The recent 4IR debate held last month in Derry explored the challenges and opportunities presented by AI – is it priceless or perilous? Emily McDaid reports.
The Fourth Industrial Revolution Challenge explores the disruptions of artificial intelligence (AI) to areas as widespread as health, energy, food and transport.
Through our new series of high-energy debates, Connect will be exploring this disruption.
The inaugural event occurred on an unseasonably warm evening in Derry on 25 May. Around 60 attendees hailed from large corporations such as Allstate, Fujitsu and Seagate; academia such as Ulster University, Queen’s University Belfast and Letterkenny IT; and smaller start-ups, including NeuroConcise.
A Twitter poll taken at the event asked attendees if AI was priceless or perilous. At the time of writing, 87pc believe it’s priceless.
The first discussion was: ‘How AI will impact the everyday experience of using a sandwich shop’, by Ciaran Murray from Creative Metrics, and author for Pure Derry and The Ulster Fry; and Mark Nagurski, co-founder of MakeMatic and founder of CultureTech.
Murray and Nagurski explored the example of a 35-year-old man named Tom buying a sandwich, and how AI could impact his decision-making.
AI can understand whether Tom is on a diet. Does he go to the gym? Does he wear a Fitbit?
If Tom goes on a diet, his AI could disable his bank card in McDonalds, decide which healthy café he should walk to (for more exercise, he can walk to the farther café), or pre-order a healthy salad and have it ready for his lunch break.
Then, he might Instagram it, or review it on Yelp. Maybe he doesn’t need to tell the app his review, because it can sense how quickly he ate the sandwich, how much he salivated, how many sandwiches he ate. Intense filtering will then occur; the one-star sandwich shop won’t show up on his results anymore.
Do we need the sandwich shop or could the best sandwiches be made in a factory outside the city and delivered to you?
In that case, do we need brands? Do we need advertising? Decisions in economics, insurance, your health and exercise can all be taken by the AI app.
The next talk was given by Ben Greene, chief technology officer at Analytics Engines, giving the view of the sandwich shop owner.
Greene said: “My personal view on AI is, at the moment, I have to think about stuff. Can I lift a little bit of that out and put it in the computer to make smaller decisions I don’t want to make?
“I’m good at some things and bad at others. Can I get an upgrade at the things that I’m not good at?
“What kind of stock do I need? What will I make money from? Maybe I’m actually making money from outside catering or other areas, not from the sandwiches themselves.”
Greene said AI can help the sandwich shop owner with the following decisions:
- It’s all about the customer experience. The length of queue is important – not too long or too short.
- What motivates my customer? What’s my target market?
- What are my risks (eg everyone going gluten-free)?
- What does the owner do if the AI engine is now making all the decisions? Is the owner now the robot making the sandwich?
- What about the location of the shop, legal services, health and cleanliness?
Prof Damien Coyle of Ulster University presented the challenges 0f AI.
“There are three main perils of AI: jobs displacement, disruptions to businesses and controversial robots.
“Artificial general intelligence (AGI) is when AI can carry out any cognitive function a human can – we’re not there yet. We are still at narrow AI, where you can train a computer to improve a process for one specific purpose.
“We can then get to super-intelligence, when one AGI builds on more AGI.”
Coyle added: “At this point, AI is here to stay. It has crossed a threshold now and become mainstream – Netflix, Amazon, Google and supply chains are all dependent upon it.”
He showed a graphic depicting how quickly AI will grow. “With exponential growth, you don’t know about it until you hit that steep curve and there’s very little time left to prepare. AI brings change and it makes people and companies uncomfortable.”
Coyle discussed the question of jobs displacement. “Up to now, machines taking the place of human labour has been slow and people could skill up, but if it speeds up, there won’t be time to retrain.”
He showed two waves of jobs displacement in AI:
- low-skill jobs in transport and administration
- the heart of upper-middle class; careers such as legal, healthcare
Coyle concluded: “Will this transformation be utopian or dystopian? The message is to plan, predict, adapt and evolve.”
Presenting on the optimistic side of AI was Fujitsu’s senior innovation bid manager, Dr Gráinne Watson.
“Already at Fujitsu, we have made strides in using AI for very real purposes. In Spain, we reviewed, using algorithms, patients’ mental health data using AI. Using AI-driven analysis, we’ve been able to prevent a few suicides.
“Other innovations in healthcare will evolve with AI. Nanotech linked in with detection systems such as Fitbits could present automatic treatments for symptoms such as asthma,” she said.
Changing topic to smart cities, Watson discussed how AI could alleviate congestion. “Information on traffic can be pushed to your phone; infrastructure projects can identify bottlenecks more easily.
“AI is here to stay. No one has gotten this right yet, so why wouldn’t we give this a go?”
‘You can either get on the AI train or it’ll hit you’
– BEN GREENE
The event concluded with a panel discussion, where each speaker summed up their thoughts.
“Is the AI machine making decisions preferable to a human making a decision?” asked Greene.
“This is something we cannot stop. Why don’t we go in hopeful?” asked Watson.
Are driverless cars safe?
“The future of air travel is a pilot and a dog. You need a pilot or people won’t get on the plane. You need the dog to keep the pilot away from the controls because he’s the most dangerous thing on the plane,” said Greene.
What about job loss? Could we work alongside AI?
“If you lose jobs in one sector, you gain them in another,” said Murray.
“I can’t see any job surviving, apart from ones that specifically involve human interaction. I don’t think it’s a question of what can AI do but what do we want AI to do?” said Nagurski.
An audience member added: “AI could never replace the human psyche – the emotions, the smiles – we will miss out on so much.”
By Emily McDaid, editor, TechWatch
A version of this article originally appeared on TechWatch