machine learning in It

What’s the latest thing to perch hopefully at the beginning of the Gartner hype cycle? Algorithmic business. Tomorrow’s companies are going to rely on algorithms more than ever before, the firm said in its latest research note, which could lead to business models that morph and change automatically. It all sounds like something from science fiction, but what does it really mean?

An algorithm is simply a set of instructions to follow when completing a process. Every piece of software is algorithmic, meaning that business has been using algorithms since the launch of the LEO I, so we might well ask why the term is being bandied about so breathlessly now.

Whereas businesses used software logic in discrete ways to automate certain processes, in the new model algorithms become a more central part of the business, making decisions that couldn’t easily be reached without them, and then even taking action on them automatically. Algorithms may have invoiced our customers and sent orders to our suppliers in the past at our behest, but now they’re set to find things that we don’t see, and possibly do things that we’re not explicitly telling them to do.

A brave new future?

We’re already seeing this in firms like Netflix, which use algorithms to tell us what TV to watch, and even to tell them what shows to buy and make. Gartner envisages algorithmic utopia in industries like retail, where companies could optimize pricing, perhaps in real time, based on data from various sources. How about digital labels that automatically adjust the price of water based on a mixture of temperature and historical regional sales data, on an hourly basis?

In recruitment, algorithmic business would take software far beyond the typical, ‘dumb’ keyword matching software used by many businesses, instead using smarter algorithms and data feeds to find people who are more likely to be looking for a new job than others.

Algorithms need a large amount of data to inform those decisions, which means that big data is closely tied to the concept of algorithmic business. Businesses with mounds of historical and real-time data stand a better chance of feeding the algorithms that can make these decisions and execute on them. Microsoft’s acquisition of LinkedIn at a 49% premium is a good example of an algorithmic business play, said Gartner analysts, and that rings true. It could use the firm’s 430 million-strong social graph to inject new, smarter functions into its software (how about recommending articles or even experts who might be able to help you with a project you’re working on?)

The ghost in the machine

AI and machine learning will naturally be a part of all this, as the kind of business smarts we’re talking about here aren’t always deterministic. They’re probabilistic, and fuzzy, just as people are, which might be a good way to describe the basis of an algorithmic business going forward. One of the key goals for Gartner’s new paradigm are businesses that use software to achieve more human-centric outcomes. You won’t have to work hard to understand the computer software, it believes. Instead, it’ll be the other way around.

It’s a great future that they’re painting, but in truth the road forward is littered with traps. Machine learning based on big data is fraught with problems. Just ask the people who programmed Microsoft’s Tay, the ill-fated online bot that started spewing racist, drug-laden diatribes online because Twitter users taught it to. Or the folks who programmed the Google AI engine that made its embarrassing gorilla gaffe last year. Autonomous algorithmic trading has crashed markets. Software isn’t god, no matter what technocrats in the valley believe.

We only have to look at this week’s hacking of smart contract on the Ethereum-based DAO, which we could define as an algorithmically-governed business, to see how easily things can go wrong. A piece of blockchain-based code used by the system was compromised, and tens of millions of dollars-worth of cryptocurrency stolen.

None of which means that smart contracts, machine learning, or big data are bad things. Used properly, they can all be useful weapons in a business’s arsenal. Technology is a great place for imagining bright new futures, especially ones built on decade-old concepts. But let’s not put on the silver space suits just yet. Gartner wants us to create markets for algorithms that enable us to automatically hire them based on the job we want them to do. That’s all very nice, but we should probably work on getting Windows 10 not to ask for updates at inopportune moments first, and then put one foot slowly and sensibly in front of the other.



Related Download
A Simple Powerful and Practical Analytics Solution For A New Generation of Users Sponsor: IBM
A Simple Powerful and Practical Analytics Solution For A New Generation of Users
Download this report for an in-depth look at the capabilities and features of IBM Cognos Analytics.
Register Now