Here’s what’s missing from the White House’s National AI Strategy

Yesterday, the Biden Administration announced new efforts to “advance the research, development and deployment of responsible artificial intelligence (AI) that protects individuals’ rights and safety and delivers results for the American people.”

There were three announcements: An updated National AI R&D Strategic Plan from the White House Office of Science and Technology (OSTP) that provides a roadmap for key priorities and goals around federal investments in AI research and development; a new request for public input on critical AI issues; and a new report on AI risks and opportunities related to education.

In addition, the White House hosted a listening session with workers — from diverse sectors including call centers, trucking, warehousing, health care and gig work — to hear firsthand experiences with employers’ use of automated technologies for surveillance, monitoring, evaluation and management.

VentureBeat discussed these announcements, as well as the federal governments AI regulation efforts overall, with Sarah Myers West, managing director of the AI Now Institute and former senior adviser on AI at the Federal Trade Commission. This interview has been edited and condensed for clarity.

VentureBeat: What were your initial thoughts on the White House announcements, and are you noticing an increase in the number of announcements the Administration has released recently?

Myers West: We’re seeing a very swift pace of movement from the White House across many different fronts, on how AI is impacting the broader public. That was helpful to see. I was a little concerned about some of the movements of the prior weeks, where things seemed much more like centered on industry interests. So I’m happy to see a broader focus from the White House.

VentureBeat: Is there anything that you think is not broad enough, or anything that you think is missing as far as a critical focus?

Yes. There’s very little in this plan that addresses the role of big tech companies and how much there are structural dependencies on these big firms, which have been in the crosshairs for Congress and the White House over the past few years. There’s very little acknowledgement of potential concerns relating to the dominance of industry in this field. That plays out across many fronts — computational power, access to data — but also the fact that in the academic fields there’s been such a rise in dual affiliations, where professors will work at a university but they’ll also work at a company.

Company participation in the leading machine learning (ML)conferences has, I think, doubled over the last decade. So there’s no acknowledgement of the heavy amount of industry concentration and how it’s shaping the future trajectory of AI.

VentureBeat: Do you have any thoughts on just how that will play out as far as AI regulation, whether by the federal government, its agencies or states?

What they released yesterday is about where they’re going to as far as the research and development focus. So, what are they investing in? What do they see as places that are needed for further development? What that map shows us is that there are some fundamental concerns around accountability. They acknowledge the need for sustainability. That there are some issues if you’re just relying on a model where it’s only up to the companies to do their own testing of the system. We need a new approach there.

They acknowledged that there’s going to be a huge impact on the workforce and that we need to be addressing the needs of workers. So I think this is more about looking forward — here’s the space where we know these things are important and we need to develop further resources to be equipped to meet the needs of the moment.

From an accountability perspective, there’s going to be this multi-layered and likely confusing set of approaches where the federal agencies say they intend to apply existing law. So, keeping an eye out for investigations or announcements that are coming from the enforcement agencies like the Federal Trade Commission and the Consumer Financial Protection Bureau.

The Equal Employment Opportunity Commission has its whole workstream on AI and hiring. So there’s a multifaceted front there. And then at the state level, there’s been a number of different bills that are starting to crop up to tackle generative AI, to form AI task forces. The state attorneys general can open up cases. So it’s going to be a very active space in the months to come.

VentureBeat: What do you think about Open AI’s recent blog post talking about super intelligence, these futuristic talks about impact? Do you think that that is relevant, or do you think it is distracting from the current issues that need to be addressed?

I feel like I’ve been like a broken record on this issue for a little while. It’s taking away from the harms that people are already experiencing in the here and now. I think, galvanizing around enforcing the existing law, ensuring that the public is being served by these technologies in the moment is what’s really key.

OpenAI has taken this position for a long time. I was looking back at some of their work from 2016 and their whole reason for being as a company is that they want to anticipate and be the first to move on the development of artificial general intelligence, to capture it so no one else can develop it and use it for harm. That’s been their orientation from the start, but I don’t think that that’s the right orientation for regulators to take.

I think our focus very much needs to be on what’s happening immediately around us, and what’s happening is workers wages are being pushed down. Their work is being devalued. There are immediate climate impacts on the deployment of these systems. There are immediate impacts on the information ecosystem. That’s that’s where our focus should be right now.

VentureBeat: Is there anything that you are expecting to come next down the pike from the administration that you’re eager to see?

The timeline is so fast on this stuff right now. The worker surveillance request for information (RFI) is one place where I think movement from the White House would be really welcome. It’s something that’s impacting workers across industry sectors, across all levels of management. And I think the the spread of this technology has been so rapid, especially post-pandemic, that it’s very prime for intervention.

The FTC has a cloud computing RFI. I think that’s really important for some of these big questions about dominance in the market by the companies that are cloud infrastructure providers like Amazon and Google and Microsoft. So that’s that’s one to watch. From these announcements yesterday, though, they will be developing a national AI strategy in addition to the research and development strategy. So I’m looking to how they intend to implement it and provide strong safeguards for the broader public.