AI researchers don’t think ethics are their problem

Hello and welcome to Protocol Enterprise! Today: why AI researchers continue to think ethical concerns are a concern for later, what’s stopping $52 billion in grants from reaching US chip-making companies, and the latest funding rounds in business technology.

not my job

Amazon, Google, Microsoft, Tesla and other AI leaders settled into one of the world’s most important computer vision conferences this June in New Orleans.

This year’s conference marked the first time CVPR organizers have “strongly encouraged” researchers to consider the negative societal impacts of their research.

  • Sure, computer vision holds promise for detecting disease faster and making autonomous vehicles safer, but it also enables discriminatory surveillance technology, controversial emotional AI, and deepfakes spreading misinformation.
  • Some researchers I spoke to at CVPR have pushed back. One of them told me that researchers “can’t tell you what the applications of their research are. It’s not their job.
  • Another who has helped advance artificial intelligence said, “It’s not the techniques that are bad; it’s how you use it.”

Researcher resistance hampered goals of CVPR organizers to move the computer vision community towards a more ethical approach to their work.

  • Although some researchers have acknowledged possible downsides to their work in papers presented to the CVPR — including research that could be used to improve deepfakes to spread misinformation — many have not.
  • The CVPR program chairs didn’t push too hard. “”The community as a whole isn’t quite there yet. If we make a very drastic change, reviewers won’t really know how to take that into account in the review process,” Stefan Roth, CVPR chair holder and professor of computer science at the German Technical University of Darmstadt.

But proponents of AI ethics balked.

  • “That’s exactly what pisses me off,” said Timnit Gebru, founder and executive director of the Distributed Artificial Intelligence Research Institute, and researcher with a Ph.D. in computer vision. “[Computer vision researchers] have convinced themselves that it is not their job.
  • Meanwhile, far away in Seoul, the main AI ethics conference – FAccT (Fairness, Accountability and Transparency) – was held the same week.
  • But hundreds of Asian computer vision researchers, many from South Korea, were thousands of miles away advancing their craft with relatively little concern for the societal implications of their work.
  • Alice Xiang, general co-chair of this year’s FAccT conference and head of Sony Group’s AI Ethics Office, told me: “We sometimes worry about whether practitioners who actually develop AI technologies might think this is just a conversation for ethicists.

Read my full dispatch from the CVPR conference and find out why research ethics are important for AI companies.

-Kate Kaye (E-mail | Twitter)

DATASTAX SPONSORED CONTENT

You are either in real time or out of time: Apps like this power our daily lives, and if they can’t access massive amounts of data stored in a database as well as “in motion” stream data instantly, you – and millions of customers – will not have this data in the – instant experiences.

Learn more about DataStax

Fools on the Hill

The $52 billion in proposed funding for U.S. chipmaking subsidies that in January appeared to be going through Congress is no longer a safe bet. It has been languishing in congressional reconciliation for weeks and has gone from a widely supported bill to partisan football.

The catch might be that, like it or not, chip subsidies have never been at the top of Congress’ list of priorities. His fate has been tied to USICA, a law that Senate Minority Leader Mitch McConnell swore will not pass if the “Democrats pursue a partisan reconciliation bill”. If some form of the bill doesn’t make it to the president’s desk before the August congressional recess, chip funding is virtually dead until the new Congress begins governing in 2023.

The chip giants who announced tens of billions of new or expanded chip manufacturing facilities in the United States are crazy. But the anger has yet to translate into a change of plans: Intel has delayed the ceremonial start of construction at its Ohio site, but the company has started building the new chip factory anyway. A TSMC board member suggested that construction — which is already underway at its Arizona site — could slow if lawmakers don’t pull together and pass the grant program. Samsung has also started construction of its Texas factory.

— Max A. Cherney (E-mail | Twitter)

Around the company

GlobalFoundries and STMicroelectronics announced plans to build a new chip factory in France as a joint venture. It is expected to open by 2026.

This major internet outage in Canada on Friday was caused by a maintenance update which broke enough routers inside Rogers to account for the disruption. Be careful there.

DATASTAX SPONSORED CONTENT

You are either in real time or out of time: Many of the challenges facing our world today are increasingly complex and critical, such as climate change, talent shortages and supply chain disruptions. Solving these problems requires quickly analyzing large data sets. Additionally, organizations need to use data to predict future problems and then determine the most effective solution.

Learn more about DataStax

Thanks for reading – see you tomorrow!

James G. Williams