‘Applied AI Letters’ – A New Journal With Emphasis on Real-World Validation

An interview with the Editor-In-Chief of Applied AI Letters, Dr Edward Pyzer-Knapp from IBM Research

Thank you for taking the time for this interview, Dr Pyzer-Knapp. There are many journals around, also in the AI field – so what are your aspirations for Applied AI Letters, and how does it discriminate itself from other publication outlets?

Edward Pyzer-Knapp (EPK): As an Industrial Researcher I found there was a gap. There’s lots of places where you can publish really phenomenal breakthroughs and methodologies, but there isn’t really a home for the applications of AI into new and emerging technologies, and I wanted to correct that. I wanted to provide a place where we can meet these new and rapidly evolving technologies but also these heavily demanding use cases that exist in the real world. I think that the rapid communication format of Applied AI Letters allows us to do that on a time frame that is meaningful to industrialists and useful to academics. We’re hoping to get very fast turnarounds in review time and we’re also launching a pre-print server which will allow you to get visibility at the point of submission. These are things that I don’t think really exist in one place, and the visibility and impactful nature of Applied AI Letters will, I believe, provide something new and valuable for the community.

Your background is in Chemistry (where you graduated with a PhD from Cambridge University a few years ago), and where ways of publishing work have historically been quite different from the Computer Science field. In which way did this shape this new journal?

EPK: Yes, you are right that there can be differences between the publication cultures between chemistry and computer science. For example, chemistry publication is dominated by journal publications, whereas a significant amount of computer science research is communicated through conference proceedings. Having been exposed to both cultures, I hope to take the best ideas and strategies from each culture. For example, we are aiming to have ‘conference speed’ (i.e. fast) review cycles enabled by our short ‘Letters’ format, but also to encourage more discussion between reviewers and authors, which is something I have particularly appreciated when publishing in the chemical domain. To this end, we are building on Wiley’s enormous experience of publishing high quality scientific journals, with high-quality and rigorous reviews. Additionally, by breaking the deadline crush which is created by fixed submission periods for conferences, we hope to be able to improve the reviewing experience for our community as well, through spreading the submissions throughout the year. Indeed, recent Turing Award winner, and pioneer of the machine learning community, Yoshua Bengio, has recently been promoting a similar idea – you can read more on his blog post ‘Time to rethink the publication process in machine learning‘. We believe that through the combination of publication cultures and experiences used to build Applied AI Letters, it will be able to perform a role similar to what Yoshua describes. 

What advice would you give to a researcher submitting their first paper to Applied AI Letters?

EPK: A paper for Applied AI Letters is short, it’s only going to be 9 pages long. You really need to focus on two things. The first thing is the value – why is what you’re doing important and impactful in the real world, and the second is the story – what can you do now because of your invention that we couldn’t do before. What avenues does it open? What impact does it gain? And if you combine those two, you will be well on the road to a very successful paper!

What impact will Applied AI Letters have on the real world?

EPK: I want Applied AI Letters to be a proving ground. A place where industry and academia can meet, can exchange ideas, can publish their really fabulous applied use cases. I would love to see one of our challenge articles really provoke a discussion and maybe spin out a whole new area of Applied AI or maybe I’d like to see one of our emerging technology articles spell out a new use case for a technology people hadn’t thought of before. And of course, of particular relevance for this audience, we certainly welcome submissions from the drug discovery, materials science and related areas which involve applications of AI – so please send us your submissions, which we are very happy to receive.

Thank you for the interview, and all the best for the future of your journal!

Further reading: Inaugural Editorial and home page of Applied AI Letters.


Leave a Reply

Your email address will not be published. Required fields are marked *