Should academics try to engage the public?

Science Hell by Tom Gauld

This may seem like an odd title given the fact that I write a blog and I also have an active social media presence. But the question of public engagement is one that does come up often in academic circles, where we are increasingly encouraged to generate impact and communicate our research to the wider community. But while many see benefits, there is still some reluctance to doing any sort of communication outside of the recognised channels of academic publishing in the shape of books and journal articles. I remember having a conversation with a colleague who admitted that they only wrote for about three or four people, as nobody else would understand or be interested in the subject. Another colleague laughed at the idea of open access and allowing publications to be read by more people as a futile exercise. She asked: “who would want to read what we write?”

This may sound elitist, but in some areas of study the intended audience is not always the wider public. In legal research, academics may write to gain impact by influencing a small number of peers or policy-makers. Or some people engage in important theoretical work that seemingly has little engagement potential, but that helps to support other work.

Nonetheless, there are plenty of us who find public engagement rewarding, and an important aspect of our research. As a person who loves to write, I find that blogging is a vital part of my academic life. I enjoy trying to convey some of what interests me as a researcher to a wider audience than that which reads a journal article. However, this attitude can often be met with suspicion from those who do not favour engagement: blogs are a distraction, lesser output, useless exercises, or as a colleague told me once, “blogging does nothing for your career”. Possibly true, but it is still a worthwhile pursuit for many.

Maintaining a blog nowadays has become difficult, with social media and one-off platforms such as Medium taking a large share of the attention. Audiences generally dwindle, and the number of blogs is therefore reduced. A viable option for engagement has become The Conversation, an independent news publisher founded by a number of UK universities which provides journalistic articles sourced from the academic and research community, edited by in-house staff and delivered direct to the public. The published articles are released under a Creative Commons licence, and can be re-published by any news source provided they attribute the author and use the attached metadata. I have started using this tool, and have now published six articles with various levels of success. These tend to be more journalistic than your average Technollama article, and they are often edited with suggestions to make the article more approachable for a wider audience. I have been satisfied by the experience, and some articles have been picked up by mainstream news organisations such as Newsweek, El Pais, Yahoo and Gizmodo.

My latest experience has a long history in the making. A couple of years ago I wrote an article for the blog entitled “Do androids dream of electric copyright? Ownership of Deep Dream images“. This was a rather successful post, so I decided to turn it into a presentation for Gikii, and then a more ambitious presentation at re:publica in Berlin (I do talk a lot with my hands). I then turned the blog post and presentation into a longer journal article which got published this year in Intellectual Property Quarterly under the title “Do Androids Dream of Electric Copyright? Comparative Analysis of Originality in Artificial Intelligence Generated Works“, and I have been very pleased with the paper’s response. This is a great example of the benefits of blogging, where you start with an idea, develop it over time, do the research, and produce a longer, high quality journal article.  Not a waste of time, as some would like to believe.

Not content with leaving it at the published article, I then sent it as an idea to The Conversation to go full circle. They picked it up and after a few suggestions we published a short version entitled “Should robot artists be given copyright protection?” I did not choose the title, but I really didn’t give it much of a thought as this is an aspect that I often leave to the experienced editors at The Conversation; one of the things about publishing a successful news article in this day and age is that the title has to have a click-bait element. Unfortunately in this case, the title does not reflect at all the content of the piece as I have never considered whether robots will be awarded copyright protection. The intent of the article was always to wonder how copyright law will react to the growing number of machine learning algorithms that are already generating artistic works. I go through the law and case law in the UK, the EU, the US and Australia, and recommend that the best approach is that taken by UK copyright law, namely to give protection to the person who made the arrangements necessary for the work to be created.

Unfortunately, some responses to the piece did not go beyond the click-bait title, and assumed that I was advocating for robot rights. A commenter in the original article wrote:

“of course they should. how else will you have job security, IP law guy?”

While the title also seriously annoyed the author of the technology blog FrogHeart, who reproduced parts of the article but seemed to concentrate on one of the examples I gave, namely that of The Next Rembrandt. I admit that The Next Rembrandt is a headline grabbing isolated example, but I go on to give many other examples of works that are being generated by machine learning.

A more thoughtful response to the piece was written by Timothy Geigner in Techdirt, but again this seems to concentrate on the click-bait title and not so much on the content. I have never proposed that copyright should be given to machines, but what happens to those works that have been generated with considerable input from a machine, as copyright law will have two options, the work will be in the public domain, or copyright will be awarded to the program creator in one way or another. Thankfully, Geigner is able to get past the title and see that this was my argument all along. However, it confuses the UK and the EU approach (they are not the same), and seems to imply that I favour rights for robots again.

This is the frustrating part of the public engagement for experts and academics. You are an expert in your field and make a long and considered argument based on the law and case law from several jurisdictions, but you run in danger of having to argue the finer points of law with a non-expert, based on a version of an argument that you never made in the first place.

So why try to engage in the first place? Why not just let the journal article speak for itself?

Despite the relatively negative experience highlighted above, I still think that publishing the article was worth it. It got published in various outlets, which gave me more coverage than the few hundreds who subscribe to academic journals on average, or the almost 200 people who have downloaded the original from SSRN. At the time of writing, the article in The Conversation has been read by almost 8,000 people, and has been tweeted 148 times, and has received 274 Facebook shares and 267 LinkedIn shares. As an academic, these are figures that would be very difficult to achieve for a journal article.

But is it worth it if you are not able to convey the complexity of the subject? I still think so, but I have to admit that my faith has been shaken.