Interview: Why “the stars are aligned” for AI and it can’t be ignored
Posted: 17 May 2017 | By Darcie Thompson-Fields
Repro Smoking Spaceman Robot – Ha Ha Toy – Silver - Close Up 1
Hi Philip. AI and machine learning have become huge buzz-words over the past year-18 months. But haven’t we been here before?
I started as an analyst 20 something years ago, and AI was the big topic then. But then it all went away. Last year marked 60 years since AI first came about, so what’s different today is a really good question. There are probably three or four factors which have all come together to make it happen. One of the biggest changes is the availability of the cloud. It allows access to very large processing power. So, all of a sudden lot of people can use vast amounts of computing power at reasonable costs to run these neural networks.
“Someone said ‘data is to AI what food is to humans’. That’s an interesting thought.”
Another is the availability of large amount of data. Someone said ‘data is to AI what food is to humans’. That’s an interesting thought. The cloud and the internet have made it possible to connect this vast processing power and vast amounts of information. That’s what’s made it possible to deliver on the academic and intellectual breakthroughs of the last 50-60 years. Another factor is the presence of these very big companies like; Google and Facebook, Amazon, Microsoft and IBM who have the resources to throw at the problem.
So, no chance of another AI Winter?
I don’t think so. The stars are all aligned and we are now seeing some real results. I read an article on how Google Translate has made more progress in the past 18 months or so than it did in the previous five years. The results are there to be seen.
Where will AI have its biggest impact?
It will have a transformative effect on businesses. There are so many examples. Take someone who replenishes vending machines. Traditionally a driver would be given a variety of chocolate or drinks, drive around on their usual circuit and drop them off with very little information on what to expect before doing so. If everything is wired
up with sensors, the driver will be able to get data on stock quantities before required before they leave. So, they may be running low on Diet Coke compared to full fat Coke, so they don’t need to load up their van with something which is ultimately going to come back with them. Likewise, they won’t run short of something and have to come back. This can all become an optimised process.
Another example is with Amazon in the US for its grocery business. The stock is stored all over the place. It’s not like a supermarket where everything is in specific aisles. The computer knows where it is, and will compute and calculate the optimal path for one of their staff to go and pick someone’s order. They say that saves them time and makes the process better. So, if someone orders some ice cream and some cleaning fluid, they’ll be sent to an area where the two are closest rather than going down one aisle and then going all the way to another. It’s created a new way of thinking and boosting productivity.
Are there any industries that won’t be impacted by AI?
It’s hard to think of one which won’t have an impact. There are personal services, like hairdressers or nail bars which may not necessarily need it in the short term. There will be clearly more impact on some more than others, but it’s hard to think of any that it won’t at some point have an impact on. Possibly one day we’ll have robot hairdressers.
What do you see as the typical benefits of AI?
I’ve not come across too many examples where people have said they don’t need people any more. It’s more about how they’re under pressure and would like the people they have to be able to do something more useful instead. There was the story out of Japan, but that’s rare. It’s more about companies wanting to do more, without bringing in extra people.
A lot of what I’ve seen is about predicting things such as outages and maintenance. Quite a lot is also about driving new business. So, more targeted emails, better recommendations on a website to encourage people to buy more. Better customer services and scheduling. There are also many examples of other ways that AI is being used, which has nothing to do with staff matters at all. For example, there is a supermarket which uses AI to manage what and how many sandwiches it needs. This is designed specifically to help fulfil demand and reduce wastage. It cut down waste by around 80%.
Is it important for businesses to start thinking about AI now?
In a word, yes. Our advice to people is you should at least be looking at this now. Definitely this year. It’s crucial to look at what’s available, how can it be used, where can be of benefit. It might be they just need to upgrade to the latest version of, for example, SalesForce because it includes intelligent sales recommendations. You can’t ignore it. The phrase we use is AI everywhere. 40% of all business transformation will include elements of machine learning and or AI. Any business of any scale, small or big, they could end up looking very last century. It’s also easy to use and incorporate.
Why do you think there is still an element of fear around AI?
Some thing’s are based on fiction. We’ve talked about systems that augment or potentially replace human decision making. Then there’s the idea of a general AI, a general conscious being. That is something which potentially would worry people, but that’s so far from today’s reality that I don’t give it any credence.
The two real fears that are reasonable are concerns over losing your job, and the other is around privacy. Big brother is watching you. For example; my iPad knows I’m going to go to the airport now. I don’t mind that so much, but in certain circumstance you might wonder who else knows where I’m going and should they know. Certainly I don’t want potential burglars knowing. AI is very good at keeping tracks on people, who they are and where they’re going. That’s a very reasonable fear and one, as a society, we need to keep an eye on.
Finally, just how big do you expect AI as an industry will get?
We’ve done some research on this. Our forecasts show that global spending on cognitive systems will exceed $30 billion by 2019 with a five-year compound growth rate of 55%.
At the moment, Europe is maybe a fifth of that and is growing slightly slower at the moment at least. North America is the hot bed in terms of companies investing in it and indeed building it accounting for around 80 per cent of the spending. Europe and the Middle East are second, but system spending in Japan and Asia Pacific will overtake the EMEA by the end of the forecast.
To read the article in its original form, as well as interviews with other leading industry analysts, professors and/or to receive advice on boosting your AI knowledge in business, please click here to receive/view our free reports.