"Rebalance power from the software guilds" to technology users and domain experts

The Atomic Human

By Neil D. Lawrence

Public Affairs, 448 pages, $32.50

In 2018, Zillow launched Zillow Offers, aimed at leveraging artificial intelligence to buy and sell houses. The business shuttered three years later as its algorithms struggled with the complex interactions between pricing predictions and market dynamics. Putting even simple algorithms into practice can get complicated.

In "The Atomic Human," Neil Lawrence -- a professor of machine learning at the University of Cambridge who has worked on supply-chain optimization at Amazon -- sets out to demystify AI, exploring "whether there is an essence of the human that can't be replaced by the machine." His goal is to empower readers to make informed choices about how they'd like "technology to be steered."

Humans are defined in part by our experiences, vulnerabilities and limitations. Likening AI to artificial plants, Mr. Lawrence notes that they may "pass for real plants" but lack the "scent of the flower, the flavor of the fruit, the ability to grow and respond" to their environment. AI, he reminds us, is made up of "computers and statistics," combined with ungodly amounts of data -- often about us. The resulting tools are transformational, but also concerning in fundamental ways.

Thanks to AI, we have already seen a "large-scale automation of decision-making," which has "sliced through our modern social fabric." Consider the role that social-media algorithms play in shaping our information environment. Matters are complicated by companies that may not always have our best interests at heart. AI is "typically not deployed to empower the individual, but to empower corporations," Mr. Lawrence writes.

Faced with "vast social power," companies have often "failed to self-regulate." Mr. Lawrence explores potential improvements. For instance, to combat surveillance capitalism, he suggests that "data trusts" might help "return the power" to the people who are creating the data -- essentially everyone.

Additionally, making it easier for "nurses, doctors, patients, lawyers, accountants" to develop their own "explainable, maintainable systems" might further "rebalance power from the software guilds" to technology users and domain experts.

There is no stopping the widespread adoption of AI. But there are opportunities to shape its societal impact. By understanding both AI and ourselves, we can engage more thoughtfully with emerging technologies and ensure that we only "serve the machine because it helps us," while remaining aware of how it might not.

---

Mr. Luca is a professor of business administration and the director of the Technology and Society Initiative at Johns Hopkins University Carey Business School." [1]

1. REVIEW --- Books -- Shortcuts: Technology: Keep the Power With The People. Luca, Michael.  Wall Street Journal, Eastern edition; New York, N.Y.. 07 Sep 2024: C.9.