ARTIFICIAL INTELLIGENCE, LARGE LANGUAGE MODELS (LLMs), and other algorithms are increasingly taking over bureaucratic processes traditionally performed by humans, whether it’s deciding who is worthy of credit, a job, or admission to college, or compiling a year-end review or hospital admission notes.
But how do we know that these systems are working as intended? And who might they be unintentionally harming? Given the highly sophisticated and stochastic nature of these new technologies, we might throw up our hands at such questions. After all, not even the engineers who build these systems claim to understand them entirely or to know how to predict or control them. But given their ubiquity and the high stakes in many use cases, it is important that we find ways to answer questions about the unintended harms they may cause. In this article, we offer a set of tools for auditing and improving the safety of any algorithm or AI tool, regardless of whether those deploying it understand its inner workings.
Algorithmic auditing is based on a simple idea: Identify failure scenarios for people who might get hurt by an algorithmic system, and figure out how to monitor for them. This approach relies on knowing the complete use case: how the technology is being used, by and for whom, and for what purpose. In other words, each algorithm in each use case requires separate consideration of the ways it can be used for — or against — someone in that scenario.
This applies to LLMs as well, which require an application-specific approach to harm measurement and mitigation. LLMs are complex, but it’s not their technical complexity that makes auditing them a challenge; rather, it’s the myriad use cases to which they are applied. The way forward is to audit how they are applied, one use case at a time, starting with those in which the stakes are highest.
Esta historia es de la edición Summer 2024 de MIT Sloan Management Review.
Comience su prueba gratuita de Magzter GOLD de 7 días para acceder a miles de historias premium seleccionadas y a más de 9,000 revistas y periódicos.
Ya eres suscriptor ? Conectar
Esta historia es de la edición Summer 2024 de MIT Sloan Management Review.
Comience su prueba gratuita de Magzter GOLD de 7 días para acceder a miles de historias premium seleccionadas y a más de 9,000 revistas y periódicos.
Ya eres suscriptor? Conectar
Ask Sanyin: How Do You Build for an Unpredictable Future?
While the pandemic was a wild ride of uncertainty for me and many of my peers in leadership, it feels like we never regained our footing.
What You Still Can't Say at Work
Most people know what can’t be said in their organization. But leaders can apply these techniques to break through the unwritten rules that make people self-censor.
Make Character Count in Hiring and Promoting
Most managers focus on competencies when evaluating candidates but it’s character that will transform the DNA of the organization. Here’s how to assess it.
Why Influence Is a Two-Way Street
Managers achieve better outcomes when they prioritize collaborative decision-making over powers of persuasion.
Know Your Data to Harness Federated Machine Learning
A collaborative approach to training AI models can yield better results, but it requires finding partners with data that complements your own.
How Integrating DEI Into Strategy Lifts Performance
Incorporating diversity, equity, and inclusion practices into core business planning can provide a competitive edge.
The Myth of the Sustainable Consumer
Companies that understand the different kinds of consumers for sustainable products can market to them more effectively.
A Practical Guide to Gaining Value From LLMs
Getting a return from generative AI investments requires a systematic approach to analyzing appropriate use cases.
Improve Workflows by Managing Bottlenecks
Understand whether process or resource constraints are stalling work.
Craft Schedules That Work for Everyone
Business leaders can improve retention and business performance with schedules that make sense for workers’ lives.
