Key Insights: Renée Cummings Discusses the AI Executive Order
All Tech Is Human Senior Fellow for AI, Data, and Public Policy Renée Cummings joined All Tech Is Human Founder and President David Ryan Polgar to discuss President Biden’s Executive Order on Safe, Secure, and Trustworthy Artificial Intelligence on Monday, Oct. 30, 2023.
Immediate takeaways
It certainly puts a lot of power in the wings of the Vice President as she heads to the UK to be at the Safety and Security Summit that's happening. It also positions the United States in a good place because we've just set the agenda as to what is probably going to be discussed in the UK. When it comes to the Executive Order, it is proactive to a particular extent, reactive because we know how long we have been waiting for something with a little more bite. It has a very paternalistic feel, but that is also in keeping with the fact that it offers particular protections.
What I was hoping for…was something that was a bit more participatory. Something that was creating some more partnerships. with communities, with underserved high needs groups, with impacted communities, individuals who have been marginalized by technology, offering some kind of leadership as well. So it is very much a sort of top-down approach. The other challenge is the technology is always ahead of any kind of legislation or compliance or anything that looks at enforcement.
So we have a great response that is required at the moment for us to really bring [a] Responsible AI feel to the space. But we know that government is also about bureaucracy - and all of what has been outlined in this Executive Order needs a whole of government approach for implementation. So we've got to now think about questions around bureaucracy when we need real time leadership at the moment in this space.
When we think about a technology like generative AI and how quickly that is moving, is this going to be executed at the same speed at the same scale?
What do you like about the AI Executive Order?
It's multifaceted. It looks at the ethics. It looks at the legal. It looks at the social. It looks at the environmental. It really handles all of the questions that have been spiraling AI for as long as we've been having these conversations.
The big question? What would be the influence on a domestic level, on a national level? What is going to be the influence? Of course, it touches all the key points: safety, security, privacy, equity, questions around justice, questions around civil rights, questions around leadership and whether or not this is really going to position the United States in a leadership position abroad. So it's something that is well received.
The Potentials and Limits of the AI Executive Order
We know it's voluntary. It's an Executive Order. It is not legislation. Ot is not about enforcement. It is about a request that's being made at the Federal level a way in which we can probably sort of leverage federal purchasing power to have some sort of impact on the totality and the broadness of the AI market".
Who is responsible for building the AI Executive Order - and who should be included in the conversation?
There are lots of things to be done. Lots of assessments to be created. Lots of guidelines to be created. Lots of different types of, cautionary procedures to be executed. So it really comes back to the question of AI literacy across the federal government and whether or not this can be impacted.
I really always think about putting the community, if not at the center, to lead. Because this technology is having real life implications on individuals. The Executive Order does pay attention to questions around equity and questions around fairness and questions around justice.
It highlights an approach that landlords need to take, federal benefits agencies need to take, and the labor industry needs to take as well. So some very critical things touch. Of course, I'm hoping that in the spirit of equity, which really guides this document, that even the White House would think about bringing a more equitable approach to the ways in which this need to be done.
We always talk about who's at the table, and we know sometimes in government, it's the same people who appear at the table all the time. So I think the White House may need to broaden its net a bit and really engage with different, more diverse stakeholders. I think that's very, very important. As I said, this Executive Order, it's comprehensive. It's a great start. It's really a document that we can really press into service. But we've got to get a much more diverse group of individuals trying to work with the different Federal agencies. There is a part of it that speaks about really an aggressive approach to hiring AI talent again. We've got to ask ourselves: Who is going to be hired? Is it going to be diverse? Are all those hirings going to be experts? We really need to sometimes rethink the meaning of experts in particular when we're seeing the kind of impact this technology is having on society. We've got to ask ourselves who we're going to hire. Is it going to be diverse? Are all those hirings going to be experts? We really need to sometimes rethink the meaning of experts in particular when we're seeing the kind of impact this technology is having on society.
How does the Executive Order balance a competitive landscape with guardrails?
It does look at collective bargaining. It does look at the future of work. It pays attention to questions around displacement and upskilling in real time.
And it really looks at building some sort of labor movement to really support the innovation at speed and at scale and looking at the ways in which we need to protect workers. There is something there and it's something that we can build upon when it comes to that question about whether or not we can do ethics or we can do innovation.
I always say ethics and innovation could live in the same house. Ethical innovation is the best kind of innovation. And I think ethics affords us an opportunity to stretch the imagination of innovation. So ethics by design and ethics by development and ethics by deployment. I think we can do those things and we can do those things well.
Ethics is not going to slow innovation down. What I would like us to think about when we're thinking about this Executive Order is: How are we going to engage stakeholders in real time? How are we going to engage communities? Where are the impacted voices? And how are those voices leading the conversation?
How can we understand this AI Executive Order? What are key ways to stay involved?
I just want people to know that this Executive Order is about you. It is about all of us. This is about our lives. This is about our future. And this is about our legacy in this country. It's about society and it's about the impact that every aspect of data and technology and AI and generative AI is going to have on all of us.
I say wherever you are - if you're in high school, reach out to your teachers, make this a discussion in class. If you're in university, have conversations around it. If you are in the workforce, speak to your HR department, reach out to politicians, get involved.
I think what is required at this moment is public oversight. Public oversight is still something that is very, very powerful when it comes to accountability and transparency, which is what we need. And if this is supposed to turn into something that protects us all, we need to be part of that process.
We need to be part of that process because at the end of the day, when it comes to the impact of this technology, I think impacted people would know how to build those protections because they have felt the brunt of it in ways that many of us may not have felt.
Renée Cummings, All Tech Is Human’s Senior Fellow for AI, Data, and Public Policy, is a 2023 VentureBeat AI Innovator Award winner. She is one of the most esteemed and respected voices in AI.