Getting smart about AI

null

It’s no secret that the prominence of Artificial Intelligence (AI) in business has grown dramatically in recent years, so much so that it is now one of the key technology buzzwords around.

The rise of cloud computing and open source initiatives have helped to foster the rapid development of these headline-grabbing, next-generation technologies, which are now the subject of significant investment from small businesses and enterprises alike.

For example, Deloitte’s Digital Disruption Index revealed that 85 per cent of senior executives plan to invest in AI by 2020, while Stanford University’s AI Index suggests that there has been a six fold increase in the annual investment levels of venture capitalists into AI start-ups since 2000, with significant investment acceleration after 2010.

If that’s not enough, Gartner predicts that the business value derived from AI initiatives will reach $3.9 trillion by 2022, up from $1.2 trillion in 2018, all of which highlights the massive potential behind the technology.

Clearly, the AI trend is showing no signs of slowing. In a world being increasingly eaten by software, AI now provides the fuel to help businesses grow, innovate and provide new services to customers.

But, although no industry is seemingly immune, people have quickly realised that implementing AI isn’t as easy as simply switching on a new tool. There are still some factors for businesses to consider before they can start reaping the rewards.

Not all plain sailing

Despite the hype, research has shown that implementing AI-based solutions hasn’t been as easy as businesses hoped. From difficulties around integrating AI tools with existing products and systems, to a lack of understanding of how the technology works and concerns around expense, the challenges are widespread and complex.

According to a recent report from Databricks, for example, 96 per cent of organisations are running into data-related problems such as inconsistent datasets, while 80 per cent cited a lack of collaboration between data scientists and data engineers.

Then there’s the issue of compute power. AI systems typically use a huge amount of processing power, which is only going to increase as data volumes continue to grow and the algorithms powering these systems become more complex, thereby presenting significant scalability concerns.

It’s also important to remember that, from a practical point of view, the technology is still in its relative infancy and is evolving rapidly. AI has been talked about for some time, but it’s only in the last few years that the rate of actual deployments has taken off.

Challenges can even arise before the implementation phase. One of the biggest AI obstacles for IT teams and business leaders to overcome, is understanding how to implement it in ways that solve real business problems and best suit their organisation’s specific needs. Rather than rolling out AI for AI’s sake, businesses have to consider where AI is likely to have the biggest impact and what specific processes could be automated and transformed.

This can be easier said than done. Although AI touches virtually every business on the planet, people with the knowledge and expertise to turn theory into tangible, profitable outcomes are often hard to find – and extremely expensive.

What’s more, AI doesn’t just refer to one technology. It can encompass a range of processes, including the likes of machine learning, data transformation, model creation, natural language processing and deep learning. Understanding the differences between these various innovations and how they could fit into an enterprise’s infrastructure is therefore essential to getting the most out of them.

So, what can businesses do to solve these issues and realise the potential of artificial intelligence?

AI all the way

With these challenges in mind, capitalising on the AI trend comes down to a few key factors. Firstly, it’s vital that businesses recognise the importance of deploying back-end infrastructure and systems that are able to support the compute-intensive tasks involved in AI and machine learning.

Operating systems have to be tuned for these advanced workloads, thereby enabling businesses to work with vast datasets, deploy applications at scale and cope with the complexity that will be created.

Secondly, businesses can’t afford to ignore the people problem. Getting AI-powered systems up and running takes a huge amount of time, effort and expertise, resources which not all organisations have at their disposal. After all, AI systems are only as good as the people who program them. The current industry skills shortage therefore has the potential to hold businesses back if they don’t have access to the right skills. That’s why partnering with expert companies that are able to guide them along the journey is essential to plugging any internal gaps.

Most importantly, enterprises have to be smart about how they deploy AI-based technologies. Rather than simply diving in headfirst, companies should think about designing a long-term strategy and investing in people with the relevant skills and experience.

Ultimately, there can be no doubting the fact that AI presents a major technological opportunity over the coming years. Harnessing its power may be easier said than done but using it in the right way and making sure the focus stays on solving real business problems will go a long way towards helping businesses achieve their goals.

We all know that artificial intelligence has been continually getting smarter for years. Now it’s time for businesses to follow its lead and start getting smart about AI in order to realise and embrace its true potential.

Carmine Rimi, Product Manager, Canonical
Image Credit: Geralt / Pixabay