David Piorkowski

David Piorkowski, Michael Hind, John Richards · ibm-research

Although AI systems are increasingly being leveraged to provide value to organizations, individuals, and society, significant attendant risks have been identified and have manifested. These risks have led to proposed regulations, litigation, and general societal concerns. As with any promising technology, organizations want to benefit from the positive capabilities of AI technology while reducing the risks. The best way to reduce risks is to implement comprehensive AI lifecycle governance where policies and procedures are described and enforced during the design, development, deployment, and monitoring of an AI system. Although support for comprehensive governance is beginning to emerge, organizations often need to identify the risks of deploying an already-built model without knowledge of how it was constructed or access to its original developers. Such an assessment will quantitatively assess the risks of an existing model in a manner analogous to how a home inspector might assess the risks of an already-built home or a physician might assess overall patient health based on a battery of tests. Several AI risks can be quantified using metrics from the technical community. However, there are numerous issues in deciding how these metrics can be leveraged to create a quantitative AI risk assessment. This paper explores these issues, focusing on the opportunities, challenges, and potential impacts of such an approach, and discussing how it might influence AI regulations.

Anna Sokol, Elizabeth Daly, Michael Hind et al.

Large language models (LLMs) are powerful tools capable of handling diverse tasks. Comparing and selecting appropriate LLMs for specific tasks requires systematic evaluation methods, as models exhibit varying capabilities across different domains. However, finding suitable benchmarks is difficult given the many available options. This complexity not only increases the risk of benchmark misuse and misinterpretation but also demands substantial effort from LLM users, seeking the most suitable benchmarks for their specific needs. To address these issues, we introduce \texttt{BenchmarkCards}, an intuitive and validated documentation framework that standardizes critical benchmark attributes such as objectives, methodologies, data sources, and limitations. Through user studies involving benchmark creators and users, we show that \texttt{BenchmarkCards} can simplify benchmark selection and enhance transparency, facilitating informed decision-making in evaluating LLMs. Data & Code: https://github.com/SokolAnn/BenchmarkCards

Erik Miehling, Manish Nagireddy, Prasanna Sattigeri et al.

Modern language models, while sophisticated, exhibit some inherent shortcomings, particularly in conversational settings. We claim that many of the observed shortcomings can be attributed to violation of one or more conversational principles. By drawing upon extensive research from both the social science and AI communities, we propose a set of maxims -- quantity, quality, relevance, manner, benevolence, and transparency -- for describing effective human-AI conversation. We first justify the applicability of the first four maxims (from Grice) in the context of human-AI interactions. We then argue that two new maxims, benevolence (concerning the generation of, and engagement with, harmful content) and transparency (concerning recognition of one's knowledge boundaries, operational constraints, and intents), are necessary for addressing behavior unique to modern human-AI interactions. We evaluate the degree to which various language models are able to understand these maxims and find that models possess an internal prioritization of principles that can significantly impact their ability to interpret the maxims accurately.

Swapnaja Achintalwar, Adriana Alvarado Garcia, Ateret Anaby-Tavor et al. · ibm-research

Large language models (LLMs) are susceptible to a variety of risks, from non-faithful output to biased and toxic generations. Due to several limiting factors surrounding LLMs (training cost, API access, data availability, etc.), it may not always be feasible to impose direct safety constraints on a deployed model. Therefore, an efficient and reliable alternative is required. To this end, we present our ongoing efforts to create and deploy a library of detectors: compact and easy-to-build classification models that provide labels for various harms. In addition to the detectors themselves, we discuss a wide range of uses for these detector models - from acting as guardrails to enabling effective AI governance. We also deep dive into inherent challenges in their development and discuss future work aimed at making the detectors more reliable and broadening their scope.

David Piorkowski, John Richards, Michael Hind

In reaction to growing concerns about the potential harms of artificial intelligence (AI), societies have begun to demand more transparency about how AI models and systems are created and used. To address these concerns, several efforts have proposed documentation templates containing questions to be answered by model developers. These templates provide a useful starting point, but no single template can cover the needs of diverse documentation consumers. It is possible in principle, however, to create a repeatable methodology to generate truly useful documentation. Richards et al. [25] proposed such a methodology for identifying specific documentation needs and creating templates to address those needs. Although this is a promising proposal, it has not been evaluated. This paper presents the first evaluation of this user-centered methodology in practice, reporting on the experiences of a team in the domain of AI for healthcare that adopted it to increase transparency for several AI models. The methodology was found to be usable by developers not trained in user-centered techniques, guiding them to creating a documentation template that addressed the specific needs of their consumers while still being reusable across different models and use cases. Analysis of the benefits and costs of this methodology are reviewed and suggestions for further improvement in both the methodology and supporting tools are summarized.

Andreea Danielescu, David Piorkowski

Previous gesture elicitation studies have found that user proposals are influenced by legacy bias which may inhibit users from proposing gestures that are most appropriate for an interaction. Increasing production during elicitation studies has shown promise moving users beyond legacy gestures. However, variety decreases as more symbols are produced. While several studies have used increased production since its introduction, little research has focused on understanding the effect on the proposed gesture quality, on why variety decreases, and on whether increased production should be limited. In this paper, we present a gesture elicitation study aimed at understanding the impact of increased production. We show that users refine the most promising gestures and that how long it takes to find promising gestures varies by participant. We also show that gestural refinements provide insight into the gestural features that matter for users to assign semantic meaning and discuss implications for training gesture classifiers.

