IT is no longer a black box

At the recent NICTA TechFest, held at Parliament House in Canberra, we had the opportunity to speak with Professor Robert C Williamson, the leader of NICTA’s Machine Learning team, about some of the big innovations he thinks CIOs need to keep their eyes on.

Big data is here and many CIOs are grappling with making meaning and gaining context from the massive amounts of data they have to collect, store, manage and analyse. Williamson explains that machine learning is the science of big data.

“At the moment you need a couple of PhDs and a lot of money to make it work. We believe it’s the kind of technology that can be ultimately pushed out there. An analogy is simple matrix mathematics, which is something many people struggle with. Once it’s embedded into a spreadsheet everybody’s using it.”

The challenge he sees is in democratising big data and simplifying how we can derive value from the data. Williamson asks, “How can you get your hands on things that you can only understand in a very poor fashion? We try to apply machine learning to natural language documents. You can’t get a computer to understand natural language like a person can.”

The second area that Williamson believes is of concern for CIOs is “the whole raft of things surrounding privacy and security”.

Over recent years, legislative and regulatory changes to businesses have landed in the CIO’s lap. Companies are looking for systems that embed those requirements into applications, hardware and data management in order to manage the organisational risk profile.

Privacy is connected to the big data challenge. “There’s empirical evidence for this. When the average person starts to realise the extent to which companies do data mining, whether it’s the Googles of the world or banks or social networking sites or supermarkets - once they understand that and what they do with that - ultimately what they do is manipulate you with it.”

Williamson said, “The challenge is that there’s many things you can do that are perfectly fine for the business and perfectly respecting the human that is your customer. The challenge is figuring out how to do that. At the moment it’s highly asymmetric. The power is all in the hands of the corporations with the machines and the computer scientists. If people start feeling exploited  - there’s just got to be some cause célèbre for this - people will rebel”.

Increasingly, companies will need to ensure that they audit their activities and that the appropriate checks and balances are in place, according to Williamson. So, tied into the capabilities of systems will be the need to put the right compliance measures in place to make the relationships between business, customers and data more equable.

One thing that’s becoming increasingly obvious and that IT leaders need to consider is that IT is not a black box hidden in the basement.

Citing examples such as the fact that 98% of CPUs are embedded and that 25%  of the cost of developing a new BMW is in software, Williamson drives the way IT can no longer be considered an isolated function but that is part of the value cycle for businesses.

“The way I think ICT has been thought of for many years or decades is people talk about ‘the ICT sector’. There are books coming out and people are now thinking this is the wrong way to categorise it. It’s a way of thinking about the world. Information touches everything. It’s hard to imagine an industry, a business, a human problem that information does not impinge upon in some way. Rather than thinking of IT as a separate thing, it’s embedded,” said Williamson.

Anthony Caruana travelled to Canberra as a guest of NICTA.