What are AI technologies?

AI technologies are programs, applications or apps that use artificial intelligence (AI) methods to offer their users added value. ChatGPT is currently one of the best-known AI tools. It should be noted that ChatGPT, like many modern AI tools, uses Generative AI technology. Thanks to the use of generative AI, modern AI technologies are able to produce new content that did not previously exist in this form. AI technologies support their users in a completely new way and are capable of scaling more and more human processes. ChatGPT is very generic and not very focused on individual use cases. However, there are many other generative AI technologies that focus on very specific applications such as podcast creation, meeting notes processing, research, image creation and many more. An initial overview of the most important AI technologies for relevant applications can be found in the Generative AI technologies overview.

What does AI technology mix mean?

If you regularly use AI technologies, you will quickly realize that there is not one AI technology for all applications. Instead, there are a large number of specific AI technologies for individual areas of application. The AI technology mix is no longer about the use of individual AI technologies, but about the ability to skillfully combine different AI technologies.

In her webinar on AI tools in marketing and customer service, AI expert Prof. Dr. Claudia Bünte combines Perplexity (for research), ChatGPT (for text and product development) and Leonardo AI (for image generation). Perplexity can search Internet resources and provides specific search results for academic and mathematical topics. The sources can be checked directly. ChatGPT for product ideas, because the AI tool can be used in specific roles, such as product developer or graphic designer, to deliver creative and specific results. Temperatures (measure for creativity) can be set. Leonardo AI is a free tool for image generation and offers daily tokens for use. It can create images in different styles and sizes.

How do I find the right AI technologies?

The number of AI technologies available today is already infinite and it is to be expected that more and more new AI technologies and start-ups will come onto the market in the future. This makes it all the more difficult to find the right AI technologies for yourself. The following tips can help:

1. use rankings and comparison lists from sources you know

AI technology overviews or rankings are a good start to finding the right AI technology. However, the number of rankings and overview lists also continues to rise daily. I therefore recommend using an AI overview where we already know the author or where we know that the source matches our values and requirements.

2. uses trial licenses

Most AI technologies offer trial subscriptions that are either completely free or can be canceled after 30 days. Use these types of subscriptions to avoid tying yourself to too many new tools from the outset. Even if they are often only small amounts, they quickly add up.

3. define your own processes and goals

Advertising and the internet often tempt people to use new AI technologies and take out expensive subscriptions. However, it is crucial that we first clearly define our expectations of AI technology in order to derive specific requirements. The search for suitable AI technology should only begin once these requirements have been precisely defined.

Unless, of course, the aim is to identify AI technologies that not only increase the efficiency of current work processes, but also take on tasks that were previously not part of your activities in this form.

How do I achieve responsible use of AI technologies?

Mindfulness and foresight are increasingly important when implementing artificial intelligence. It is essential that new AI technologies are used responsibly. Not only do data protection and transparency need to be taken into account here, but the skillful integration of AI technology and human action is also an essential challenge for anyone who wants to use AI technologies wisely and responsibly. More in-depth tips and thoughts on the mindful and responsible use of AI technologies can be found in my articles: Using AI technologies mindfully and responsibly and 5 tips for using AI technologies mindfully and responsibly.

From the field: AI technologies and humans for podcast creation, editing and publishing

In the following, I will show you how I use AI technologies today to optimize the production and marketing of my podcast. These technological solutions allow me to achieve a higher quality than was the case four years ago, when my podcast was launched. I can also increase my efficiency and even publish the podcast in various media formats.

1. the podcast recording

The process of publishing a podcast begins with planning the content and selecting the interview partners. Theoretically, I could create the conversation guide with ChatGPT and use AI technologies such as Perplexity.ai to research suitable conversation partners. But in the spirit of responsible integration of humans and AI, I take on these tasks personally. I decide who I invite to an interview, arrange appointments and conduct the interviews without prepared notes, but completely spontaneously.

2. the podcast edit

When editing the podcast, I use the AI technology of my publishing tool “Buzzsprout” for the first time, which I have been using since my first podcast recording four years ago. Although many tools were originally developed without AI technology, they now offer functions based on artificial intelligence. These are often associated with higher costs, so the decision to use them should be carefully considered. In my case, I use the AI functions of Buzzsprout directly instead of introducing a separate technology.

3. the podcast transcript

Buzzsprout does not currently offer a solution for transcribing or creating additional texts, so I decided to use “Descript”. This technology first creates a complete transcript of my episode, identifies different speakers and inserts time stamps. The AI-generated script can then simply be copied and pasted into my publication tool.

4. the podcast title

Descript not only transcribes, but also enables the creation of texts and media. I use it to generate the title of the podcast. I configured the prompt for the Descript AI assistant so that it always contains the term “AI” and the name of the interviewee.

5. the right podcast image

Even though Descript offers a wide range of options for image composition, I decided against using it. The original image of my podcast was created by a graphic designer friend, and I don’t want to replace it with an AI-generated image. The responsible use of AI technologies also includes the decision to leave certain tasks to humans. Only minor additions, such as adding a new partner logo, were made afterwards using AI technology, for which I used additional functions from Canva. Canva is another example of an application that was successful for a long time without AI and now offers AI technologies in addition to the classic functions.

6. the blog to the podcast

The majority of my podcast episodes are available exclusively as podcasts and can be accessed via my general podcast landing page. However, since the implementation of Descript’s AI technology, I have increasingly been publishing blog posts to accompany each episode. Descript makes it much easier for me to create summaries, show notes or other relevant text elements for my podcast. For example, I created the articles Innovations in Generative AI: A conversation with Marcel Rassiger and AI technologies in studies: A conversation with Julius Oberli with the support of Descript’s AI technologies.

Descript currently still has the weakness that it does not always write all texts in German. In the event that a text is generated in English, I use ChatGPT for translation.

7. the social media contribution to the podcast

I regularly publish posts on social media to ensure the timely release of my new podcast episodes. In the process, I realized that the existing AI technologies do not fully support me, but they do provide valuable suggestions. I usually use Descript to generate five suggestions for a post on LinkedIn. From these suggestions, I then develop a personalized post that carries my individual style, the “Sophie touch”. I also make sure to highlight relevant people who could potentially be interested in the respective episode or have a connection to my interview guest. The combination of my own human creativity and AI technology is of particular importance to me in this process.

Conclusion: AI technology mix – using humans and AI responsibly

This article and my example of podcast creation, including the creation of other media content and content elements, underline the need not to rely exclusively on a single AI technology. Instead, it is necessary to combine different technologies as required. It is generally advisable to use a smaller number of AI technologies instead of taking out new subscriptions every day. Nevertheless, a well-considered mix of AI technologies is becoming increasingly important for sustainable success with artificial intelligence. At the same time, it is essential to integrate people into this technology mix. Certain tasks and processes should continue to be carried out without the use of AI in the interests of careful and responsible use of AI technologies.

What happens next?

If you liked the topic of AI technology mix and would like to learn more about it, I recommend the AI Masterclass that I am offering together with Prof. Dr. Claudia Bünte in Berlin and Zurich this fall.

Or are you more interested in the responsible and mindful use of AI technologies? Then take a look at my offer on AI with foresight.

If you are still undecided, send Prof. Dr. Claudia Bünte a nice e-mail and ask for the video recording.

*I used SwissGPT’s AI technology to optimize the language of this article.