With potential applications that include robotics, self-driving vehicles, natural language process, virtual agents and machine learning, Artificial Intelligence could radically change many industries. But at the present time, AI isn’t all that widely used across many industries.
Not all types of AI are at the same stage of development and some will need to be adapted more than others to suit the needs of particular industries. Some early adopters are already getting started; implementing AI into their operations and even creating their own solutions and technologies.
Companies in the US and China are the ones most actively using AI right now. Commercial use of AI is still fairly novel in both markets but it’s almost impossible to believe there won’t be a huge uptake of this technology soon.
Some brave and innovative businesses have already started to experiment with the applications of AI in their operations. These early adopters are blazing a trail that the rest of us will probably soon follow. So what are their experiences and what are they using the tech for?
China is facing a shortage of doctors, with up to 650 milion people having no access to a GP and long delays trying to get a brief appointment. It’s a particular problem in rural areas where it isn’t easy to get trained doctors to work.
Doctors are also at high risk of being attacked by angry patients, making it a less appealing field to work in. AI could help bridge the gap between patient needs and doctors’ availability. But regulation restricts tech solutions to providing general advice on health rather than a full AI consultation – something that there’s arguably a need for in China’s poorly-served health market.
Many health-related apps targeting patients have emerged in China to tackle the issue of releasing pressure on the healthcare system. Search giant Baidu has launched its own app. Ping An’s Good Doctor app had close to 20 million regular users at the end of 2016. But it’s unclear how profitable these are.
Tech start-up Koboro offers a health app called Daxia Health, which can provide medical advice and offer health management plans. The founder claims a high profit rate of between 30 and 40%, but other apps have not achieved this. Baidu is considering a withdrawal from the healthcare sector, partly because it’s been entangled in a medical mis-selling scandal.
Other AI tools offer support and advice to physicians rather than to patients. Huiyihuiying uses AI to interpret CAT scans. Some Chinese doctors are also using IBM’s Watson medical AI technology to guide the treatment they are advising for patients. This AI tool analyses data to advise practitioners on the best care plan for cancer patients – but at $1000 per consultation, it’s hugely costly.
There may be a huge number of apps coming out of China’s medicare market but their profitability is not assured and it’s unclear how the authorities will continue to support and regulate this market in the future.
It seems that the recent flurry of investment is now levelling off and some start-ups have inevitably burned through their initial investment capital. Right now it’s a highly competitive market where it’s tough to find a profit. It now remains to be seen which apps emerge victorious from the competition and whether regulations are relaxed to allow AI a bigger role in diagnostics.
One of China’s biggest insurers, Ping An Insurance, is both using AI for its own operations and also innovating in technology for wider use across the industry. The AI tools focus on reducing costs and improving efficiencies, as well as creating a better customer experience.
One product available for the car insurance industry is designed to read photos of the scene of the crash and recognise how much damage has been caused. It can then calculate the payout in a very short amount of time.
Ping An is using AI for biometric identification of customers using their faces or voices. The company says this practice can reduce disputes with clients and reduce the time it takes to resolve claims. It’s also employing AI to fulfil most of its customer service interactions – in Japan, the use of similar technologies have led to layoffs in the insurance industry.
So far, the results seem positive for Ping An: the company enjoys a high net promoter score and improved quality control.
A 2017 report by Accenture spells out how AI is likely to revolutionise the insurance industry in the US, probably within the next 3 years. Companies working in this sector are under pressure to respond quickly to the new technology, or face being left behind their competitors.
Not every organisation dares to be an early adopter; for many AI is not yet even on the radar. Amazon’s one of the early adopters now making serious use of the technology in its warehouse operations.
At the company’s Manchester warehouse, AI robots pick inventory items for order fulfilment and pass them to human workers. This significantly reduces the amount of mileage the warehouse team need to cover in an average working day. Amazon’s order fulfilment is now operating much more like an assembly line rather than a customer meandering around a supermarket. Similar technology is used by Amazon’s Chinese counterpart Tmall.
The advent of AI in the logistics industry raises obvious concerns for jobs, but Amazon actually claims that the technology will increase not threaten jobs.
Online supermarket retailer Ocado makes a similar claim – that AI will support rather than replace human workers and is already helping chase margins in the competitive world of groceries. It seems the changes to the employment landscape wrought by AI may be more complex than previously envisaged. AI may not directly replace warehouse workers, but it will almost definitely change the nature of their roles.
Getting started with AI
Most organisations presently using AI have brought in outside help to both create and implement the AI they need. There’s still a shortage of skills and experience in this field and even major tech companies don’t have the skills in-house just yet to implement AI.
Few organisations are developing their own AI technologies, simply because it’s so expensive and few have enough scale to do so.
Right now, those already experimenting with AI tend to be the ones that are already highly invested in digital technologies. Not only do these tech-savvy companies have the mindset, but they’ve also gone through a digital transformation process that presumably leaves them well-structured for new technology adoption.
Early adopters face both advantages and disadvantages of being the first off the block with this new technology. They’re likely to face higher costs, risks and development time to create and implement the very new technologies themselves, but they’re at an advantage over later competitors if they manage to implement it successfully.