If you squint at today’s AI policy debates, you may see the Telecommunications Act of 1996 in the distance.
In this conversation, Matt Perault, head of AI policy, a16z, sits down with Adam Thierer, resident senior fellow, technology and innovation, R Street Institute, and Blair Levin, policy analyst, New Street Research and non-resident senior associate, Center for Strategic and International Studies, to revisit their first-hand experience tackling a similarly significant moment in tech policy: a small number of incumbents with entrenched market power, a messy patchwork of federal and local rules, and misaligned governing authority. The result then was federal preemption coupled with a comprehensive national framework for telecommunications—all through a bipartisan deal.
A few themes from their discussion stand out:
Preemption was a tool for achieving a common goal
In the 90s, Congress wasn’t debating preemption in the abstract. They were trying to solve a concrete problem: how to open up competition in telecom markets that had been shaped for decades by monopoly and regulation.
As Blair explains, the “big bargain” of the Telecom Act allowed local phone companies (the Baby Bells) to enter new lines of business like long distance and video. In exchange, they had to open their last-mile networks to new entrants on fair terms. Federal rules were important to prevent state regulators from simply protecting the incumbents they were closest to.
Adam points to Section 253, which explicitly limited state and local barriers to providing telecom services, as one of several provisions aimed at creating “a more unified, clear-cut national framework.” Preemption here was in service of a larger goal: making sure new entrants and disruptors could compete in order to provide more choice and better experiences for customers.
Comprehensive doesn’t have to mean clean
Judging by their war stories, no one at the table has any romantic illusions about the Telecom Act. Blair quotes Justice Scalia’s comment that the law was “a model of ambiguity,” and notes the 110 rulemakings that followed as the FCC and courts wrestled with text shaped by legislative horse-trading.
The act mixed strong assertions of federal authority with carve-outs preserving state and local powers–sometimes in the same section. Adam recalls working on language to open wireless and satellite markets, only to see parallel clauses preserving local control over tower siting and other details. That tension fueled years of litigation and FCC proceedings.
Yet both Adam and Blair argue that, on balance, the national framework worked, not because it resolved every issue cleanly, but because it created the conditions for new markets to form.
Leadership and vision matter
If there’s a through-line from the Telecom Act to today’s AI moment, it’s that major technology frameworks don’t just materialize, they happen when our country’s leaders come together around a common goal and vision for our nation, and are willing to build it.
In the 1990s, that leadership took multiple forms: a bipartisan belief that innovation was good for Americans; a willingness to negotiate across industry lines; and a White House that articulated a clear, pro-growth, pro-innovation framework for US technology leadership. Rather than use the same analog rulebook for governing legacy networks, the tech policy of that era helped to define the open environment from which internet companies grew.
Today’s AI debate sits in a similar moment of possibility. As Adam describes it, this is a “constitutional moment for technology,” where the U.S. must choose whether AI will develop under a unified national framework or splinter and stall through 50 competing regulatory regimes. Blair underscores the stakes from a different angle: public skepticism is rising, institutional capacity is thin, and Congress has not yet stepped into the role only it can play.
This transcript has been edited lightly for readability.
Blair Levin (00:00)
So the big idea was competition, but it required a certain kind of regulation to make sure that competition was fair.
Our biggest fear was the Bells would just dominate everything. And that would not lead to the competition we wanted. And so we kind of bent over backwards to interpret the law. We were usually upheld to enable the disruptors, the new entrants, to be able to make a play for local voice.
Adam Thierer (00:27)
When we passed the Telecom Act, it was a new day and we basically started shifting more the focus to this being a national concern and a national marketplace. And we got that right. And I talk a lot about innovation culture and about Congress and the White House sending a clear signal on certain key technologies and sectors for the American economy and for American strength. And we did that finally. We said, no, this is a national matter. We need that for AI too.
This is what I’ll call a constitutional moment for technology. Like what will our new innovation policy be for the next 30 years?
Matt Perault (01:07)
We are in the middle of a discussion of preemption and like many discussions in policy and politics, the feeling of it is if this is the first time that Congress, the federal government has thought about what the relationship should be between an important topic in tech policy between the federal government and states. Blair and Adam, you lived through a really important debate where Congress and the states were thinking about how to balance affirmative governance and also outlining clear guardrails between what the federal government should be doing and what state governments should be doing. So you are the perfect two people to be joining us today to talk about the history of preemption and important tech and telecom policy debates. Thanks so much for joining us.
Blair Levin (01:46)
Thank you.
Adam Thierer (01:47)
Thanks for having us.
Matt Perault (01:48)
Blair, maybe you could start if that’s okay. Could you tell us a little bit in the mid 90s as the Telecom Act was coming together, what was the project? what was Congress and the administration, what were they trying to achieve? And then how did preemption figure into that conversation?
