Advanced Computing in the Age of AI | Friday, March 29, 2024

Google-Backed Study Seeks Middle Ground in AI Debate 

The debate over the benefits and pitfalls of AI is evolving as alarmists and cheerleaders are joined by realists operating on the assumption that machine learning is here stay, workers must be retrained along with other technology investments while access to open data is broadened.

In a report on the economic impact of machine learning over the next decade, The Economist Intelligence Unit seeks a middle ground between AI pessimists and optimist. (Google, an optimist, sponsored the study.) The researchers conclude that AI fears may be overblown while optimists have yet to make a convincing case.

For now, the study released Monday (Feb. 5) concludes that machine learning has made inroads in segments such as energy, health care, manufacturing and transportation. However, those benefits so far represent "incremental improvements" in safety and efficiency rather than technological breakthroughs.

Machine learning has advanced to the point where policymakers "must focus on investing in skills and training, keep data safe, and [invest] in R&D and technology," the report concludes. It also found a communications gap between the developers of machine learning and potential users, including the need to manage expectations about what the technology can and cannot do along with the need to acknowledge risks.

In addition to worker retraining to meet growing demand for "soft skills," policy makers must also address concerns about the privacy and security of personal data, the report stressed.

Beyond that, the Google-backed research seeks to move away from what the authors call the "hyperbole" on both sides of the AI debate to "identify the middle ground by developing quantitative and qualitative scenarios on the impact of machine learning for a select number of countries and industries."

The report places particular emphasis on the importance of "upskilling" workers while investing in machine learning technologies. In a best-case scenario, U.S. investment boosts the average skill level of the American workforce, "thereby ensuring that the complementary effect is stronger than the substitution effect, and that the dominant impact of machine learning technology is to enhance, rather than replace, human labor," the study notes.

Workforce issues aside, critics warn that AI and machine learning only add to the evolving threats associated intelligence networks. Among them is respected security expert Bruce Schneier, who warned during a December AI conference that machine learning and other technologies can be used both to protect and attack networks.

A related issue is what effect AI will have on data privacy. Again, the machine learning report advocates greater transparency about how algorithms are being used, arguing that secrecy would ultimately result in stricter controls that would rein in the technology and dampen AI economic impact.

"Our objective with this report [is] charting a path between the techno-utopians who believe these technologies will solve all the world’s problems and the pessimists who warn that they are dooming us to a jobless, dystopian future," said report editor Chris Clague.

 

 

 

About the author: George Leopold

George Leopold has written about science and technology for more than 30 years, focusing on electronics and aerospace technology. He previously served as executive editor of Electronic Engineering Times. Leopold is the author of "Calculated Risk: The Supersonic Life and Times of Gus Grissom" (Purdue University Press, 2016).

EnterpriseAI