AI misdescription
October 24, 2018

AI misdescription and real world applications from Asia

By Paul Laughlin

In between blogging, from the two events I recommended, let’s return to the challenge of AI misdescription.

As I shared in my review of Tony Boobier’s new book, he calls out the problem of mis-labelled AI. That is analytics or even BI applications being called AI, mainly because it is fashionable.

Below, Tony returns as a guest blogger, to not only confirm that this is still the case, but also call out some genuine AI applications. As a globetrotter, who have previously reported for us from Vegas & Chile, Tony has had the opportunity to see first hand AI applications around the world.

Most of his examples below come from Asia, especially China. See which inspire ideas for you, as this blog keeps seeking to bring the real world implications of AI to life…

The misconception causing AI misdescription

The AI agenda is a global one, and misdescription remains widespread. What many businesses overseas describe as AI is in reality no more than the clever use of specific Apps – face recognition, voice recognition, predictive analytics and workflow cleverly bundled together.

Some even still confuse predictive analytics (such as that used in marketing) with AI, simply because of the ability of the technology to predict or ‘anticipate’ the future – at least in terms of the likely next purchasing action by a consumer. I wonder if ‘fortune tellers’ are an endangered species?

Progress on Autonomous Vehicle AI

But there are many who seem to be shaping up to meet the AI challenge. Last week the Chinese technology heavyweight Huawei announced a partnership with German car maker Audi which will increasingly focus on autonomous vehicles, in an attempt to apply industry insight to specific business issues.

Huawei forecast the global AI industry will reach a value of $380 Billion by 2025, and that 90% of the market will come from specific industry applications. Effective partnering is clearly high on the agenda.

Xu Wenwei, chief strategy marketing officer of Huawei says that they ‘will partner with universities, industry associations and others to nurture talents and offer capital support’. They aim to have one million developers and partners in their AI ecosystem within 3 years. Cynics might argue that it is the ‘quality’ rather than the ‘quantity’ that matters but that is a different story. Whilst innovation currently and (probably) unarguably rests in the Western world, few would dispute the ability of the East to replicate, and at scale.

Robots taking over Chinese customer touch-points

When recently staying at a major Asian hotel, I was greeted in the reception area by a ‘robotic‘ advisor which (who?) was able to explain the facilities of the venue. I was slightly underwhelmed, not least because of the use of local language without translation. But it’s part of a growing trend to use AI services robots in public and conference areas.

During China’s National Week, a team of robotic policemen were being used, with the ability to scan human faces, as part of an initiative to improve visitor safety. Standing slightly shorter than humans, they are equipped with cameras, loudspeakers, flashing lights and wheels that allow them to ‘walk about‘. It’s not really AI but it gives an indication of the direction of the technology.

AI misdescription

In fairness both the police and the hotel robots look more like R2D2 of Star Wars fame, rather than any attempt to be humanoid. Perhaps appearances do matter in our perception of what is an acceptable robot or AI assistant?

One major Chinese hotel chain, Kuntai, are already planning to use robots in the reception area where language and accounting skills are needed. By 2021 they plan to deploy AI in all their rooms,conference facilities and in corridors, to coincide with the 10th anniversary of their business. The key drivers of change ( as they describe them) are not only around operational efficiency, but also to attract the growing number of tech-savvy guests, especially from younger generations.

China’s push to beat the USA in rush to mainstream AI

Looking specifically at China, It’s all part of their ‘push’ to improve their capabilities in the area of AI, aiming to build a 1 Trillion yuan ( $147 Billion) artificial intelligence industry by 2030, which in real terms is just around the corner. Their target is to use AI as the major driver in industrial upgrading, especially in manufacturing, smarter cities, agriculture, defence and other areas. Consultants PwC forecast that the adoption of AI could boost China’s GDP by as much as 26 per cent in 2030. Anaud Rao, an AI consultant at PwC, suggests that initially the Us will have the fastest productivity gains because of its more developed infrastructure, but that China will start to pull ahead in 10 years.

Being practical, we should recognise that in a country as large as China, different regions inevitably move a different speeds. The development of AI is likely to happen first in the bigger cities. (To give a sense of scale, Beijing and it’s sister port city Tianjin have a combined GDP equivalent to that of Australia.

Zhang Yaqin, president of Baidu, one of the largest technology companies says that China is likely to outstrip the United States in the application of AI. Looking beneath the covers, there are at least two big reasons for this. The first is that of government support, not only in the form of funding but also in terms of vision – the National AI development plan published on the State Council’s website details a three-step strategy, which ensures a high degree of focus. The second reason is perhaps a little less obvious, and that is one of the flow of knowledge, in so far as it appears easier for Chinese experts to gain access to, and read, research papers written in english language, whereas its harder for Western experts to read about what’s going on in China.

How fast is too fast for consumer of AI-led CX?

I wonder if that hotel’s 3 year programme to introduce mainstream AI ( even in its broadest meaning) is realistic, even with industry-specific intentions ? Perhaps. It’s about three years since I last visited China, during which time the country has increasingly and quickly moved towards a cash-free society through companies such as Ali-Pay and WeChat Wallet. Part of the speed of change is due to government aims to crack down on the Black Economy, and ensure correct payment of taxes. The other part is the high level of convenience it appears to bring to the consumer. It seems that if the wind is blowing in the right direction, a lot can happen in a short time.

So, as a result of eventual consumer demand, coupled with government support, might this stuff actually be with us before we know it? If the latter, then don’t we all need to start thinking harder about the implications? Especially fortune tellers.

What do you see, AI misdescription or rapid AI mainstream?

Thanks to Tony for that global view on the acceleration of AI & robotics into mainstream customer experience.

What do you see happening? Is most of your experience AI misdescription, with a lot of hype & not many applications at scale? Or, do you see things changing faster than expected?

Whatever your perspective, as a Data/Analytics/Insight Leader, I’m interested to hear your perspective.

As well as being struck by the emerging role for Product Managers in Data Science teams – the 2018 Data Leaders Summit taught me that Machine Learning has progressed fast. Data Science teams are becoming more mature & many have moved beyond pilots into live commercial propositions & an emerging production line.

So, reason for us all to call out AI misdescription & be ready for meeting the insight needs of widespread AI adoption (when it comes).