Blair Levin (02:05)
So it really started quite a bit before the bill passed in 96. As soon as the breakup of AT &T ended, that is to say the judge says, okay, here’s the consent decree. You have your local ⁓ carriers, what were known as the baby bells. You have the long distance companies. The baby bells who were prohibited from entering certain lines of business were pressuring Congress to say, let us do various things. And the momentum for that grew. There were hearings in the 80s. There were lots of hearings in the early 90s. The bill almost passed in 94 and then went to 96. And the big idea behind it was the local phone companies could enter long distance, they could enter video. We would no longer have siloed, protected businesses. But in exchange for that, the local companies had to provide on some basis, and that was a big debatable point, on some basis, they had to provide access to the last mile network to long distance carriers and to others. And so that was the big bargain. And the fear was that the states would be very friendly to the incumbent local exchange carrier, the ILECs the Baby Bells, because they had a lot of employment, they had a lot of political power with the state utilities.
And so the idea was you give the federal government, particularly the FCC, the authority to set the terms and conditions by which the new entrants to voice could have, particularly local voice, could have access to certain networks, facilities, in exchange for the local guys then being able to enter any other kind of market. So the big idea was competition, but it required a certain kind of regulation to make sure that competition was fair. At the heart of it was the so-called 14-point checklist, which said basically the local guys had to complete 14 things, all of which were controversial, before they could enter a long distance. And that was at the core of it.
There were a lot of other things in the telecom act, broadcasts and cable and a bunch of other stuff. Adam, did I get that right?
Adam Thierer (04:26)
Yeah, no, that’s all correct. You know, the amazing thing about this history, first of all, let’s remember what the history is, because a lot of people today haven’t lived it. They didn’t grow up in the age of rotary dial telephones and simple broadcasts from radio and television operators that you could count on your hand and how many you had in your community. You know, the history of American telecommunications and media policy has not been a pretty one.
For the most part of the past century, a convoluted thicket of federal, state, and local rules and policies controlled information and communications technologies through a variety of clunky mechanisms meant to ensure the public interest was served. All too often, unfortunately, it was not served very well. We had essentially regulated geographic monopolies in telephony, cable, broadcasting, not due to natural technology circumstances or market circumstances.
But mostly to policy, in my opinion. We had operational licensing and permits. We had line of business restrictions that created direct barriers to entry and exit from markets that limited innovation. We had price controls, rate of return regulation, subsidies that distorted markets in strange, bizarre ways. We had technical device and equipment regulations and quality of service requirements.
And this wasn’t just the Federal Communications Commission doing all this enforcing. A lot of it was done, as Blair noted, by public utility commissions at the state level. And then there was still more meddling even at the municipal level in some cases. And on top of all of this economic and technocratic regulation, there was a whole bunch of speech regulation too. We can’t forget about that.
But as Blair noted, the thing that really set the wheels and notion for federal reform was the fact that following the 1982 consent decree that the Department of Justice struck with AT&T, the so-called modification of final judgment, a lot of authority passed to the courts and really to one man, a man by the name of Judge Harold Green, to basically be the enforcer of this consent, this antitrust consent decree. And it had sort of removed from Congress and the FCC and even the PUCs a lot of authority.
Well, nothing wakes up members of Congress like a good old fashioned turf war. And like a question of they’re like, who runs the show? You know, who’s in control here? And so I think it was this, Blair, you can correct me on this. I believe it was as early as 1988, we saw the first bill dropped. And I believe it was by Senator Bob Dole that basically would have done something as simple as just say, we’re going to take the power that we ceded to, you know, the DOJ to do this and MFJ and Judge Green, we’re going to pull it back into Congress.
And so it started with the simple idea that Congress should have more say about the state of competition policy in the telecommunications market. That led to numerous debates over numerous sessions of Congress where that bill or that simple idea started growing and growing. And it went from just a few pages to dozens to by the time of the Telecom Act, hundreds of pages and a lot more technical, technocratic deregulation as I’m going to put in air quotes in our conversation because a lot of people think we massively deregulated the Telecom Act, which is just not true.
Blair Levin (07:39)
We had 110 rule makings that we had to do that’s not exactly deregulation. And I would just note that, yeah, it is kind of an ugly path, but as Shakespeare, I believe, wrote, the path of true love never did run true. It’s always messy
Matt Perault (07:41)
Yeah, well, in both of you, I think have a path of true love with this set of issues. So I’m curious, what was the vision that the federal government, the Congress had for what would be federal and then what was left to states? And how would you characterize that balance? Adam, what’s kind of like your top line take on like how you would think about what Congress is doing and what was left to states?
Adam Thierer (08:16)
Yeah, well, Congress came right out in the Telecom Act and said, and all the floor talk about this, even from people who are on opposite sides of it, pointed out the need for some sort of a national framework of a more unified approach to ensuring that telecommunications and media markets became more competitive. There was a legitimate desire by all to ensure that result, but it was just a question of the means to getting there. But they made it very clear that, and in fact, maybe I’m jumping ahead in conversation, but what inevitably went into the Telecom Act was language in section 253 that said no state or local statute of regulation may prohibit or have the effect of prohibiting the ability of any entity to provide any interstate or intrastate telecommunication service.