Soya Park, April Wang, Ban Kawas et al.

Data scientists face a steep learning curve in understanding a new domain for which they want to build machine learning (ML) models. While input from domain experts could offer valuable help, such input is often limited, expensive, and generally not in a form readily consumable by a model development pipeline. In this paper, we propose Ziva, a framework to guide domain experts in sharing essential domain knowledge to data scientists for building NLP models. With Ziva, experts are able to distill and share their domain knowledge using domain concept extractors and five types of label justification over a representative data sample. The design of Ziva is informed by preliminary interviews with data scientists, in order to understand current practices of domain knowledge acquisition process for ML development projects. To assess our design, we run a mix-method case-study to evaluate how Ziva can facilitate interaction of domain experts and data scientists. Our results highlight that (1) domain experts are able to use Ziva to provide rich domain knowledge, while maintaining low mental load and stress levels; and (2) data scientists find Ziva's output helpful for learning essential information about the domain, offering scalability of information, and lowering the burden on domain experts to share knowledge. We conclude this work by experimenting with building NLP models using the Ziva output by our case study.

David Piorkowski, Soya Park, April Yi Wang et al.

The development of AI applications is a multidisciplinary effort, involving multiple roles collaborating with the AI developers, an umbrella term we use to include data scientists and other AI-adjacent roles on the same team. During these collaborations, there is a knowledge mismatch between AI developers, who are skilled in data science, and external stakeholders who are typically not. This difference leads to communication gaps, and the onus falls on AI developers to explain data science concepts to their collaborators. In this paper, we report on a study including analyses of both interviews with AI developers and artifacts they produced for communication. Using the analytic lens of shared mental models, we report on the types of communication gaps that AI developers face, how AI developers communicate across disciplinary and organizational boundaries, and how they simultaneously manage issues regarding trust and expectations.

David Piorkowski, Daniel González, John Richards et al.

A vital component of trust and transparency in intelligent systems built on machine learning and artificial intelligence is the development of clear, understandable documentation. However, such systems are notorious for their complexity and opaqueness making quality documentation a non-trivial task. Furthermore, little is known about what makes such documentation "good." In this paper, we propose and evaluate a set of quality dimensions to identify in what ways this type of documentation falls short. Then, using those dimensions, we evaluate three different approaches for eliciting intelligent system documentation. We show how the dimensions identify shortcomings in such documentation and posit how such dimensions can be use to further enable users to provide documentation that is suitable to a given persona or use case.

John Richards, David Piorkowski, Michael Hind et al.

As AI models and services are used in a growing number of highstakes areas, a consensus is forming around the need for a clearer record of how these models and services are developed to increase trust. Several proposals for higher quality and more consistent AI documentation have emerged to address ethical and legal concerns and general social impacts of such systems. However, there is little published work on how to create this documentation. This is the first work to describe a methodology for creating the form of AI documentation we call FactSheets. We have used this methodology to create useful FactSheets for nearly two dozen models. This paper describes this methodology and shares the insights we have gathered. Within each step of the methodology, we describe the issues to consider and the questions to explore with the relevant people in an organization who will be creating and consuming the AI facts in a FactSheet. This methodology will accelerate the broader adoption of transparent AI documentation.

Michael Hind, Stephanie Houde, Jacquelyn Martino et al.

AI models and services are used in a growing number of highstakes areas, resulting in a need for increased transparency. Consistent with this, several proposals for higher quality and more consistent documentation of AI data, models, and systems have emerged. Little is known, however, about the needs of those who would produce or consume these new forms of documentation. Through semi-structured developer interviews, and two document creation exercises, we have assembled a clearer picture of these needs and the various challenges faced in creating accurate and useful AI documentation. Based on the observations from this work, supplemented by feedback received during multiple design explorations and stakeholder conversations, we make recommendations for easing the collection and flexible presentation of AI facts to promote transparency.

Matthew Arnold, Rachel K. E. Bellamy, Michael Hind et al.

Accuracy is an important concern for suppliers of artificial intelligence (AI) services, but considerations beyond accuracy, such as safety (which includes fairness and explainability), security, and provenance, are also critical elements to engender consumers' trust in a service. Many industries use transparent, standardized, but often not legally required documents called supplier's declarations of conformity (SDoCs) to describe the lineage of a product along with the safety and performance testing it has undergone. SDoCs may be considered multi-dimensional fact sheets that capture and quantify various aspects of the product and its development to make it worthy of consumers' trust. Inspired by this practice, we propose FactSheets to help increase trust in AI services. We envision such documents to contain purpose, performance, safety, security, and provenance information to be completed by AI service providers for examination by consumers. We suggest a comprehensive set of declaration items tailored to AI and provide examples for two fictitious AI services in the appendix of the paper.

Tommy Sandbank, Michal Shmueli-Scheuer, Jonathan Herzig et al.

Virtual agents are becoming a prominent channel of interaction in customer service. Not all customer interactions are smooth, however, and some can become almost comically bad. In such instances, a human agent might need to step in and salvage the conversation. Detecting bad conversations is important since disappointing customer service may threaten customer loyalty and impact revenue. In this paper, we outline an approach to detecting such egregious conversations, using behavioral cues from the user, patterns in agent responses, and user-agent interaction. Using logs of two commercial systems, we show that using these features improves the detection F1-score by around 20% over using textual features alone. In addition, we show that those features are common across two quite different domains and, arguably, universal.