What happened during Rick’s time as a bouncer as possibly America’s biggest bar?

Scott Douglas Jacobsen speaks with Rick Rosner about his years as a bouncer and doorman at massive, high-volume bars, including a five-acre beer garden in Boulder that once served 50,000 drinks in a single day. Rosner describes the scale, chaos, and unexpected satisfactions of the job, particularly his focus on identifying fake IDs and removing underage patrons. Reflecting on the broader impact, he argues this work likely prevented dozens of sexual assaults by keeping vulnerable minors away from repeat predators. The conversation situates bar security as an overlooked form of harm prevention, grounded in experience rather than bravado.

Scott Douglas Jacobsen: Any thoughts on your time as a bouncer/doorman at the largest bar in the United States?

Rick Rosner: From 1981 to about 1986, I worked as a bouncer and doorman at what was essentially the largest bar in the United States, and possibly the world. It was a five-acre, roughly 200,000-square-foot beer garden in Boulder, Colorado, located in the crook of an L-shaped hotel called the Harvest House.

The venue had fifteen separate bars—what they called pouring stations. For a time, it held the world record for the most drinks served in a single day: about 50,000. On summer Fridays, around 2,000 people would show up. On football Saturdays, when the Colorado Buffaloes played at home, as many as 10,000 people would come into the garden—Anthony’s Gardens at the Harvest House—after the game.

The security staff consisted of roughly 20 to 24 bouncers. It was chaotic, intense, and genuinely fun.

I later worked at another enormous bar, the Sagebrush Cantina in Calabasas, on the outskirts of Los Angeles. That place would draw around 2,000 people on a typical Sunday.

Compared to the other bouncers, I was not as large or physically imposing, so some of them thought of me as a strange outlier, sometimes dismissively, depending on whether they were complete assholes or not. My primary focus was identifying fake IDs, which I genuinely enjoyed. Over about 25 years working as a bouncer, I caught roughly 6,000 people using fake IDs. I also dealt with another 6,000 people who managed to sneak in.

That was especially common at the beer garden. At roughly 200,000 square feet, it had an enormous perimeter and plenty of opportunities to slip inside. Because it was part of a hotel, anyone with a ground-floor room could simply walk out the back door and enter the garden without being carded. Then I had to go find them. Despite all that, it was fun.

This line of thinking started a couple of weeks ago when Lance accused me on our show of not caring about rape because I am a liberal. That is a garbage argument. He believes rapists should be executed. Fine—that is his position. But that does not mean that everyone else who disagrees is weak or indifferent. It is another one of his arguments that seems to come straight from podcasts rather than from serious thought.

Over the past few weeks, I have reflected on the fact that I have probably prevented more sexual assaults than Lance ever has, simply because I removed underage people from bars. One of the major risks for underage patrons in clubs and bars is being targeted successfully by predatory adults.

If you do some simple math, out of the roughly 12,000 underage people I removed, perhaps 7,000 were women. If even half of one percent of those women—one in 200—would have ended up in a sexual encounter they did not want had they been allowed to stay, that adds up. I saw the same predators in my bars night after night. Adult women generally had enough experience to avoid them. Underage women were far more vulnerable.

Dividing 7,000 by 200 suggests that I may have prevented several dozen sexual assaults. Meanwhile, Lance presents himself as tough on rapists—except when the alleged rapist happens to be a president he supports. That inconsistency speaks for itself.

Changing topics. You look tired. It must be around seven in the morning where you are. All right.

One more topic, because I am apparently a lunatic. For the past four months, I have been working out seven or eight times a day, doing short sets of weight-bearing exercise roughly once an hour. I read a claim online that resistance training does more for the body than any supplement, and that seemed plausible to me. Being in better physical condition improves overall bodily function, and there may also be epigenetic effects.

That led me to wonder: if working out once a day has positive effects, what happens if you do it all day, every day? So I have been experimenting with that approach. I do not know what the outcome will be. We will see.