And there were several other provisions in there trying to figure out how to create a more unified, clear-cut national framework, or at least clean up some of those old messes of the past and limit some of those state powers, although not completely, we can get into the details here, because there was a balance. And Congress, in one breath with Section 253 said that, but then in the very next clause said something about basically other exemptions to it. And a variety of other provisions came into the bill over time that I’ll hand it to Blair to go into more detail.
Blair Levin (09:32)
I think the details, there’s a million details. And of course, there was a big Supreme Court case on this at which Justice Scalia wisely wrote, the act is a model of ambiguity, which was confirmed by a senator who said to Reid Hunt, the chair of the FCC, Reid, here’s what we did. We put in everything that one side wanted, and then we put in everything the other side wanted.
Like I said, there was this 14-point checklist and there were eight terms and conditions. There were a thousand different things. And a lot of the debate was, what was it that the FCC could preempt? We were basically on the side of preempting states a lot because what we saw was, look, whenever you do these things, you’re going to make a mistake one way or the other. And you can argue we made certain mistakes.
Our biggest fear was that after all of this work and after so many thousands of hours on Congress, that the real fear was that the Bells would just dominate everything. And that would not lead to the competition we wanted. And so we kind of bent over backwards to interpret the law…we were usually upheld… to enable the disruptors, the new entrants, which included the long distance industry to be able to make a play for local voice. I think as history went on, the markets changed and it didn’t really matter that much in some ways. We could chat about that. But our basic point was where there was ambiguity to interpret it to enable that new investment, that new innovation.
Matt Perault (11:09)
Blair, in your work for New Street Research, which is a Wall Street research firm, you produce regular research for them. You’ve often gone back to the value of ambiguity in terms of policy achievements, like the political value for bringing people to the table. Blair, and Adam look at this from different political perspectives.
I think one of the miraculous things about the Telecom Act is just that it exists, that it was able to get through the political process. I think sometimes when we look at current issues in AI policy, it seems like that would be incredibly challenging.
So I’m curious about the political mechanics. Like how did this get to done? What can we learn from that and thinking about AI policy and then Blair, you’ve done so much thinking about specificity and ambiguity, like how does that play into the achievement of the Telecom Act?
Blair Levin (11:58)
It’s a great question. Look, there’s a lot of credit that goes to a lot of different people and I won’t name all of them, but really on both sides of the aisle. I will, Bob Dole, for example, was a real terrific leader in the Senate. Tom Bliley, the former mayor of Richmond, the Republican head of the committee, and Ed Markey, who we relied on a lot, Fritz Hollings. It was a very different Congress. They held a bazillion hearings on this stuff. They worked on all of these different elements. But I’ll tell you what I think is really different. I’ve had a lot of conversations with various Democrats about how did the Democrats lose the tech industry? I actually disagree with almost every bit of analysis I see. But I also don’t think it’s that relevant. Here’s the big question. How did the tech industry lose the public?
When we were doing this act, I think by an order of magnitude, 75% of the public was excited about the opportunity to use the internet better, about what technology could bring. You look at the polls now, it’s exactly the opposite. It’s like 25 % think that AI is positive. We could go into why the public believes that. It’s not what I believe necessarily, but having said that, being pro-tech, being pro-innovation is a much harder thing to do politically. Now I actually think the tech industry in the last year has undercut their own case in a variety of ways, we could go into that. But I think that’s part of the problem, the tech industry is not trying to persuade people in the way that I think the long distance guys, the local guys, the cable guys, everybody in Washington in that run up to the act actually tried to do a lot of persuasion to meet various folks.
By the way, you know, as Adam will recall, Reid and I met with Adam, who was very much on the other side. I used to go to meetings with the Progress and Freedom Foundation. I’ve never talked about this publicly. They had a big influence on us. We didn’t usually adopt what they said, but they kept us from doing things that were a lot stupider than we were thinking about doing. And I just don’t think that process exists today in the same way. There’s not the same good faith.
Adam Thierer (14:08)
You know, of the interesting things about the Telecom Act era, as Blair alluded to, I mean, the politics were quite different. There were very strange bedfellow alliances that went on behind the scenes. But a lot of it was defined a lot by where you sat as a congressman in this world in terms of who you were aligned with in industry. There were like, are you a long distance man? Are you a Bell man? Are you wireless guy?
And it was very strange because you had Republicans who are going against each other, each other’s throats, because one would be more affiliated with long distance thinking versus local RBOC regional bell operating company thinking. So that was different. There were other things that were going on behind the scenes. There were a ton of different special interests involved in shaping the Telecom Act and getting very strange things inserted line by line that are relevant to the preemption stuff.
Adam Thierer (14:57)
Because as I said, with each line that says we need to have a national framework and ensure greater competition, we’d have another line that undercut it by saying, but we want to make sure that state and localities can do all of these other things. This was true for things like wireless markets, where we were trying to open those up at that time. People forget we didn’t have nationwide spectrum and wireless markets, but we were starting to for that and for direct broadcast satellite. And we tried in the Telecom Act to negotiate some language about that.
