We’ve all seen the pop culture depictions of AI...
In a nutshell
- The majority of the world’s largest companies are investing heavily in narrow AI applications. While most companies cannot afford to compete with this level of investment, AI remains something Kiwi businesses need to be thinking about and creating plans for at any level.
- Marketers and data analysts are beginning to come together to realise the potential of AI to save time and costs, radically improve the customer’s experience, alter how businesses operate, and develop entirely new products and capabilities.
- Prescriptive analytics is a developing field that uses data to advise on the future rather than telling us about the past.
... A smarter-than-human robot set on over powering us, often in some sort of dystopian near-future.
It’s therefore quite understandable that AI has become feared. And this is only emphasised by the rate of change we’re seeing, a factor that makes AI seem beyond our control. But this fear is stopping many from being open to the possibilities that AI holds to benefit business and our broader lives.
The AI we see presented in sci-fi films like I, Robot and The Matrix is artificial general intelligence (AGI) where a machine can ultimately perform any intellectual task that a human can and often achieves consciousness. As well as taking on human intelligence, this type of AI is generally presented in pop culture with human physical characteristics too. However, we can breathe a small sigh of relief because there’s little evidence to suggest that AI development will take this route.
In reality there’s much more development happening in the space of ‘narrow AI’ – artificial intelligence focused on a single task, enabled by the computing power and data we have at our fingertips.
All applications of AI that we currently see are narrow AI including chatbots, an example that most people would be familiar with. Chatbots range from anything like an online chat or troubleshooting service to more complex applications like virtual assistants Siri or Amazon Alexa. Recently Google demoed Google Duplex, a chatbot voice that is nearly impossible to distinguish from a real human. Google Duplex adjusts to the person it is speaking with, uses phrases like “um” and “ah”, can handle interruptions and elaborate on its points. People can use Google Duplex for things like booking a hair appointment on their behalf.
Keeping up with change
The majority of the world’s largest companies are investing heavily in this space. The likes of Amazon, Google and Facebook have baked analytics and AI into the very fabric of their product, and have research labs dedicated to making developments.
While most companies cannot afford to compete with this level of investment, AI remains something Kiwi businesses need to be thinking about and creating plans for at any level.
If we’re to believe Moore’s Law (which says that computing power doubles every 18 months) there will come a point soon when the growth curve of computing development becomes vertical, where you can teach a model to build a better model than itself, and so on. Some experts believe this will happen in 2040 – within our lifetime the world will be a very different place.
This isn’t in a dystopian future way like the movies would have you believe. This is about solving real problems and making our lives easier.
Bridging the gap
"We’re at a point where marketers and data analysts are beginning to come together, and this is where we’re seeing things take off."
We’re at a point where marketers and data analysts are beginning to come together, and this is where we’re seeing things take off. Up until now analysts haven’t really been thinking about how exactly you apply AI to business. And at the same time marketers haven’t understood a lot of the potential that data holds.
A lot of marketing organisations, even the most advanced ones in this space, have been thinking of analytics and AI as marketing automation – better targeted ads to improve click through rates, automating email sequences, or mapping their segmentation to a Facebook audience. But the potential of AI is about more than creating incremental improvements to marketing efforts. It’s about saving time and costs, radically improving the customer’s experience, altering how our business operates and developing entirely new products and capabilities.
Retail businesses, for example, can begin using AI to optimise the range of products they have in store, how those products are laid out, what their pricing and promotion strategy is, and where the best place is to build their next store.
Data that tells us what to do
Much of this potential falls into the realm of prescriptive analytics – analytics that can advise on the future rather than telling us about the past, as most descriptive analytics currently do.
This likely sounds a bit of a stretch to many, but the tools to do this are currently developing. Content generation, for example, is a regular and necessary evil for many marketers and one that many would appreciate being taken off their hands I’m sure. As the systems evolve, we can expect to see prescriptive analytics manage the process of content generation from writing, to posting and optimising.
We’re currently doing work in this space with Paydar, our analytics tool for retail and hospitality businesses. In the evolution of prescriptive analytics, we’re currently able to tell our clients what the key factors are driving their business performance. With a bit more development, we’ll be starting to identify what a business should do to capitalise on that insight. From there we’ll get to the point that Paydar can automatically control for factors, for example adjusting the pricing on a retailer’s website and assessing the impact of the change through reviews and feedback.
It would appear the future of AI is more promising than pop culture would have us believe.