PRW 2024: Responsible AI and Navigating Mindful Innovation
Artificial Intelligence (AI) is rapidly evolving and shaping the world as we know it. AI, a term first coined back in 1955, actualizes the ability for computers or other digital systems to perform tasks and analyze data by mimicking the natural intelligence and discernment of humans. Becoming increasingly adopted by publishers to automate tedious manual tasks, maintain quality, and even support research integrity during various stages of the manuscript lifecycle, AI has earned its place within our industry – but where exactly and at what capacity? With recent advancements specifically in the area of generative AI (GenAI), we now see a wave of enthusiasm resulting in some forming hasty conclusions that “AI is the answer to all our problems”. Furthermore, this eagerness to realize the known or perceived value of generative AI technologies broadly can mask the realities of the limitations of these emerging technologies. Without proper due diligence or equal consideration of other perhaps more viable solutions to our most important problems to solve, we run the risk of prioritizing the solution over problems. In honor of Peer Review Week 2024 – which is dedicated to the theme of Technology and Innovation – Aries is highlighting our approach to responsible innovation and how the product development journey can differ when AI is the driving force.
Within the traditional product development process, technology providers typically begin with a known problem or pain point that is raised through customer interactions, internal assessment, or changing industry needs. To better understand this problem, in-depth discovery and analysis is conducted through various research channels and a thorough consideration of the impacts and risks involved. Based on this comprehensive examination of the identified problem, teams begin to brainstorm and experiment with potential solutions before eventually reaching the best conclusion and shifting to the final stage of implementing a resolution. This linear process it iterative and rigorous, often with multiple cycles to ensure all channels are explored before moving on to the next stage. Starting the product lifecycle with the problem at the forefront follows the “problem-solution fit” (PSF) philosophy for a strong foundation to validate whether proposed solutions address the identified problems. Many solutions never move past that stage because it is only through this iterative experimentation that we validate our assumptions and ideas. Failure in this stage is as likely as success and equally as valuable in the product development lifecycle. The below process is designed to identify which ideas deliver on, exceed, or fall below expectations so that we can focus/move on to new, better-informed ideas towards the intended value.
In contrast, this journey can look a bit different when the solution is already prescribed from the start – as is often the case with recent generative AI conversations – and used to try and validate known (or unknown!) problems, causing a shuffle in the stages of traditional product development. This approach is not necessarily wrong or incorrect – but it does require a new perspective and rigor around defining success as a team. Without which, product teams can unintentionally spend too much effort trying to find a home for a desired solution. To break it down, we have mapped out what the product development journey could look like with generative AI, such as Large Language Models (LLMs), as the presumed solution.
Taking what we currently know about LLMs – understanding their capabilities and limitations – what pain points can it possibly address? This involves reviewing problems that are labeled as a current high priority, problems that have been shelved or deprioritized, and problems that have already been solved to see if they can be further optimized. Once one or more problems have been identified that could match the prescribed LLM solution, the research and experimentation completed in the Discovery phase determines whether the LLM solution is truly the right path. In this example, this may include consulting technical vendors and cloud services about what offerings and support they can provide, speaking with customers and internal data scientists to gauge accuracy and trust in proposed use cases, and carefully considering the various impacts of this solution.
In response to emerging technologies and needs within the industry, Aries Systems has followed both product journeys to innovate responsibly. We recognize the higher level of care and consideration content under peer review represents. In doing so, we analyze potential opportunities to develop of integrate generative AI into Editorial Manager to ensure it’s the right solution for our customers’ most critical pain points. As a member of the wider Elsevier family, we are committed to our Responsible AI framework for mindful innovation to ensure we produce the right solutions. We have taken part in similar exercises with potential AI solutions, and are currently in the midst of some product discovery. Aries has recently integrated with some AI-based tools, including Paperpal Preflight and Clear Skies Papermill Alarm.
“Whether it is a problem looking for a solution, or a solution looking for a problem, Aries is open to exploring the possibilities that any emerging technology could offer,” said Kate Horgan, Director of Product Management. “Our hunger for smart and considerate innovation means taking a step back to ponder the best possible outcome. The ongoing debate of AI’s role in peer review is an exciting one, but just because we can doesn’t mean we always should, and we strive to achieve the most ethical, robust, effective, and responsible solution available. If our support of the researcher community has taught us anything, it’s that we should always be willing to change our mind as new data or insights are validated.”
The theme for the tenth annual Peer Review Week (PRW), held from September 23-27, was selected via a global survey opened to the publishing community earlier this year. Aries Systems is pleased to collaborate alongside more than 35 participating organizations on the Peer Review Week Steering Committee!
Follow Along & Connect
#PeerReviewWeek
X: @PeerRevWeek
LinkedIn: @Peer Review Week
Join the Committee: contactprweek@gmail.com