I remember working on it very directly with Senator Pressler’s office and Senator McCain and others to try to get more amendments on that. Each time we did, there was an effort by the other side to come in and say, no, we need to have this preservation of state authority. It ended up being here. I’ve got it here, section 704 of the telecom act. Basically, after saying we’re going to open up the market, said nothing shall limit the authority of state or local government regarding the placement, construction, or modification of personal wireless facilities.
And so you had this immediate tension, right? We’re going to have more competition. We’re also going to have more local regulation and NIMBYism. And this pervaded a lot of what was in the Telecom Act. And we lived with that for many years after. Like going back and forth, were numerous FCC proceedings about this and cases about small cell orders and 5G siting issues, Muni broadband.
And, you know, back in the day, I was much more in favor of like striking those compromises and a lot more sympathy for a lot of those state and local governments than I do today because I now see the mess that it created and that it really did undermine a lot of important innovation and competition and led to like only certain larger players being able to benefit from from that market.
And the areas where we did not touch in the Telecom Act were the oldest and the newest of the information and communications technology world. We didn’t touch publishing newspapers and books. We never really did with federal regulation. And we didn’t touch computing for the most part. We left those alone. The oldest and the newest were essentially born free. But everybody else was born into a different type of regulatory arrangement. And we tried to optimize for the suboptimal by saying, well, let’s balance here. Let’s try it with a Goldilocks formula. But where we left things alone, I would argue, we actually got the best results.
And I would include in that, to some extent, the wireless market, which at the time, Blair and the FCC did wonderful work in helping open that market. Originally, we had beginnings of regionalized monopolies there. We were going to give local carriers a half monopoly on the market and then one other carrier. And then we opened it wide up. started auctioning spectrum. So where we allowed markets to work and technology to work its magic, things worked out best in my opinion. And where we didn’t and micromanaged it, we’re still today fighting over those issues.
Blair Levin (17:39)
Yeah. Can I just say, I agree with all of that, but I have to add one important point because there were two things that we did that history has completely ignored. And I would argue that may have been the most important things. And they both go to the same issue of what can the incumbent local phone company charge and what’s called terminating access charges. So whenever you connect to another, to that local carrier, they would charge you. And that this was part of long distance, that’s why long distance cost what it did, et cetera. As to wireless, obviously in the early 90s, when very few people had wireless phones, most calls had to be connected to the wired network. We basically said, you can’t really charge the wireless guys. It was only after we did that, that AT&T started what was called the OneRate and it started to be a mass consumer product. We did the same thing with the internet.
We didn’t regulate it, but we regulated on behalf of it. A very big issue was could the local, this was the only access to the internet was dial up in those days, could the local exchange carrier charge on a permanent basis, whatever they wanted? And we said, nah, you can’t do that. And we might’ve had, but the Bells pushed various members of Congress to require us to do that.
We might’ve had something to do with getting Steve Case, who ran AOL, to start a letter writing campaign. I think there were 400,000 emails, which seemed like a lot to Congress, and gave us the kind of political window to be able to say to the Bells, no, you can’t charge terminating access. And by virtue of that, internet service providers charged a set rate. You could use it as much as you wanted. I think those two things were really critical to the growth of the internet and to wireless. But it required essentially saying, we’ve got this existing monopoly that we have to make sure doesn’t carry it over to the next thing.
Matt Perault (19:41)
I think anything that’s trying to govern this comprehensively would have a governance component and then something that outlines the scope of what the federal government should do and then some carve outs to that for what is going to be left to the states. And there’s like this balancing between those three things, like how much are we governing or choosing not to govern, choosing to just leave decisions to the market or leave governance to the market.
What is the scope of what we’re trying to do at the federal level and how large or small should that be? And then what is the nature of the carve outs for states where we would say states can continue to act. It seems like in each of those, there was a set of compromises.
Adam, the way I hear you saying it is less like we got to the Goldilocks was, was a bug, not a feature. The nature of trying to get everyone, the compromises to please everyone got us further from something that was optimal. But it does seem at least from the outside and someone like from my perspective, I didn’t live this, that you get concepts that might not have been thought of as bipartisan to eventually become bipartisan and get us to a point when there’s actually some form of governance. I’m curious how you see it. How did you, for each of those three buckets, how do you get enough Democrats and enough Republicans on board to say, we’re okay with this moving forward?
Adam Thierer (21:01)
What are we talking about on the telecom act or AI policy? Those are two very different things.
Matt Perault (21:04)
I guess both. I think the question, the focus for us, I think, is like, what can we learn from the Telecom Act for AI policy?
Adam Thierer (21:13)
Yeah, I think it has some application. And again, I’ll go back to what I already said, that where we were clearest and where we had, I think, the most sweep in terms of offering a more serious national framework, had the most effectiveness. And where we had the most sort technocratic micromanagement, specifically with local telephony, we spent years in courts and wrangling in endless filings to the FCC and to the courts about how to allow for anything to happen at both the federal and the state level.