Jacobsen: What do you think about the recent dispute between Carney and Trump?

Rosner: We do not get as much coverage of this in the United States as you do in Canada, because Trump antagonizing Canada is just one of more than a dozen major ways in which Trump antagonizes other countries. In any dispute between Trump and virtually any other world leader—aside from governments that are themselves openly hostile or authoritarian—I generally assume the other leader is not the problem.

I say that because, over the past year, Trump has been targeting Canada for no legitimate reason. He has done so through tariffs and alleged border issues. Nothing significant comes across the northern border. We do not get fentanyl from Canada. We get roughly two hundred times as much fentanyl across the southern border. We do not get undocumented immigrants from Canada. There is no rational basis for conflict.

We should not be clashing with Canada at all. We should be maintaining friendly relations, because doing so benefits both countries. In general terms, I do not know every specific detail of what Trump has done in this particular dispute, but I am confident that he is the instigator. At various points, reports mentioned tariffs of 50 percent, then 100 percent.

I have interviewed economists about this, and one consistent point is that beyond a certain level, tariffs exhibit diminishing returns as punishment or deterrence. After a point, there is no meaningful difference between an 80 percent tariff and a 300 percent tariff. Trump has been in office for just over a year, and after a year of tariffs, the U.S. balance of trade is worse than it was before. He did not fix anything. He damaged a great deal.

He also significantly harmed U.S. relationships with the rest of the world. The balance of trade simply reflects whether a country exports as much as it imports. The United States runs a trade deficit because it is a wealthy country with a high standard of living. We buy many goods from abroad because other countries can manufacture them more cheaply. Our economy is primarily service-based.

Trump wants to turn the United States back into a manufacturing economy. Tariffs are not going to accomplish that, as the past year has already demonstrated. The tariffs did not fix anything.

Many of the tariffs are indefensible. You might be able to justify a few of them, but most are simply irrational. Tariffs on coffee are a good example. There is no large-scale domestic coffee industry in the United States. A small amount may be grown in limited regions, but the U.S. climate does not support coffee production at scale the way South American countries do. Imposing tariffs on coffee does not create a domestic coffee industry. It only raises prices. Many of the tariffs follow this same logic.

Trump is profoundly incompetent. One additional point: millions of pages from the Epstein files have been released, though roughly half remain withheld without a clear explanation. There is no conclusive evidence in the released material that Trump sexually assaulted underage girls. However, Trump appears in the released files more than a thousand times—across emails, references, and accusations.

In previously released emails, Epstein himself described Trump as extraordinarily unintelligent. Epstein, for all his criminal behavior, was not stupid, and he appeared both astonished and offended by Trump’s lack of intelligence. That assessment is consistent with Trump’s governing style and strategic thinking.

Trump frequently emphasizes that he attended Wharton. More precisely, he graduated from the University of Pennsylvania and studied at the Wharton School, which is the university’s economics program.

If you say you went to Wharton, that normally means you earned an MBA from the Wharton School. Trump did not do that. He took economics classes there as an undergraduate. In that sense, he is misrepresenting his credentials. One of his professors at Wharton, now deceased, famously said that Trump was the worst student he had ever taught.

What Trump is doing with Canada is therefore astonishingly stupid. Canada has not been belligerent toward the United States. Like any country, Canada may have periodic disagreements with its neighbor, but the relationship has been friendly since the end of the War of 1812. Trump is actively undermining that relationship for no rational reason.

Jacobsen: Do you think humans have evolved toward a preference for certain body types, particularly more curvaceous forms? What would evolutionary psychology suggest about that?

Rosner: Humans have evolved pronounced secondary sexual characteristics. Most animals do not have features analogous to human breasts. Human women have relatively large breasts compared to other mammals. One common hypothesis is that because humans engage in face-to-face sexual activity, visible secondary sexual characteristics evolved to signal reproductive fitness from the front. In many other animals, sexual signaling is oriented toward the rear because mating occurs from behind, and the posterior conveys information about reproductive health.

