Threequarters of Irish businesses are at risk of being outmatched by their European counterparts due to their lack of artificial intelligence (AI) integration.
A new study by Microsoft has found that while three quarters Irish businesses are in the planning and piloting stage of integration, full implementation is yet to occur. This creates the fear that there will be a lack of investment and buy-in from a board level.
“AI encompasses lots of different things. It’s not just robots; it’s also computer processing, so in many instances what is understood as ‘AI’ is actually ‘machine learning’.
Pretty much everything we do now in our research area (human language technology) is data-driven, where the machine makes inferences (about what sentences mean, or how to translate a sentence, or what a user wants to see given a specific query in Google, say) from large collections of data,” said Professor of Computing, Andy Way.
Way said that these systems been shown to work best on a wide range of tasks and that AI systems outperform traditional machine systems.
A common example of AI in businesses are chat-bots. European countries have incorporated these systems into their customer service, often serving as the first line of contact for customers.
Across Europe, 65 per cent of organisations expect AI to have an impact on their core business in the future. About half expect it to impact areas that they haven’t even considered. However, only 4 per cent consider themselves advanced in this area.
“So, if we interpret the Microsoft study in this light, I think it is probably credible that three quarters of Irish businesses have indeed integrated AI. If we’re talking machine learning, that’s relatively straightforward; if it’s replacing boring human tasks with robots, clearly that’s much harder,” said Way.
The exponential development of AI has been influenced in large part by smartphones, wearable devices, the internet and the power of the cloud that helps harness that data and make sense of it.
Humans are crucial in the development in AI, with research indicating that companies that put an emphasis on emotional intelligence have better success in adopting AI.
“One issue though, even with machine learning is that today’s deep learning approaches require GPU architecture rather than CPU, which are expensive, and the models take longer to train, so the architecture is more expensive and there is undoubtedly an impact on climate as these models take a long time to run, so more computer power in both senses of the term is needed,” said Way.
David Kelly
Image Credit:Pixabay