Bring Human Values to AI

Speed and efficiency used to be the priority. Now issues such as safety and privacy matter too.

by

and

Yves Lostanlen

by

and

Yves Lostanlen

From the Magazine (March–April 2024) · Long read

Giulio Bonasera

Summary.

When it launched GPT-4, in March 2023, OpenAI touted its superiority to its already impressive predecessor, saying the new version was better in terms of accuracy, reasoning ability, and test scores—all of which are AI-performance metrics that have been used for some time. However, most striking was OpenAI’s characterization of GPT-4 as “more aligned”—perhaps the first time that an AI product or service has been marketed in terms of its alignment with human values.

In this article a team of five experts offer a framework for thinking through the development challenges of creating AI-enabled products and services that are safe to use and robustly aligned with generally accepted and company-specific values. The challenges fall into five categories, corresponding to the key stages in a typical innovation process from design to development, deployment, and usage monitoring. For each set of challenges, the authors present an overview of the frameworks, practices, and tools that executives can leverage to face those challenges.

Idea in Brief

The Problem

Products and services increasingly leverage artificial intelligence to improve efficiency and performance, but the results can be unpredictable, intrusive, offensive, and even dangerous.

The Solution

Companies need to factor AI’s behavior and values into their innovation and development processes to ensure that they bring to market AI-enabled offerings that are safe to use and are aligned with generally accepted and company-specific values.

How to Proceed

This article identifies six key challenges that executives and entrepreneurs will face and describes how to meet them. Companies that move early to acquire the needed capabilities will find them an important source of competitive advantage.

When it launched GPT-4, in March 2023, OpenAI touted its superiority to its already impressive predecessor, saying the new version was better in terms of accuracy, reasoning ability, and test scores—all of which are AI-performance metrics that have been used for some time. However, most striking was OpenAI’s characterization of GPT-4 as “more aligned”—perhaps the first time that an AI product or service has been marketed in terms of its alignment with human values.

A version of this article appeared in the March–April 2024 issue of Harvard Business Review.

JA

Jacob Abernethy is an associate professor at the Georgia Institute of Technology, a cofounder of the water analytics company BlueConduit, and a research scientist at Google.
FC

François Candelon is a partner at private equity firm Seven2 and the former global director of the BCG Henderson Institute.
Theodoros Evgeniou is a professor at INSEAD and a cofounder of the trust and safety company Tremau.
AG

Abhishek Gupta is the director for responsible AI at Boston Consulting Group, a fellow at the BCG Henderson Institute, and the founder and principal researcher of the Montreal AI Ethics Institute.
YL

Yves Lostanlen is the founder of SenseiAdvisory, where he advises CEOs of Global 2000 companies, and has held executive roles in scale-ups including Element AI.

Bring Human Values to AI

Idea in Brief

The Problem

The Solution

How to Proceed

Partner Center

Explore HBR

HBR Store

About HBR

Manage My Account

Follow HBR