But with wireless, for example, and I’ll just give the particular example, direct broadcast satellite, which is already a technology that’s almost come and gone in this world in some ways, but we still have DBS technology out there. But the reality is, that that was a market that was just developing then. And we had to make sure that we essentially protected it as it came into being from a lot of state and local meddling and taxation. And so there were provisions that tried to deal with that new, exciting, new form of competition to cable television and basically allowed that to develop. And it was really crude. We basically, I remember sitting around in debates in the Senate trying to figure out, how do we do this, but also give the state some balance. And there was a lot of discussion about how to essentially just have almost like a pizza dish size rule that says if it’s as big as a pizza or a bit smaller, it’s gonna be national. If it’s bigger than that, let the states figure it out.
You know, these things are messy. They’re not easily done. But I think that where we did have the clearest rules and the more national framework and focus, especially with the internet, things turned out best in my opinion.
Blair Levin (22:49)
Yeah, it’s a great example, but if I could illustrate another aspect of it, the DBS really started to grow once the cable access rule, the program access rules of the cable act were passed. And what those rules said was that the cable networks, which were largely owned in part by the cable networks, I just said the cable channels were owned by the cable networks. So John Malone of TCI owned BET and CNN and Turner, and he prohibited them from selling to DBS.
Once you had the program access rule saying you have to sell on equal terms and conditions, DBS rose. That was great in terms of multi-channel video competition. But the big unknown story about that was it caused Malone and the cable industry to go, holy shit. These guys have a better cost structure and a better regulatory structure, 10 years from now, they’re going to totally beat us because now we don’t have them. We don’t have an advantage. And therefore we have to figure out another business. And that’s what got them into broadband. And that’s what got into the two wire of the competition. Look, the telephone guys had developed DSL technology in the 80s, but they didn’t want to use it for the internet because it was so profitable to sell dial up lines and sell two lines and all that.
And it was not the intent of Program Access Rules to create a new broadband competitor, but it happened. And I think it was one of the great accidents of policy in doing that. But the other more direct answer to your question, Matt, is I hate to sound like an old guy, though I am an old guy, but there was just an element of leadership. Like we’re here to do something.
And so you had Clinton and Gingrich and Gore and the other people that I’ve mentioned saying, we have to fix this problem, but it’s going to be great for America. Right now, have there been any hearings? Have there been any negotiations? It’s bipartisan, but really kind of going the other way. There just isn’t the kind of leadership that thinks it has to persuade people, that thinks it has to compromise. If you don’t have that, sure, you can do an executive order. The executive order, I’m not sure, has any legal standing. But there was a style of leadership that understood this. I mean, it was really interesting watching Clinton and Gingrich communicate with each other. And they’re both kind of Southern good old boys and they knew how to talk to each other, but they wanted to do things as opposed to simply command things. And I just, don’t, I don’t see that happening on the Hill though. Adam, you’re on the Hill a lot more than I am.
Matt Perault (25:35)
I think it’s striking the arc that you’re describing is actually often what I hear from Adam in the stuff he writes and what he says where he says, there’s a problem. We see X number of state bills. We need to solve it through leadership and a vision for what technology governance should be for America.
And that will benefit Americans in all these different ways, economic and social value. So Adam, is that right? And if that’s what you’re trying to say, and let’s say Blair is also correct that we’re not seeing that reflected, at least in how the policy process has been able to move forward, what do think the gap is?
Adam Thierer (26:09)
Yeah, okay. So first of all, let’s step back just a little bit and go back to the Telecom Act and what it did and did not do with regards to the internet and the world that we are living in today, because this is an important story. And this is a story that’s being changed as we speak. But in the 1990s, the internet was still very much an nascent technology and what we used to clunkily call the electronic super highway and electronic commerce was slowly building.
The Telecom Act largely forgot about the internet. To the extent it talked about it, it was to regulate decency online, right? We had the Communications Decency Act, which was later struck down in the courts. But Blair used the term accidents of history. Well, we had one of the most momentous accidents of history in this country when the Communications Decency Act was struck down, but not included in that strike down was Section 230, right?
And that really became the basis for the explosion of electronic commerce that followed. Likewise, the president had a vision, President Clinton had a vision that was articulated in the framework for global electronic commerce about the internet being a growth engine. And he predicted that we would see a massive explosion in commerce because we were going to treat the internet differently and not trying to pigeonhole it into yesterday’s analog age rule book.
And so this really became a crucial part of the story in terms of how the internet unfolded during this period. And really important to this story, especially for the preemption question, is the fact that the states were very much in the backseat on this. In fact, they weren’t even in the car. When it comes to the regulation of the internet, we did not have 50 different state internet bureaus, like we had 50 public utility commissions. We did not have comprehensive regulation of computing at the state level.
And so that became the new baseline, that and the fact that the president had established this new innovation culture with the global electronic commerce, the framework for global electronic commerce that basically treated it as the principal said, a global resource and platform. And that talked about policy should be facilitated on a global or national basis and avoiding convoluted borders and barriers wherever possible, learning from the past. That was then, this is now.
