Always at your service.


Does anyone remember the NOKIA 3210? The first commercially successful mobile phone with T9. Thanks to the “Text on 9 Keys” software, users did not have to type each letter individually to write an SMS text. Instead, the software provided the right words – and that without being intelligent itself. T9 transformed the act of “writing short messages” into “texting”. Just the sound of this newly-coined word brings the invaluable benefit of word-finding automation to the point: simplicity and ease.

Surrounded by ‘4.0.’ elves.

Originally, bots were software applications which carried out automated tasks – called scripts – via the Internet. These scripts are both simple and structurally repetitive. Scripts are executed faster and with fewer errors than the same task completed by a human.

One of the best known applications for these bots is search engines. These use web crawlers to evaluate sites. Their automated scripts retrieve sites, upload the contents of web servers and analyse and evaluate them. Other bots can, for example, buy and sell stocks. By 2014, more than 75 percent of shares traded on US exchanges arose from automated trading system orders.

There are also bots that act together – called botnets. In a botnet, bots give each other commands, or get commands from only one source. But such botnets tend to have a bad reputation because they are primarily known as malware. There are spambots that can send up to 51 million e-mails per minute. As well as this, social media botnets cause unrest within the public domain. Bots that respond to the requests of people are called chatbots. And perhaps it is precisely these chatbots that will make the transition from scripted bots to bots with artificial intelligence. Then you could find yourself talking to machines without realising it.

Coffee from the machine.

Imagine not clicking through a menu to place an order, but doing so in a discussion with a chatbot. That has already been possible for a long time. My Starbucks® Barista lets you order your favourite coffee at a machine. To do this, you start a conversation with a chat messenger. The bot responds like a real service person, accepting the order and forwarding it to the desired shop. The advantages are obvious: the customer does not have to queue, the name is always properly written on the cup and the company can collect a lot of valuable information about each individual person. Your favourite roast in exchange for premium data mining – served at the perfect temperature.

Of course, bots can make more than just coffee. At KLM Royal Dutch Airlines, flight ticket booking can be supported by an algorithm; ShopBot browses on eBay on behalf of a user for rarities and bargains; the chatbot lawyer DoNotPay acts against traffic tickets for illegal parking; the bot from H&M, the world's largest clothing retailer, finds matching outfits based on a piece of clothing; Wordsmith transforms data into text; Digit helps save money; the bot Lisa seeks out new customers – and even cities need their own bots.

Even Mercedes-Benz has one, but so far not one that automatically answers emails for employees. And because the Facebook Messenger alone has more than 11,000 bots, it needs its own web directories, called botlists, to find a suitable bot. Chatbots may have their origins back in the 60s, but their heyday is the present. Facebook, Google, Microsoft, IBM – they and many other companies are all vying to be the universal platform for all types of bots. They are sure that bots will replace apps. But if the latest CES in Las Vegas has proven one thing, it is that bots controlled by Amazon's Alexa are the next big thing.

Service needs to be learned.

Some of the fields of application need more than just going through a pre-structured process. Bots have to learn more about the customer and their preferences to be able to always get them the best deal. And it is precisely at this point that control by means of artificial intelligence (AI) comes into play. These bots are able to interpret emotional and semantic contexts, and in this way they can communicate in an increasingly human way. A new form of customer loyalty is another reason why more and more companies rely on the development of bots. Unlike a call centre agent, the bot usually has direct access to the customer database.

Instead of having to identify a person repeatedly, the bot remembers the personal profile. In addition, it draws straight from the existing product range and can thus offer the best deals. The bot is also always friendly and accessible around the clock. However, if it is not able to satisfactorily settle a transaction, the bot can still pass the discussion on to a human. Maybe in the future you will be able to decide whether you prefer to speak with a bot or a human. The latter would then possibly be a paid premium service.

Coverage is strength.

In addition, the coverage of each messenger plays a major role. Facebook Messenger is currently used by one billion active users. They do not all need to download a new app to access a bot service. Instead, they simply ping each bot and use the required service.

The Japanese company Fukoku Mutual Life Insurance recently replaced 34 employees with AI bots. These scan medical reports, keyword them and collect data that is relevant for a payout. AI bots are apparently taking over jobs

Bots in every walk of life.

It took just 18 years to forget that T9 was once a pioneering innovation. A standard that no-one questions any more today. Presumably it will be similar with bots. If cash is abolished, they might become normal payment channels, just like credit cards. Bots can already help with the purchase of a new car, suggest insurance and arrange servicing appointments. Why shouldn’t they also help to book a car2go or help with route planning? One day, all communication could be as simple as texting.