Big Data and Artificial Intelligence have recently become important tools for managing the tsunami of scientific data generated across the disciplines. To ethically address the biggest problems in this space, attention must be paid to multiple computing, data, and communication approaches. Also important is attention to the grounding of all data and analysis into the domain of consideration. All analytical approaches, including artificial intelligence, (defined as intelligent machine approaches to data analysis), cannot be carried out without attention to the collection, cleaning, organization, access and sampling of data. This usually begs for statistical, computational, data management, and communication expertise to define and plan solution strategies and to present results. The domain expert must be securely in the loop [1]. Fail any piece of this, the AI algorithm can be rendered useless, with "garbage in and garbage out". This has ethical implications that must be considered because the results of data and AI analysis are frequently used to develop actions and policies that impact individuals, groups and society in general [2]. For example, the well-being of people may be impacted by biomedical decisions or treatments that are informed by these analyses [3,4,5,6]. In this workshop session, we will discuss the places in the ecology of AI and data analysis that need the attention of a team of experts with diverse perspectives, and how NBT [7] can address the ethical concerns that could arise if these teams do not function well.
When considering AI and data analysis, many focus on the analytical algorithms and procedures, and frequently forget about other important aspects such as the data and the communication of analysis results. We will consider and discuss these points in the conduct of data analysis that may lead to erroneous results, thereby threatening the safety of entities that are impacted by resulting decisions, actions, and policies. All forms of data analysis, including AI, require an array of activities, each calling for disciplinary expertise. No-boundary Thinking can help teams possessing diverse perspectives to work well together to perform and communicate robust and ethically safe data analysis.
Permission to make digital or hard copies of part or all of this work for personal or classroom use is granted without fee provided that copies are not made or distributed for profit or commercial advantage and that copies bear this notice and the full citation on the first page. Copyrights for third-party components of this work must be honored. For all other uses, contact the owner/author(s). ACM-BCB '22, August, 2022, Chicago, Illinois USA https://doi.org/10.1145/3535508.3545595 WOODSTOCK'18, June, 2018, El Paso, Texas USA F. Surname et al.
data. Data must be collected, cleaned, organized, archived, accessed, and sampled correctly in order for AI analyses to be reliable. We will discuss the places in the ecology of data and AI where attention and expertise are needed.
. Once an analysis model is decided and the analysis is complete, the work is not done. Analytics experts must communicate to domain experts how the data has been managed, the procedures that are used to analyze the data, and how the results might be interpreted. Ideally domain, data analytics, and communication experts work together from problem definition to analytical results using best practices of No-boundary collaboration to achieve this. We will discuss where and how important communication modalities are needed for the whole team to understand both process and results. It important, for example, that domain experts are able to assess if analytical results are consistent and explainable by the core scientific principles of the domain. Furthermore, all team members should be able to understand what has been done with the data, and how to interpret results with confidence.
Team building can be difficult when participants are from the same discipline or sub-discipline, but needs special attention when participants are from different disciplines [9]. Challenges arise when participants have different perspectives, use different vocabulary, and have different cultural views on what constitutes good problems and solutions. Essential to NBT teams is proper formulation of the problem to be solved; a basic tenet is that the NBT team must come together to synthesize diverse perspectives to decide the problem before solutions can be considered. Given that participants come with different views on problem formulation and solution, it is important to consider a robust process for team formation and maintenance. This takes extra effort and time, but scholars studying teams of experts with diverse training have found that they are better positioned to be successful in solving even deep and difficult problems [8] especially if they have learned to work well with each other.
In this workshop session, we identify the places in the data analysis and AI ecology where ethical concerns may arise. Then we will consider how NBT might be helpful. We will first briefly present some examples, and then we will engage with the workshop participants to consider and brainstorm other issues, and other approaches to address the ethics of data analysis and AI using NBT.