And to get back to your question, Matt, now is a world where the states have not only caught up, they’re in the driver’s seat. They’re completely controlling technology policy in the world of advanced technology, the internet, and now AI, with 1,000 plus bills ⁓ moving across America and all sorts of different types of flavors of regulation being proposed. This is what I’ll call a, sort of from a policy perspective, a constitutional moment for technology.
Like what will our new innovation policy be for the next 30 years? Congress and the Clinton administration gave us one between 96 and 98 that I think worked marvelously. But we’re in the process of now upending it and going with totally different route. And that’s why we have the discussion about preemption of AI at the federal level happening right now.
Blair Levin (29:11)
Yeah, look, I largely agree. But I guess we should state the most obvious point. There was preemption in 96 because Congress wrote it. And they wrote the law and said, here’s the way we’re going to do this. Now what Congress seems to be saying is, we’re not going to say anything, but we’re not going to let you say anything either. And I don’t think that’s really the thing.
That reminds me of that old scene from Duck Soup with Groucho Marks. He’s running the parliament meeting and some guy raises something. He says, you can’t bring that up. talking about old business and that’s new business. The guy says, okay, all right. Any other old business? None okay. Any new business? The guy raises his hand again and Groucho Marx says, no, no, you raised that before. It’s now old business.
You know, or by the way, it’s also like what the FCC did under a [G pie] to say we have no authority to regulate broadband, except the authority to tell states that they can’t. This was all in the context of net neutrality And the court said, no, you can’t do that.
Adam Thierer (30:13)
This is where there’s a serious policy disagreement between Blair and myself, because I Congress can absolutely do that.
Blair Levin (30:19)
Well, Congress can, but the FCC can’t.
Adam Thierer (30:20)
But Congress can declare its intent to essentially keep the field clean and essentially preemptively deregulate and create a policy firewall, you know, between old rules and regulations and new ones. And, you know, I agree with Blair, Congress needs to act on AI policy and they can learn some things from the Telecom Act era. But I will agree also with Blair that the policies, the politicians have grown more hardened. The issues are more R and D than they ever have been in my lifetime in terms of how tech policy plays out. I don’t think there’s a single Democrat in Congress today that would probably vote for any sort of preemption on AI policy because they like what’s happening at the state level.
Blair Levin (31:00)
You know, one thing you learn at the FCC is it’s all about the kids because the most powerful argument is always that if the kids have access to this, it’s bad for them. But there’s no doubt that there’s a lot of stuff that can happen on AI that is not good for kids.
Earlier, I mentioned that one of the major differences between now and when I was at the FCC in 96, but also when I did the national broadband plan in 2009 and 10, is there’s a real negative view, which really about technology that I think started in serious form in about 2012. And Jonathan Haidt, the professor at NYU has done a lot of work on this. And you see this with states on a very bipartisan basis saying, we need to get cell phones out of schools. That kind of thing. Well, there is an understanding that there is a negative side to technology.
Of course, you always understand there’s always trade-offs. Nothing is perfect. But I don’t think congressional leadership has risen to the occasion of trying to address this in any kind of way on a bipartisan basis. It’s a lot harder for the reasons Adam mentioned earlier. The partisanship is much tougher. But at the end of the day, there’s something weird going on in Washington where Congress is just, other than passing one piece of legislation, what are they doing about anything?
Adam Thierer (32:23)
Well, there’s a lot to unpack there. I’ll just say this, like, look, Congress is sort of fundamentally dysfunctional on multiple levels right now. There’s a lot more going on here than just technology policy dynamics. It’s just really, really hard to get a lot of things through Congress these days and bickering, partisanship is part of it. But also these issues have become more complicated than ever and more numerous than ever.
One of the interesting things about the old days is I could sit around as an analyst at the Heritage Foundation where I spent 10 years covering these issues in the 90s, and I could really cover almost everything myself. I really had a really good feel for what was going on in the world of telecom law and policy. I absolutely could not do that anymore. I absolutely have to get more siloed and focused. And members of Congress and their staff, their very young staff, are confronted with a knowledge deficit about the unbiased.
[You have an] unbelievable explosion of new technologies, sectors, opportunities, and potential threats. And they’re trying to figure out how to handle it all preemptively. Like, how do we get this all handled right now? And I disagree a little bit that, Congress, is not caring about this. I mean, there’s a hearing in the Energy and Commerce Committee where 19 different child safety bills were being considered for AI policy and internet policy.
A lot of it is just how you get those things over the finish line and how you also overcome constitutional objections and other barriers to these things. But there’s a lot of interest in doing things. It’s just hard to get them done.
I think the other problem here is just the way that the legislative clock takes seemingly faster and faster makes it so that you have to come up with quick fixes and gimmicks compared to previous sessions. I mean, we used to work on things over a couple of sessions of Congress and then they finally come to fruition.
