Scenes from a D.C. summer
Conversations in Washington reveal how things really work
I taught a course on history and policymaking this summer at the Maxwell School for Citizenship & Public Affairs, and as a result I’ve spent June through August at our house in Washington, D.C. I used to spend every summer in D.C. when I worked at the Library of Congress. But since then, I’ve had four years as the founding director of the Lepage Center for History in the Public Interest and two years as a speaker and author traveling across the U.S. and Europe. I would be in D.C. for part of the summer, but rarely in D.C. for the entirety of the summer.
Summer in D.C. has its own idiosyncrasies. For starters, it’s brutally hot and humid outside, which when coupled with the unrelenting air conditioning inside, means you’ll sweat through your clothes on the way to work, then need a sweater while at work. (As an aside, in the early 19th century one of the reasons people didn’t believe there’d ever be career politicians was because no one wanted to live in D.C. year-round.)
Summers in D.C. are also full of events: breakfasts, lunches, happy hours, dinners, briefings, conversations, panels and screenings. I’ve been fortunate to attend quite a few, including breakfasts with Members of Congress, lunches with diplomats, happy hours with overseas visitors, dinners with journalists and conferences with advocates. That has led to some interesting conversations. Allow me, then, to share a few of those conversations with you, and explain how they connect to the most frequent topic of conversation in D.C. this summer, artificial intelligence—revealing, I hope, some of how Washington works.
Scene I
Earlier this summer, I went to an evening function at a townhouse in Kalorama. This is the neighborhood where the Obamas and the Kushners own properties, among many other well-to-do Washingtonians. At the event, I was chatting with a well-placed Capitol Hill staffer. We had a lovely chat about a variety of subjects, during which he shared a story about a recent piece of legislation.
He told me that he and his team had been drafting a piece of legislation that would have attempted to reign in a powerful segment of the economy. That’s when the lobbyists emerged. Corporate interest groups descended on Capitol Hill and met with every Member of Congress involved in the negotiations. The lobbyists pressed them on why this legislation would not be in their best interests. They spent $300 million attacking the proposed bill. In the end, it worked; the legislation did not go forward.
Whether this piece of legislation would have had much impact, we’ll never know. For our purposes, though, it’s a good reminder of how much money is spent in D.C. ensuring that certain laws do not get passed. Many people are invested in the status quo, and will spend a lot of money to maintain it.
Such lobbying efforts are business as usual in Washington, though some might say they are a violation of the public trust. Is $300 million to defeat a bill evidence of regulatory capture, or a manifestation of free speech? Is it corruption, or simply a side effect of democracy where societal actors express their desires through paid advocacy?
(As an aside, $300 million would provide stable housing for 8,500 homeless veterans.)
Scene II
A bit later in the summer, I conducted an on-the-record interview with Elton Skendaj of Georgetown University for an article I wrote for Devex (you can read the article here).
Elton is a scholar of corruption—and being from the Balkans, he has a lot of first-hand experience. He witnessed much corruption in Albania in the post-Yugoslavia years, and he noted that what constitutes corruption varies from location to location. “Sometimes what’s locally accepted would, internationally, be considered problematic,” he told me.
He continued: “In the U.S., we have a practice in which you have lobbyists from a particular industry that write the legislation for said industry and provide that legislation to Members of Congress that they have sponsored in electoral campaigns. In the rest of the world, we would call some of this state capture. In the U.S., this is fully legal.”
In the Balkans—which readers will recall I visited in May when I traveled with the U.S. Department of State—Elton said there is another issue that some may view as corrupt while others do not: nepotism. In his view, much of the poor administration in the Balkans results from people holding positions that they are not qualified to hold. They get government jobs because they are personally or politically connected, not because they are qualified. For many, this is business as usual. To others, it’s a violation of the public trust.
“In the Balkans, we found that when international organizations gave authority to local parties to build state administration, the state administration became bloated, hiring people based on connections,” Elton said. “That meant the state capacity was weak.”
Scene III
As mentioned, I taught for eight weeks this summer at the Maxwell School for Citizenship & Public Affairs, housed within the Center for Strategic & International Studies, one-half mile from The White House. The course, which I designed, focused on how to bring history into policy, and we spent much of the class talking about the structural challenges in bridging these two worlds.
For example, Members of Congress and political appointees cycle through their positions every two, four or six years, and there can be immense pressure to demonstrate accomplishments during their tenures. Meanwhile, it might take a historian six years to write a single book. (Mine took five-and-a-half). How do you marry scholarship with lawmaking when they operate on such different lifecycles and have such different incentive structures?
My students—who were amazing—had some ideas. Yet, as we delved into solutions, we returned to a recurring problem: you can research and present the most compelling and evidenced-based historical scholarship to a decision-maker, yet if it does not align with their worldview, party position or lobbying interests, they might reject it. One person told us that he and his colleagues presented evidenced-based historical research to a key decision maker within the U.S. government, only to have that person ignore their recommendations and fire the people who presented them.
How this connects to A.I.
These conversations with scholars, students and policymakers underscore how things do or don’t get done in Washington:
Sometimes people in government want to act, but they get stifled by corporate interests;
Sometimes people in government act, but in the interests of themselves or their close allies, not the broader public;
Sometimes people in government refuse to act, even when evidence is placed in front of them.
All of this has been apparent in the discourse around A.I., a topic that has dominated D.C. conversations this summer. A.I. has been discussed by think tanks, news outlets, committee hearings, working groups, commissions, reports, universities, corporate entities and civil society organizations (including my History Communication Institute).
