­An Anonymous Evaluator: Applications in the Age of “Artificial Intelligence”

On Making Applications—the Administrative Gap

Let’s face it: making a (good) application for any kind of funding or for an artist residency is a lot of work and will most likely demand quite some time (that could have been spent otherwise). On the one hand, making up a new project within a specific framework can sometimes lead to new and exciting outcomes; on the other hand, it is a reality that competition is pretty tough, and chances to succeed might appear relatively low at first glance.

So no wonder many artists do have mixed feelings about application work and consider it a tedious duty that might even be frustrating at times. Some artists, though, are really good at it and seem to have found an effective way of dealing with the matter. How did they do it, and what might be the right balance between artistic and administrative work? How much time should actually be reserved to make an application work?

Having done quite a few applications myself in the last 15 years—obviously with mixed results due to the nature of the cause—I really feel you. The reality of being a professional artist truly does not correspond well with romantic ideas at times.

But isn’t there any shortcut?

A New Reality—Unexpected Encounters

It seems like many artists believe they have found such a shortcut these days.

As a former Saari Fellow, I was asked to participate in this year’s peer review—specifically as an expert in the field of visual arts. To me it was an honorable task that I was taking very seriously, dedicating a lot of time and thought.

My procedure was to always start with the work proposal and then move on to the portfolio and the CV. In retrospect, it took me quite some time to realize that a very high number of work plans were completely made up by AI or had been “powered” by AI. Finally having found an explanation why many texts appeared somewhat weird and why a lot of layouts were reminiscent of a business report, I was gradually developing a sense for it. While the layout and the same repeating structure of the proposals did not bother me, it was the content that became arbitrary, speculative, or even missing.

So how should I deal with that new reality? I was well aware that AI had found its way into people’s daily lives but didn’t expect it to have gained such a substantial impact on art applications yet. As a serious evaluator, I kept investigating every text, trying to figure out its content. But as you can imagine, there is a limit.

So finally reading texts that actually had a human author felt like a relief, and made these texts stand out positively. Of course, when dealing with AI texts, it made a big difference if “the machine” was responsible for the whole process—so also for “finding” the idea—or if it was only used as a language tool to shape the text.

After I had finished the evaluation process, the topic was still on my mind, so I decided to actually go through every work plan once more—this time evaluating only the usage of AI. In some cases I could not really tell, so I introduced the category “uncertain.” It felt good to finally have some numbers.

Statistics

The sample for this statistic is 220 applications, around one-third of all applications in the field of visual arts. Please note that in other categories things might be quite different.

It turned out that 54% of the applicants were obviously and excessively using AI for their work plan. I was uncertain about 28% of applications, and only 18% I was completely sure had been written fully by a real person.

I did not investigate if portfolios had been put together with the help of AI, as the aesthetics of the layout are not of importance—only the artistic quality is.

Guidelines

I expected there had been no guidelines about the use of AI, but to my surprise, I learned that shortly before the open call begun, the instructions of the online grant system had been updated, and the following could be read:

You may use AI to support your application preparation, but please remember that you are responsible for its content. While AI can help you express your ideas more clearly, aim to retain your own voice and personality.

In my opinion, this is a fair and realistic approach—unfortunately, this advice has been vastly ignored or overlooked by applicants, as the lack of personal involvement was evident in so many texts.

The Future of Applications

I believe the discussion might not be over yet, as large language models are able to produce very fast and generic applications that would not have been possible back in the days. Consequently, the number of applications might rise further and require significant resources on many levels. I hope, considering AI’s energy footprint, the question of “digital awareness” will become more relevant in the future.

These days work proposals made with large language models (LLM) are still relatively easy to detect, but as technology is progressing fast, I am curious about how or if “the machine” will be able to mimic artistic “thinking” in the future and how this development will affect the evaluation process.

Recommendations to Applicants

If you are applying for the Saari Residence, do not underestimate the importance of the work proposal. Reserve some time and consider writing your text all by yourself, as it definitely will stick out and show your commitment. This is a personal opinion (?) of an anonymous evaluator and, in no way, an official guideline. I wish you all the best with your application; you can do it!