Now it just seems like we move our way to get to something like on driverless cars or privacy or whatever, and then it all just peters out, implodes, and then the next issue hits you in the face. So the quick ticking clock and the fact that the issue set has exploded and become so much more complicated, weighs against the fact that Congress is gonna be able to be a major actor on technology policy as they were in the past, leaving it to the sporadic activities of a lot of different states and localities and international governments to sort of fill in gaps here and there, but not in a consistent coherent fashion, which is the patchwork problem that we’re facing today.
Matt Perault (34:39)
So we’re talking a lot about the Telecom Act and all the things that worked well. Adam, I know that you pointed to some of the things that were more challenging about it. What are, can we go into a little more detail? Like what are the kind of mistakes that were made or what are the things that if you could go back and do them again, you would do them differently?
Blair Levin (34:57)
I’m going to be like Donald Trump and say that everything we did was perfect.
It’s a much longer conversation. I would say for the most part, we got a lot done. We got it mostly right. I would say that ironically, the things that really worked well, the core bargain was that long distance local thing. But at the end of the day, what mattered was the cable industry developed its own broadband. We essentially destroyed the voice market. What are you paying for long distance these days? Nothing.
Because the internet replaced everything and by the way Al Gore and Reid Hunt understood that in an analog world you could have these silos of voice video and data but in a digital world they’d all come everything just becomes data what we’re doing right now. I mean I went to the 1964 World’s Fair in New York where it was a video phone that was like three bucks a minute.
And of course, no one adopted it because like three bucks was really a lot of money in those days. But we’re doing this essentially for free.
There are so many things that actually worked out well, but not necessarily because of the way we did them. But it was things like the program access rules that forced cable in. It was like getting rid of the terminating access charges for wireless. So it could grow. It was spectrum policy. There were a lot of different things.
One of the advantages that we had was we had an agency, the FCC, that had expertise. And one of the most important things about expertise is to understand when you’re wrong and course correct. Because the truth is you’re never gonna get it 100% right. And there are gonna be these court proceedings and other things. So part of the problem we have now is, and by the way, we spent a lot of time in the background, talking to members of Congress and saying, if you write it this way, we have this problem. If you write it that way, it’s a little bit better. Who’s the expert agency on AI? Where is there a locus of people who aren’t looking at this from a partisan basis, but just want to see America succeed with AI? It doesn’t really exist in the government. Maybe it’s a little bit at the FTC, but I don’t think so. And that’s what I think we have an institutional capacity problem.
Adam Thierer (37:16)
So let me briefly touch upon where I think the telecom succeeded, because we had a hundred years worth of really misguided policies and specifically a lot of misguided sort of geographical constraints and regional monopolies and everything.
When we passed the Telecom Act, it was a new day and we basically started shifting more the focus to this being a national concern and a national marketplace. And we got that right. And I talk a lot in my work in my books about innovation culture and about like Congress and the White House sending a clear signal on certain key technologies and sectors for the American economy and for American strength. And we did that finally. We said, no, this is a national matter. We need that for AI too. We need that kind of a vision.
And then secondly, I think I’ll just go back to the fact that there were important sectors that we essentially created a policy firewall for, not perfectly.
But we basically said that certain things in the world of wireless and the world of hardware and equipment, computation, computing, software, we’re not completely off limits, but we’re going to take a lighter touch. It was a new day. We were going to allow those technologies to have a new chance and break away from the analog era mindset and give them a different ability to prosper. I think that was crucial. It wasn’t perfectly articulated. There isn’t any one moment in the telecom act where we got this exactly right. And of course, there are other things we haven’t discussed like the Internet Tax Freedom Act and some other things. There was a bill that was pending in 1997-98 called the Internet Freedom Act, the IFA. Rick White and I forget who else, Billy Tauzin had a bill that basically said what I just said. And then that became what the White House put together in the framework for global electronic commerce. I wish Congress would have better articulated all.
So it sort of became this accidental fortuitous development of having a national framework, but it clearly got started with the telecom. Congress absolutely has to do that again. We can debate the details, but we just can’t say, off the wheel, anything goes, patchwork of a thousand plus policies on every single issue under the computational algorithmic sun. That is not gonna work for the United States or its strength and its ability to prosper.
Matt Perault (39:40)
So maybe this can be closing portion of the conversation. Adam, given what you just said, what do you think would be, what would be the pillars of what you have in mind? So like if we were thinking of a telecom act essentially for AI, what would the elements of it be?
Adam Thierer (39:55)
Well, clear assertion of national authority over certain matters that are clearly interstate algorithmic commerce. We can debate what that phrase means, and I know it’s controversial, but the reality is Congress has to take more authority over this marketplace. Second of all, I think we need an ongoing standing process in our government to actually try to deal with federal-state relations. This is a hard issue. Blair points out there’s no institutional capacity.
I would say actually we do have a lot of agencies. I’ve advocated some of them go away, but they never do. And the reality is, is not only those, but we’ve added to it. Joe Biden added an AI Safety Institute, which is now the Center for AI Innovation and Standards.
