Validity Is What You Need

While AI agents have long been discussed and studied in computer science, today's Agentic AI systems are something new. We consider other definitions of Agentic AI and propose a new realist definition. Agentic AI is a software delivery mechanism, comparable to software as a service (SaaS), which puts an application to work autonomously in a complex enterprise setting. Recent advances in large language models (LLMs) as foundation models have driven excitement in Agentic AI. We note, however, that Agentic AI systems are primarily applications, not foundations, and so their success depends on validation by end users and principal stakeholders. The tools and techniques needed by the principal users to validate their applications are quite different from the tools and techniques used to evaluate foundation models. Ironically, with good validation measures in place, in many cases the foundation models can be replaced with much simpler, faster, and more interpretable models that handle core logic. When it comes to Agentic AI, validity is what you need. LLMs are one option that might achieve it.

Designing Fiduciary Artificial Intelligence

A fiduciary is a trusted agent that has the legal duty to act with loyalty and care towards a principal that employs them. When fiduciary organizations interact with users through a digital interface, or otherwise automate their operations with artificial intelligence, they will need to design these AI systems to be compliant with their duties. This article synthesizes recent work in computer science and law to develop a procedure for designing and auditing Fiduciary AI. The designer of a Fiduciary AI should understand the context of the system, identify its principals, and assess the best interests of those principals. Then the designer must be loyal with respect to those interests, and careful in an contextually appropriate way. We connect the steps in this procedure to dimensions of Trustworthy AI, such as privacy and alignment. Fiduciary AI is a promising means to address the incompleteness of data subject's consent when interacting with complex technical systems.

Racial categories in machine learning

Controversies around race and machine learning have sparked debate among computer scientists over how to design machine learning systems that guarantee fairness. These debates rarely engage with how racial identity is embedded in our social experience, making for sociological and psychological complexity. This complexity challenges the paradigm of considering fairness to be a formal property of supervised learning with respect to protected personal attributes. Racial identity is not simply a personal subjective quality. For people labeled "Black" it is an ascribed political category that has consequences for social differentiation embedded in systemic patterns of social inequality achieved through both social and spatial segregation. In the United States, racial classification can best be understood as a system of inherently unequal status categories that places whites as the most privileged category while signifying the Negro/black category as stigmatized. Social stigma is reinforced through the unequal distribution of societal rewards and goods along racial lines that is reinforced by state, corporate, and civic institutions and practices. This creates a dilemma for society and designers: be blind to racial group disparities and thereby reify racialized social inequality by no longer measuring systemic inequality, or be conscious of racial categories in a way that itself reifies race. We propose a third option. By preceding group fairness interventions with unsupervised learning to dynamically detect patterns of segregation, machine learning systems can mitigate the root cause of social disparities, social segregation and stratification, without further anchoring status categories of disadvantage.

Don't Fear the Reaper: Refuting Bostrom's Superintelligence Argument

In recent years prominent intellectuals have raised ethical concerns about the consequences of artificial intelligence. One concern is that an autonomous agent might modify itself to become "superintelligent" and, in supremely effective pursuit of poorly specified goals, destroy all of humanity. This paper considers and rejects the possibility of this outcome. We argue that this scenario depends on an agent's ability to rapidly improve its ability to predict its environment through self-modification. Using a Bayesian model of a reasoning agent, we show that there are important limitations to how an agent may improve its predictive ability through self-modification alone. We conclude that concern about this artificial intelligence outcome is misplaced and better directed at policy questions around data access and storage.