Call centre nightmares averted
The collection of data from virtually any source is a common business practice today. If you ever had to call technical support, you are probably familiar with that short recording disclaimer notifying you that the call is monitored and can be used for quality purposes. Many of us have encountered cases when our issues weren’t resolved and we had to patiently wait for weeks until someone competent finally got to them. It isn’t hard to imagine a situation where you your call is transferred from one agent to the next, because you were initially transferred to the wrong department and they can’t help you. You are then forced to explain your situation over and over to the next agent in line, making you wish you had recorded a speech before you called them. At that point you are probably thinking that there must be a better way to solve this. Now there is – and it’s artificial intelligence.
AI- powered virtual assistants
Unresolved customer service or technical issues are frustrating not only for the customer, but also for the supplier. Unhappy customers take their business elsewhere and they tell everyone they know in the process. Bad customer service will soon be a thing of the past, largely due to machine intelligence. AI can help solve such issues to each side’s contentment by acting as a first line of support, collecting all the information about the issue at hand, analysing the request and transferring the customer to the most appropriate department or person, along with short notes explaining the issue.
There are other, more advanced techniques that could also be applied in the pursuit of quality customer service, which is becoming a prerogative for many businesses, small and large. These include an analysis of the tone of voice of the caller, helping determine their emotions at the time and offering exceptional remedies or perks when there is a perceived risk of losing them as a client.
AI, virtual assistants and ubiquitous computing
We live in a tech revolution, with connected devices at its forefront. It is not unusual to own and use four or more devices and to be bombarded with the same notifications on all of them. The concept addressing this problem is called ubiquitous computing. You can imagine it as an omnipresent assistant, which will send notifications only to the device you are currently using. It can read multiple sources of data such as emails or calendars, and automate a variety of tasks based on your habits. Perhaps the most useful thing it can do, by far, is order flowers on your behalf, on your partner’s birthday or your anniversary.
Warehousing and storage
Have you ever wondered how big warehouses, like Amazon’s, are able to ship so many parcels every day? Everything must be perfectly arranged and stored, right? Using AI and machine learning, Amazon’s systems are able to create a variety of locations containing, what may seem like, random items. In fact, these collections of items are customers’ future orders. Instead of sending the picker to 10 different locations spread all over the massive warehouses, the computer algorithm can calculate the shortest distance and the lowest amount of stops it would take to fulfil the order in the fastest time. Another example of such futuristic warehousing is online supermarket Ocado, which can deliver groceries across the UK on a daily basis, thanks to its fully robotised storage facilities.
E-commerce and physical stores
Ultrafast processors, coupled with deep learning algorithms have the useful ability to find patterns in vast datasets of unlabelled data, which would certainly escape most human data scientists. Just by observing the browsing patterns of in-store shoppers, or looking at their online equivalent: heatmaps for e-shops, machines are capable of preparing tailored offers containing suitable products, likely to be purchased.
AI on the stock market
New initial public offerings take place every second. Will a stock open high? Will its opening rate affect the stock prices of other companies inside or outside of that industry? Predicting stock market movements is the dream of every investor and trader, yet it is constantly getting more complex and more difficult even for the most promising of analysts. Once AI was introduced to this sector, it became feasible to evaluate the hypothetical interplay between a large number of related factors and act upon them in a timely manner, earning or saving investors sheer millions.
Hollywood and photography
AI and machine learning are gradually finding their way into numerous industries, including cinematography. Animation of 3D characters based on the movement of real actors, face swapping, object recognition and automatic scene alteration are just some of the new tricks special effects experts have in their repertoire now, all because of AI. Goodbye, photobombers (or ex-partners, for that matter) – soon we will be able to re-touch our pictures and remove any unwanted persons or objects at the click of a mouse or at the swipe of a finger.
Pitfalls of AI
As with every new technology, AI comes with certain flaws, which are yet to be worked out by developers and data scientists. Shortly after Facebook announced the introduction of AI in their Messenger app, for example, instead of learning to communicate with people, the algorithm behind it started communicating with other Messenger bots, thus creating their own language. The programme was then promptly shut down by Facebook. Although this isn’t necessarily a failure and the newly developed algorithm can be used in machine-to-machine communications, the AI in this case failed to fulfil its intended purpose – to communicate with people.
Copywriter: Ina Danova