Deus ex machina? A practical look at AI tools

It’s hard to avoid the term "arti­fi­cial intel­li­gence" these days: Almost every day, new AI-powered pro­grams and tools appear on the mar­ket, designed to sim­pli­fy writ­ing, research and cre­ativ­ity pro­cesses. It's a devel­op­ment that is caus­ing both excite­ment and con­cern — among research­ers as well as those who will be most affected by the AI tool revolu­tion: Cre­at­ives and desk workers.

Image 1: Gen­er­ated with "Gen­er­at­ive Fil­lung" in Pho­toshop with the prompt "AI Tools, Dall‑E, Mid­jour­ney, Stable Dif­fu­sion,, Designs AI" 

How­ever, machine learn­ing meth­ods have already been sup­port­ing us in our every­day lives for a num­ber of years. They work in the back­ground, for example, in search engines, in voice assist­ants and social media algorithms. Since the release of the AI-powered chat­bot Chat­G­PT in Novem­ber 2022, the top­ic of AI tools has also been more present in pub­lic discourse.

With the "Deus Ex Machina?"-series, we want to put selec­ted AI sup­port tools to the test: What can they do, what algorithms drive their AI, and how intu­it­ive are they to use? What are the rhet­or­ic­al advant­ages of these tools and what prac­tic­al assist­ance do they provide? We are focus­ing primar­ily on tools that aim to be a sup­port for every­day life in sci­ence and sci­ence com­mu­nic­a­tion, that are inter­est­ing for stu­dents, or that can (poten­tially) lessen the work­load in journ­al­ist­ic prac­tice — at least that is what the man­u­fac­tur­ers claim. But is this really the case? That is what we want to find out.

You can find our tool tests here:

Image 2: Gen­er­ated with Dall‑E with the prompt "AI Tools, Dall‑E, Mid­jour­ney, Stable Dif­fu­sion,, Designs AI"