” to take and show people.
I could just imagine them giving this guy a box with perhaps a plasma ball in it and saying, “this is Watson, take good care of it”.
Of course the presenters said, “Well obviously we’re not going to be able to give you Watson to install at your site, I imagine the data will need to be sent offsite”.
Having your cake…Now although I'm having a bit of a laugh at someone’s expense, it was by no means a dumb question; merely an interesting one -- and one which makes a great example to use when explaining the boundaries of these “on-prem” solutions.
The professional asking the questions, followed his question up with a few more about the safety of data in transit and the ability of other IBMers to intercept and read the data being sent to Watson but in the end, I felt that he more or less gave up. IBM didn't have the answers he wanted. It’s far too early to make calls about the accessibility of Watson in on-prem solutions.
What I felt didn't get explained very well was that Watson doesn't operate terribly well with a single piece of data. In order for Watson to work effectively, it needs to be exposed to all of your data. This is what helps it to spot trends and themes and provide real business-value to the organisation.
Sending a random email off to Watson uses some basic capabilities but not full-fledged AI. It may be able to translate the document into another language but can't we already do that with Google Translate? It doesn't mean that Translate is an AI, it's merely a very good interpreter.
In the end, you really can’t have your cake and eat it too. You can’t have the privacy offered by on-prem solutions while also having the benefit of systems which rely on significant levels of access to your data. You have to start trusting somewhere.
It’s clear that the next frontier (after social computing) is AI. IBM has Watson, Microsoft has Cortana and Tay (In Tay’s case it’s probably A without the I), Google has Google Now and Deepmind and Apple has… well, Apple has Siri (and perhaps Emotient).
If you’re planning to stay ahead of the competition, you need to be prepared to use AI. If you’re planning to use an AI from the market leaders, you’ll need to decide who you want to trust. Right now, my best guess would be IBM’s Watson.