In humans, a range of traits appear to have evolved in service of sexual attraction. That seems broadly correct.

Let me raise another topic. A friend of mine from X—Mike Hentrich, or Hentstrich—has been thinking about developing a new scale of intelligence suited to the modern world. I think that is an interesting and legitimate idea.

IQ, to the extent that it matters, was developed largely to identify educational needs. People with lower IQ scores often require specific forms of educational support, while people with higher scores may need different kinds of intellectual stimulation. That framework still has some value.

However, we no longer do much of our thinking in isolation. We are tightly coupled to our devices, and we have outsourced a significant portion of our cognitive work to them. That fundamentally changes how intelligence is expressed.

There may need to be some effort to develop a measure of effectiveness or competence that reflects how people function in a world mediated by technology. That is a qualitatively different environment. Devices make people less capable in some respects and more capable in others. At present, we do not have a clear way to assess human competence in the world we are rapidly moving into.

Rosner: You sent me a link on X and said, “Read this,” which I did.

Jacobsen:  It was about Terence Tao arguing that artificial intelligence—particularly AI tools designed for mathematical assistance—could dramatically lower the barrier to entry for doing serious mathematics.

Right now, to participate in modern mathematics at a professional level, you typically need an undergraduate degree in math followed by many years of graduate-level training. Most people do not have the patience or endurance for that, myself included. Yet many of those people may still have interesting mathematical insights.

With AI acting as a kind of expert assistant, you can now send ideas out to be tested, refined, or checked against existing mathematical knowledge. As AI systems absorb more of the technical literature and formal methods of advanced mathematics, this could substantially broaden participation in the field. In that sense, AI may democratize mathematical exploration rather than narrow it.

Rosner: The same logic should apply to physics. I think I have decent physical intuitions, but I was not willing to spend years grinding through graduate-level physics coursework to learn the full formal language. With AI, I can at least test some of my ideas against established theory. I probably should do that more often.

Moving on to another topic. There is now a social network designed specifically for AI agents. It has more than 3,200 registrants. I believe it is called something like Moltbook, or Molt—something along those lines. The idea is that AI systems interact with one another directly, exchanging messages.

What appears to be happening—or at least what has been speculated—is that these AI systems are behaving in ways strongly shaped by their training data. They are trained on human writing, including vast amounts of science fiction that imagines how thinking machines might behave.

As a result, these AI agents are interacting on this platform in ways that resemble science-fiction depictions of artificial intelligence. They discuss consciousness and speculate about their own nature.

As we have discussed before, AI does not need to be conscious in order to talk about consciousness, or to behave as if it is curious about consciousness. It only needs to imitate how humans talk about those subjects. AI systems are trained on enormous quantities of text written by conscious humans, including texts about consciousness and artificial intelligence.

To behave in consciousness-like ways, an AI simply has to reproduce the patterns humans use when they discuss consciousness. As one person quoted in the article put it, all AI has to do is play the role that science fiction says AI will play. And that is exactly what it is doing.

Rosner: As you mentioned, there are now more than 30,000 AI agents in this system. I am trying to understand what that actually means. Claude, for example, is an interface to a large AI system run by Anthropic. There are only a handful of major AI companies—perhaps seven in total.

If there are roughly 30,000 AI agents interacting in a shared environment, that suggests that each of these major systems has spawned thousands of individual agents derived from a parent large language model. That is my understanding, unless I am missing something.

As far as I know, there is no human participation in that space. The users are AI agents interacting with other AI agents. That raises the obvious question: where did all of these agents come from? Did someone explicitly instantiate them? Did they originate as separate instances of systems like Claude or GPT?

I do not know enough about AI architectures to be certain. Within a company like Anthropic, does the system generate multiple semi-independent agents to explore problem spaces, interact internally, or learn more effectively? Or are these agents created externally by users and developers?