Let’s have a standing body that tries to help figure out standards for federal state matters. If we’re going to allow the states to be a player on things like AI chat bot rules and algorithmic discrimination and all these other things that I personally prefer comprehensively preempted, we at least need to have consistency in that process. We have to have some sort of a federal overlay that says, here’s how we make it simpler so that Little Tech and other types of players can thrive and that consumers in the nation can benefit. And so I think that’s really, really crucial.
A third component we haven’t discussed here is the federal government has to play a role because there are also speech related matters in play here. This is about making sure that this is a technology of freedom and a technology that can benefit the public by having a lot of different voices heard.
I think that’s something that’s often overlooked. We kind of got that wrong in the Telecom Act, but look, the court smacked it down. But I would like to see that made a priority as well.
Matt Perault (41:26)
Blair, what about from your perspective?
Blair Levin (41:27)
Well, from my perspective, Adam has done a lot more work on this recently than I have, but number one, I would say as a political matter, the first thing is to protect the children. I kind of make light of that in some way, but in some way I’m quite serious about that. Because I think that what we’ve seen from, and again, Jonathan Haidt has documented this. There are people who argue with him and all that. But certainly if you poll parents and kids, there’s a lot of negative that’s coming out of the way the internet is being used. And so there’s a fear that that negative sensibility, causes of depression and other things, we’ve got to do something about that. We’ve got to protect it. I agree with that. And it’s better to do it on a national level than a state by state level. But when you look at what the state laws are doing, a lot of them go to that. So I think that’s actually the top priority.
Then I think there’s a certain kind of infrastructure need. It could be that capital markets, they seem to be funding all of this. That doesn’t seem to be a problem. But there are various problems, particularly with the power, the grid, and all that for the data centers. And you see this in the race in New Jersey, where it’s really a concern to consumers, because they don’t see they’re getting the benefit, but they’re facing higher electrical prices and they think it’s because of AI. And so you’ve got to deal. There are some issues related to that.
But the third thing, and this reflects that I was chief of staff when we were doing this, it’s about institutional capacity. I’ll make the following deal with Adam. Let’s get rid of the FCC, which I think has outlived its usefulness. That’s a whole other question. Move spectrum over to NTIA, but create a digital regulator. I know Adam would not want to give that power to a federal authority. But I think that the way to do it is to have a federal authority that actually does have the power of regulation, uses it wisely, sometimes they’re going to and sometimes they’re not going to. But I think you have to, but I’d rather have that it’s very difficult for Congress to write something that’s going to last more than a year. And that’s why you have an FTC or an FCC or an SEC so that would be mine, but I suspect that Adam would disagree with that.
Matt Perault (43:49)
Adam, what do think?
Adam Thierer (43:49)
Yeah, I think we have a lot of institutional capacity already we need to optimize for. And if we’re going to get rid of a couple of agencies.
Blair Levin (43:56)
It’s out all over the place and it doesn’t and it also hasn’t been given a mandate from Congress, right? I mean, Biden created this thing.
Adam Thierer (44:04)
Yeah. And there are many bills in Congress that actually would cede new authority to the Department of Commerce and the Center for AIA Standards. And we could have that discussion. I’m open to that. I have written things for the R Street Institute that basically say, here’s how you can actually have some federal regulation of some of the concerns that Blair’s outlined and others we haven’t discussed in terms of AI safety, transparency, and other things like that.
We have a lot of regulatory capacity at the federal level, 411 agencies at the federal level. But if you want to add one more, I guess go for it. You know, the reality is, yeah, I just think that, you know, bureaucracy is not always the solution and we already have a lot of capacity. So let’s optimize for it.
Blair Levin (44:33
Well, I’ll add one and subtract one.
Bureaucracy isn’t the answer, but capacity is. And I just think that’s a big issue here. Again, I could be wrong about this. You’re involved more with this than I am. But I don’t see who is helping members of Congress understand what the real problems are.
There’s a conservative analyst Yuval Levin, who unfortunately is not related to me because I really like his work, but has talked about how Congress has become formative, performative rather than formative. That is to say, because of the cameras on the hearings, everyone’s just, you know, performing for TikTok. And I testified a couple of times last year and it was so strikingly different than when I had testified before in terms of nobody was really asking a question to try to learn anything. They were there to make speeches and that’s just a big problem.
Matt Perault (45:34)
Blair, Adam, thanks so much for this great conversation.
Blair Levin (45:37)
Thank you and good luck.
This newsletter is provided for informational purposes only, and should not be relied upon as legal, business, investment, or tax advice. Furthermore, this content is not investment advice, nor is it intended for use by any investors or prospective investors in any a16z funds. This newsletter may link to other websites or contain other information obtained from third-party sources - a16z has not independently verified nor makes any representations about the current or enduring accuracy of such information. If this content includes third-party advertisements, a16z has not reviewed such advertisements and does not endorse any advertising content or related companies contained therein. Any investments or portfolio companies mentioned, referred to, or described are not representative of all investments in vehicles managed by a16z; visit https://a16z.com/investment-list/ for a full list of investments. Other important information can be found at a16z.com/disclosures. You’re receiving this newsletter since you opted in earlier; if you would like to opt out of future newsletters you may unsubscribe immediately.