As part of being media literate information consumers, it’s incumbent upon us to ask why so many people are talking about A.I. and not, for example, homeless veterans. The answer, in large part, is money. The federal government is on the verge of committing billions of dollars towards artificial intelligence, and numerous stakeholders are lobbying to dictate where that money will go and what it will (or won’t) regulate. Consider:
Congress and The White House have agreed to create a National Artificial Intelligence Research Resource (NAIRR), with a proposed budget of $2.6 billion;
The Biden Administration requested from Congress $1.8 billion in defense spending on A.I. in fiscal year 2023;
The National Science Foundation (NSF) received $700 million for artificial intelligence research;
The Department of Energy and National Institute of Standards and Technology are each asking for millions of dollars in A.I.-related funding;
Corporations and universities are asking for funding to train students in machine learning, cybersecurity, processor design and other A.I.-related disciplines.
We should recognize, then, that much of the conversations in Washington these days about A.I. are, essentially, an extension of lobbying:
Companies are lobbying government to ensure legislation and policy aligns with their interests;
Universities and professors are lobbying government to ensure that funding is directed towards their labs and research programs;
Individuals are positioning themselves as experts in order to be considered for government jobs, roles on advisory boards, witnesses at Congressional hearings, invitations for speaking opportunities and even potential book deals. (This happened to me; someone advised me to rush out a book proposal on A.I. in order to capitalize on the buzz).
Talking about A.I. can be almost as lucrative as developing A.I.
As it relates to the aforementioned scenes, then, per Scene I: corporate interests have descended on Washington speaking publicly and spending mightily to ensure that A.I. legislation aligns with their interests and maintains the current course of business. The goal is to continue to develop A.I. while also continuing to earn billions of dollars. Microsoft, IBM, Google, Meta, Amazon and others have embraced A.I. regulations so long as they get to shape them. According to a story in the L.A. Times, 123 companies, universities and trade associations spent a collective $94 million lobbying the federal government on issues including A.I. in the first quarter of 2023. Part of that lobbying effort includes hosting events, publishing op-eds and sending representatives to speak at responsible tech conferences.
That dovetails into Scene II, where, if you recall, Elton explained that in the U.S. we have a system wherein lobbyists from a particular industry write the legislation for said industry and provide that legislation to Members of Congress that they have sponsored in electoral campaigns. The U.S. Chamber of Commerce spent $81 million on lobbying last year (including donations to both parties), and this year has drafted legislation and executive orders for several A.I.-related technologies, according to Open Secrets. When California Representative Anna Eshoo, a Democrat, unveiled her recent CREATE AI Act, she specifically touted the support it received from Microsoft, Anthropic (an AI startup), and Seed AI, whose CEO is a former government affairs staffer at NVIDIA.
As it relates to nepotism, here the view expands beyond Washington to include universities and civil society. Schools such as Carnegie Mellon, Case Western Reserve University, Harvard and Stanford are spending large sums of money in Washington because they want their faculty in computer science, business, economics and advanced computing to sit on government advisory committees, serve as expert witnesses, and ultimately, have relationships with federal agencies who will approve grants, research labs and infrastructure. Their experts often have deep connections to government and corporations. Take for example Stanford’s Fei-Fei Li, who was a member of the NAIRR task force that recommended its $2.6 billion budget; she formerly served as a Google executive and her professorship is endowed by Sequoia Capital, a multi-billion dollar venture capital firm in Silicon Valley.
This brings us to Scene III, and the structural challenges of fitting history and humanities scholarship into this ecosystem. Hype cycles such as A.I. often take on their own momentum, which compels policymakers to strike while the proverbial iron is hot. Political mandates can be fleeting, and policymakers want to demonstrate to their constituents that they’ve taken swift action when they have the opportunity to do so. On the flip side, rigorous humanities scholarship that grapples with technologies and their effects takes a long time to produce. If I was to write a book on A.I. with evidenced-based recommendations, it could take years before it was published. (There is a chapter about A.I. in my existing book, and it is quite insightful, if I do say so myself.)
Humanities scholarship is being produced about A.I. and its effects. But by the time much of it appears, the hype cycle may be over, billions of dollars in funding will have already been allocated, and the conversations may move onto other subjects. Even if the scholarly work was presented to Members of Congress, would they put aside the interests of Google or Sequoia Capital to engage with it?
Closing thoughts
As mentioned, we have had far more conversations about A.I. in Washington this year then, say, homeless veterans. In fact, in all the conversations I’ve had in D.C. this summer, homeless veterans, and the homeless more broadly, have not come up.
This, too, is an important point for media and historical literacy: what we choose to talk about at certain moments in time is not a neutral creation. Conversations don’t just happen. They are shaped and steered, often by powerful interests who seek certain outcomes. There are numerous agendas—corporate, academic, personal—at work in any discourse; our task as historically literate and media literate information consumers is to identify them and determine what their agendas may be, positive or negative.
That reminds me of a final conversation I had this summer, with my wife. “During the Obama Administration,” she and I remarked to each other one evening with some humor, “everyone was an ISIL expert. Then during the Trump administration, everyone was a disinformation expert. Now during the Biden Administration, everyone is an A.I. expert.”
Whenever the federal government is investing billions of dollars, people will try to position themselves accordingly and lobby for their particular interests. Is that corruption, or just an expression of how America does democracy? Whether you believe the answer is yes, no, or some of both, in this town, it’s how things work—even in the summer.
Speak to you in September,
-JS
Insightful comments. It reminds me of all the political, scientific and governmental hype years ago about nanotechnology. The lyrics change but the song remains the same. (Apologies to Led Zeppelin, and songwriters James Patrick Page and Robert Anthony Plant.)