ChatGPT-Interview: Facts vs. fiction about AI-generated content

As an IT company, ChatGPT and AI-generated content has obviously also had an impact on our projects, from interesting partner tools to self-developed apps. We asked one of our AI experts how they use ChatGPT to create better user experiences - and why the human touch is still crucial.

Content: 

1. Meet our AI expert
2. A ChatGPT solution for eCommerce & SEO 
3. The 5 basics for efficient AI-generated content
4. The human factor
5. Advantages of content-generating AI

Meet our AI expert

Question: Hello Felix, can you introduce yourself and tell us a little bit about your job at DIGITALL

Hey there, my name is Felix, and I am an e-commerce developer with a primary focus on web development, where I'm deeply involved in crafting exceptional online shopping experiences for our clients.

As an e-commerce developer, my role at DIGITALL revolves around creating and optimizing websites that not only showcase products and services in the best light but also ensure seamless user interactions. My team and I are also constantly exploring innovative ways to improve customer engagement and satisfaction through cutting-edge technologies.

A ChatGPT solution for eCommerce & SEO

Back to overview

You're currently working on a ChatGPT integration for an eCommerce solution that basically helps generate SEO-optimized product information. How does it work?

Within our project work for eCommerce customers, we wanted to provide SEO support based on ChatGPT that is easy to use. So, we've built a Chrome extension that seamlessly integrates with an existing eCommerce platform and allows users to leverage SEO optimization features right within the web interface. This saves time and effort.

The functions range from generating meta data like titles and descriptions as well as keywords and URLs. Additionally, all suggestions are SEO improved, so every text has a good probability to positively affect the search engine ranking.

The solution uses product information like name and category, target keywords and unique SEO requirements, so the results really fit the specific product range and brand.

Furthermore, the user can choose from a selection of results which adds a human touch, plus, the eCommerce, sales or marketing expert can further add their know-how by picking the best choice. The whole process is designed to be as intuitive and seamless as possible. It's a solution that can cut a lot of work from a very repetitive process without losing quality.

The 5 basics for efficient AI-generated content

Back to overview

What are the most important things you have to consider when working with and integrating content-generating AI?

There are five things that good AI-models for content generation need to provide, especially when companies want to use them widely: accuracy, relevance, scalability, consistency and flexibility.

The AI-generated output needs to accurately reflect the desired information request and fit the target audience, relevant keywords, and other requirements for that specific prompt.

As such, the content also needs to be relevant. For example, content for people who are looking for sports-wear, needs to include more information on the material quality under duress, breathable fabrics, etc.

Scalability is incredibly important since these AI-models work with a massive amount of data which needs to be processed in a short amount of time.

No one wants to wait for results, especially if they might have been faster writing the content on their own.

Especially for companies but in general, consistency is important if you want to use AI-generated content for different purposes but under one "umbrella", such as a brand or your own personal voice. The content needs to reflect a specific style and should not deviate too much. Imagine, if product descriptions for one shop had wildly different writing styles - it would alienate customers because it would make the brand's voice unclear.

And last but not least, a good AI-model is able to "learn" and adapt given its information input. In our project example, it would stay flexible when it comes to changes in SEO algorithms, market dynamics, and customer behavior.

The human factor

Back to overview

The results of your integration are manually chosen by the user, which was a deliberate decision by you and your team. Why is it important to include human decision-making into the process and not simply automate all of it?

At the end of the day, the AI - despite its name - is not "intelligent" as we understand it. It does not have contextual understanding (yet) or can actually produce individual thoughts. For our solution, we deliberately designed the functionality in a way that a user can choose and edit the final result to make sure that it really does fit the brief/prompt.

In general, there are many different reasons why AI-generated content works best if it is managed and "guided" by humans.

A unique voice or identity is hard for an AI to develop, especially if it has to sound authentic.

Sure, the AI can summarize all the information in a very digestible text but we all know that it's not just the information that counts, it's the style, the humor, the personality that comes with it.

I'd say that ethics are a major key in developing and understanding AI. There have been so many examples, where AI could be manipulated into turning aggressive or insulting. It also can have potential biases simply based on its input and logical conclusions. A critical human hand can help get the bigger picture and take countermeasures.

I also think that there always should be room for the unexpected which is something that humans can deal with much better than AI. By allowing users to choose between options, for example, they can still shape the content on their own according to specific requirements, experiments, etc. It's just like with everyday work: if you only make decisions in your own bubble, you often lack impulses and inspiration from other viewpoints. It helps to collaborate.

The last reason why human collaboration with the AI is actually need has less to do with the quality of the work but rather with how we incorporate AI into our human processes and how it also has been discussed publicly.

AI is not here to replace people.

But of course, if we develop solutions that work in their own algorithmic decision silo, we absolutely perpetrate that myth. By involving users in the decision-making process, businesses can give ownership to the users and create engaging experiences. By being actively involved in the (final) results, the users are still the managers of their topics which motivates them to actively work on the iterative improvement of the AI system.

Advantages of content-generating AI

Back to overview

Now, plenty of people are currently talking about ChatGPT and other content-generating AI. In your opinion, what are the biggest advantages of this technology?

There's so many advantages and it really depends on how you use them. From the top of my head, you can be more efficient and save time, support creative processes, be more productive, check quality, scale and adapt, reduce costs and combine human with AI know-how for much better insights and decision-making.

For the everyday life, I think it's mainly about taking off tasks that are easy to do but incredibly time-consuming.

It's much easier to work on a summary of a meeting, if you already have a text to work with, even if it needs adjustments. The same goes for our eCommerce solution: writing these titles and descriptions for dozens, if not hundreds or even thousands of products, might be easy for each one of them but they add up and consume so much time, cognitive effort, and creativity.

With the AI-generated suggestions, it's much easier to focus on more context tasks and also keep the whole picture in mind which can easily get lost if you're always working on the small details. It's like your own sous-chef for your daily work: they make sure all the ingredients are measured and prepared, so you can create a wonderful dish.

Our AI- and data experts support you on your way to an efficient and smart data management. Whether it's business intelligence, artificial intelligence or data security - we transform your business.