More than just a buzzword - what really is AI and how should it work in your business?

Buzzwords like “AI”, “machine learning”, and “big-data” are thrown around with reckless abandon by start-ups and multi-nationals alike, with promises to revolutionise cancer diagnostics and – at the Hollywood end of the spectrum – spur the rise of a dystopian society where robots replace large swathes of the human workforce or worse. Amongst all this noise and hype it is essential to get back to the basics of AI and understand what it is and how it can drive real value. 

AI isn’t magic or the dark arts. At its most basic, AI is simply taking vast amounts of complex data, identifying patterns, and pointing it at a specific problem to solve. 

That ‘problem’ might be detecting fraud at a bank, powering driverless cars, or your preferred personal assistant technology, from Siri to Alexa. All real-world applications that AI is powering today.

So while the application of AI is indeed more complex today than at its birth, the tools and technologies that are powering AI have existed for half a century. It’s the underlying compute power that has increased exponentially, far outpacing humans’ ability to create the underlying software of the same complexity.

As Afiniti CEO Zia Chishti puts it, “we have not moved a byte forward in understanding human intelligence.” While our computers are faster thanks to Moore’s Law, the algorithms of today are largely identical to those powering machines 40 years ago. 

From VisiCalc, to Excel to a company like Afiniti today, the complexity of the underlying software of these technologies is almost identical. It’s the increasing processing capacity of the hardware that’s the most significant driver in AI’s advancements.

And rather than a future where cyborgs are babysitting our children, the best application of AI today doesn’t replace human jobs, it enhances them, improving productivity and efficiency. 

AI expert Kai-Fu Lee argues that “the age of ‘artificial general intelligence’ - or when AI will be able to perform intellectual tasks better than humans - is far in the distance”. 

Humans remain and will remain for a long time, far better at solving problems that are native to humans. Take visual processing for example. Without a structured problem, a computer couldn’t walk into a room and tell you what it saw. Whereas a human might recognise a familiar face, see one red tie in a sea of blue, without a question to answer, a computer simply couldn’t respond.

AI shouldn’t be viewed with awe and glazed eyes as someone pitches “neural networks” and “Bayesian inference”. Think of AI as a powerful yet practical, advanced business tool that can work for you to find business optimisations and give customers a better experience more efficiently. Some of the most powerful applications of AI empower and free humans from low-level tasks allowing them to be more productive and impactful at their primary job. 

So with AI demystified and the myth debunked, how can corporations best benefit from AI?

Firstly and chiefly, does the CEO understand the value the AI technology can deliver? Is the business case exceptionally clear? If not, then it’s likely the project isn’t worth the time. 

Secondly, AI projects that often reach across an organisation, absorbing massive amounts of data from every corner of a business, must be driven from the top. CEOs must both understand the benefits and drive the benefits. Afiniti’s lived experience, is that delegating is almost certainly a recipe for failure. 

And finally, beware of marketing gimmicks and hype. Organisations must focus on two fundamental questions. One, what problem does this solve for? And, how can I measure the outcome? Both these answers must be immediately apparent to any leadership team adopting AI within their business.

If you can reduce the puzzle to something specific and quantifiable, that’s the correct use of AI today. We’re teaming up with Avaya in late February at Customer Contact Week to analyse the puzzles that Australian organisations are currently solving, and exploring practical strategies not just to implement AI, but do so in a way that yields outcomes for the benefit of their customers. 

The problems that are yet to be solved by AI are as endless as they are varied and I remain thoroughly optimistic for the future. And for now, I’ll be optimistic, but realistic.