Artificial intelligence – an increasingly important part of websites and intranets

Recently, I published a post on LinkedIn External link. about how artificial intelligence will probably be more involved in the writing process and become a conscious or unconscious part of the texts we consume, such as social media posts or news articles. The plot of the post was that the whole post was written by just artificial intelligence which was given the simple instruction to "Write an article in English about how AI will write all texts in the future".

We've known for a long time that companies such as Tesla are using advanced artificial intelligence to make their cars self-driving, and back in 2017 Apple started including the Apple Neural Engine in iPhones and other devices, which puts artificial intelligence in your pocket wherever you go. It's the Apple Neural Engine that allows you to get great-looking photos, quickly find any photo where a particular person appears, authorise payments with your face via FaceID and of course, even talk to Siri.

We were also recently met with articles about how ChatGPT's  External link.rise caused "Code Red" at Google as services like ChatGPT could be a threat to the traditional search engine as it gives us an articulate answer to the visitor's query instead of a list of links like Google displays today.

Moreover, Microsoft has already integrated OpenAI (of which ChatGPT is a part) into its Bing search engine and its Edge browser. With AI integrated directly into the browser, you can, for example, quickly get a summary of the page or document you are currently reading, help with writing a post on LinkedIn and much more.

Proof that we are only in the early stages of AI support, however, and that we have yet to review and assess the results we get from AI, came earlier this week. Google recently announced its AI chatbot called Bard External link.. Unfortunately, the demonstration was less successful as Bard showed factual errors in its first demo.

However, for a long time I have been thinking about how artificial intelligence can help us in our efforts to create the best digital workplace or the best customer journey on the website.

In 2017, I conducted a proof of concept (POC) with a client where we had an AI trained on the website's content. The aim then was to have a smarter more user-friendly alternative to the website's search engine and the idea was a new kind of FAQ feature that had could interpret the visitor's questions and refer to the right answer. Unfortunately, this proof of concept was never launched as a finished feature, the main reason being precisely that the artificial intelligence was not yet good enough for us to be satisfied with the result, especially not in the English language.

Since 2017, a lot has happened in the AI world and many of the web solutions we deliver today in the form of intranets or websites often have one or more features using artificial intelligence. Among other things, we find recommendations based on visitor patterns, buying patterns and seasons, but also tools such as automatic translation into other languages or search engine enhancement.

Instead, the big question is what the next step is and how the customer experience (CX) and employee experience (EX) will change going forward based on the advance of artificial intelligence in multiple forms.

If services like ChatGPT and Bard become more common where the user experience is that you ask a question and get an instantly formulated and personalised answer, rather than a link to a page, then sooner or later the user will also expect the same interaction on the organisation's website and intranet.

For example, if an employee is wondering how vacation days are affected by VAB, today they have to look up the page about vacation days on the intranet and read up on how it might affect them. What an employee will expect is to be able to easily ask the intranet "I've been on leave for 10 days this year how will that affect my holiday days?". The answer you would expect is a personalised and articulate answer as if you were talking to someone from the payroll department. An AI who has learned the HR manual, who knows who is asking the question and who can also see and refer to data from the payroll system.

By allowing artificial intelligence to train on internal data such as visitor patterns, customer data and sales data, websites are guaranteed to offer a new kind of personalisation and experience where text, images and offers are personalised to you. The user experience will probably consist of an artificial intelligence knowing what you want, when you want it and how you want it without you having to ask the question. This is just based on past patterns from you and other customers.

Out of curiosity, I have already conducted a POC where OpenAI is integrated directly into Sitevision. The first version is mainly focused on helping the editor to write correct texts, but the integration could help the user or visitor in a variety of ways.

Consid External link. currently has an innovation project where AI with its potential is interesting and where internal and external data is used to optimize assortment and product offering in e-commerce. For example, it takes into account factors such as sales trends, stock availability, events, seasons and weather when suggesting purchases. This in turn optimises sales and reduces waste and environmental impact.

We are already seeing artificial intelligence that can generate different types of product images. You will soon be greeted by product images in e-commerce that are customised for you with colours you like and inspiration from product images of your previous purchases. An example of this already today is how the streaming service Netflix uses different cover images and previews for the same series and movies for different people, all based on your history in your profile.

Finally, I also asked my colleague Patrik Zetterström, one of Sweden's leading AI experts, how he thinks websites will be affected by artificial intelligence in the coming years.