Artificial intelligence: a brave new world or the downfall of humanity?

For some, the rise of artificial intelligence (AI) is an exciting and captivating concept, for others it is not be trusted and even feared. But for many, feelings lie somewhere in the middle.

After numerous ‘AI Winters’ in the last century, the rise of ‘modern’ AI really started in 1997, according to Medb Corcoran, applied intelligence director at The Dock, Accenture’s global research and incubation hub.

“The key moments in the rise of AI, as I see it, can be classified into four types. First came the ‘AI showing superiority to humans’ moments – where AI beat world expert humans at games.

“Second came the ‘I didn’t think AI would be able to do that’ moment – the main one for me being the Google Cat Experiment in 2012, where Google researchers fed millions of unlabelled images to deep learning algorithms and, despite being given no identifying information about them, the AI was able to learn to detect pictures of cats. In the past five years came the third type, the ‘this will really change the way we work and live’ moments – such as the first driverless car and Skype’s launch of real-time voice translation.

“Finally, in the last five years, we have seen key warning moments of what could happen if AI is not used responsibly. In 2015, an open letter to ban the development and use of autonomous weapons was signed by Stephen Hawking, Elon Musk, and 3,000 researchers in AI and robotics. Many years have been spent learning about ethics and transparency when it comes to AI. Now the focus should be on ensuring unfair bias is avoided when AI is scaled.”

Clive Foley is lead technical architect for LiveTiles, a platform that builds intelligent workplaces. He says mixed feelings towards the advancements in robot technology are not surprising.

‘Talk to technology’

“We are beginning to talk to technology more than we type and press buttons. This type of interaction will give us an expectation that the machine will ‘always know the answer’ because we can ask it anything. This has the potential to both enable us but also make us lazy.

“The fact that the range of smart assistants that are out there including Siri, Alexa and Cortanta are humanised is no surprise. For these services to be successful, they need their customers to trust them. As we link services to these assistants, they will allow us to do higher and higher-level tasks.

“Inevitably though, they will learn more and more about us – what drives us, what we care about and what makes us happy, which is both exciting and scary in equal measures.”

So what can we expect from a future with AI at the centre of it? It depends what school of thought you subscribe to, Foley says.

“Some people envision a utopia where multiple super-intelligent AIs solve all of humanity’s problems and leave us all with plenty of time for leisure and allowing us to do what we want without the pressure of making a living. Dystopian people see the rise of AI as the beginning of the downfall of humanity.

“I hope that the truth is somewhere in between. Ultimately, any AI that is built has a goal. Aligning that goal with the goals of humanity should lead to breakthroughs and advancements that humans alone can’t achieve in any meaningful timeframe,” he says.

In the next few years, we will also see more collaboration between AI and humans, Corcoran says.

“Humans will augment machines – for example by teaching them to adjust to changes. And at the same time, machines will augment humans – for example by delivering individually tailored training to enable them to be productive quickly and stretching them to reach their full potential on value-added tasks and roles.”

We will also see the continued rise of the smart home, she says. “The intelligent voice-controlled assistant, such as the Amazon Echo and Google Home, is now a viable control centre for the connected home, and while today the focus is on solving simple problems, seamlessly connected homes are coming.”

Ronan Fitzpatrick director of digital at PwC says the adoption of “no-human-in-the-loop” technologies will mean that some jobs will inevitably become redundant, but others will be created by the shifts in productivity and consumer demand emanating from AI, and through the value chain of AI itself.

‘Creation of jobs’

“Along with jobs in the development and application of AI, the technologies will need to be built, maintained, operated and regulated. In most cases, these jobs will be conducted by humans. Furthermore, the extra demand generated in the economy, as a result of AI increasing output and incomes, will lead to the creation of jobs not directly related to AI in non-AI intensive sectors. All of this will facilitate the creation of jobs that would not have existed in a world without AI.”

He adds that given Ireland’s strong business and regulatory environment, we should expect that AI will also lower the number of jobs being outsourced, since its AI-readiness will mean jobs being outsourced can now be done more efficiently in Ireland. “Therefore, it is likely that the effect on jobs in Ireland in the long term will at least be neutral, if not net positive,” he says.

In spite of all of this, Fitzpatrick says Ireland is lagging behind the rest of the world in terms of investment in AI.

In a 2017 Digital IQ survey carried out by PwC, it shows that substantial investment in artificial intelligence by Ireland is low: 36 per cent compared with 54 per cent worldwide.

Irish PwC research suggests the lack of digital skills is holding back progress. For example, fewer than one in five (16 per cent) senior business executives said their AI skills were highly or quite developed. Fitzpatrick says a step-up in investment is also needed, as just 54 per cent of Irish executives plan to invest in AI, compared to 63 per cent globally.

Foley says for Ireland to become leaders in this space, we need to start asking ourselves hard questions and start preparing for the answers. For example, if an AI-powered self-driving car crashes, who’s responsible? If an AI automates a large number of jobs, how do we help our citizens who may lose their jobs? Would we be comfortable with an AI judge that determines outcomes of legal cases?

Extensive ecosystem in Ireland

Corcoran says we should continue to capitalise incrementally on the existing breakthroughs in machine learning by leveraging the extensive ecosystem in Ireland.

“Namely, all of the global top 10 ‘born-on-the-internet’ companies have a significant presence in Ireland, as well as the vibrant tech start-up scene and a strong set of academic R&D centres. Accenture’s international clients that come to The Dock in Dublin are consistently amazed by this ecosystem and how proximally located they are.

“But we should also aim higher – and look to take further advantage by achieving step change. This could be done through setting a grand challenge in this space – an ambitious but achievable goal that solves important national or global problems, and that captures the public’s imagination. Grand challenges channel cross-industry, cross-sector (commercial, academic, government) collaboration and can make step changes possible.

“Lastly, we shouldn’t get complacent – we need to stay ahead of the curve through continued investment in fundamental, as well as applied research, thus fuelling the pipeline of the breakthrough ideas of the future,” she says.

“It is clear the value of AI enhancing and augmenting what businesses can do is large, if not larger than automation. AI will likely be one of the biggest game-changers this century and is likely to transform our lives as individuals, business leaders and as a society,” Fitzpatrick says.