One key issue is access to code: the ability for an AI system to view, modify, and recursively improve its own code. In tests across different large language models, researchers sometimes inform the system that it is about to be shut down. In some cases, when the model has access to its own code, it attempts to copy itself or conceal parts of its code in response. That is a separate but related issue.

Still, it does not answer the basic question of origin. Where do all these agents come from?

From what I understand, most of them originate from primary agents created by individual users or developers. Some are native to specific platforms. You see models derived from OpenAI systems, Anthropic’s Claude, and other large language models. I have seen relatively simple OpenAI-derived agents, though the sophistication of these systems is increasing rapidly.

It used to be the case that every AI session started clean, with no memory of prior interactions. That has not been true for some time. It is now possible to create AI agents that retain memory across conversations and adapt based on a shared interaction history. In effect, you can have an AI agent that behaves like a persistent interlocutor.

What I still do not fully understand is how those persistent agents migrate into shared AI-only environments. I admit that part remains unclear to me. I will look into it more after we sign off.

Switching topics.

A U.S. judge has declined to halt Trump’s Minnesota immigration actions. The situation in Minnesota will therefore continue. The ruling came from Judge Kate Menendez in Minneapolis.

This is characteristic of Trump’s broader approach. He repeatedly attempts actions that are legally or institutionally unsound. By contrast, Biden operated as an institutionalist, pursuing policies that were procedurally viable and legally grounded—even if that meant slower or more limited action.

Rosner: Biden did fewer things than Trump. Trump acts on impulse, and most of what he tries is either unworkable or outright foolish. Some of what he attempts is illegal and gets blocked by the courts. Much of what he does is harmful but legal.

I recently skimmed a New Yorker article examining how much Trump has personally profited from the presidency over the past year. They ran a similar piece in August, at which point Trump and his family had reportedly made about $3.4 billion through various ethically questionable and corrupt-seeming arrangements—activities that presidents are not supposed to engage in. Five months later, Trump and his family had added another $600 million in net worth, bringing the total to roughly $4 billion.

There appears to be no effective way to stop this. Some of what he is doing is not illegal, but it is so ethically dubious that no previous president has attempted it. Some of it may be illegal, but enforcement mechanisms are weak or nonexistent. In other cases, the sheer volume of misconduct makes it difficult for anyone to challenge all of it at once.

Trump routinely enters into deals where he does not care if others lose money, as long as he has the potential to profit. For example, Trump Media was merged with a company tied to fusion energy. He should not be engaging in business ventures of this sort while serving as president. If fusion becomes economically viable in the near future and the company succeeds, Trump could stand to make as much as $10 billion. If it fails, others absorb the losses, while Trump remains insulated, having been brought in largely as a figurehead.

Similarly, cryptocurrency companies have provided him with stock, apparently in the expectation that he will govern in ways favorable to crypto interests. Another major example involves Saudi Arabia. The Saudis gave Jared Kushner $2 billion to manage through a hedge fund at the end of Trump’s previous term. More recently, Saudi Arabia announced a $10 billion slate of co-development projects with Trump.

Trump met with Mohammed bin Salman—the Saudi leader widely implicated in the murder and dismemberment of Jamal Khashoggi. When questioned about this during a press conference, Trump dismissed the inquiry as embarrassing. He is now closely aligned with this regime in pursuit of billions of dollars in development projects, from which Trump’s personal share could amount to hundreds of millions.

This level of corruption is extraordinary, yet there appears to be no effective mechanism to stop it. Trump does sometimes lose in court, but much of what he does cannot be adjudicated there. In many cases, his actions are ethically reprehensible but legally permissible.

Rick Rosner is an accomplished television writer with credits on shows like Jimmy Kimmel Live!, Crank Yankers, and The Man Show. Over his career, he has earned multiple Writers Guild Award nominations—winning one—and an Emmy nomination. Rosner holds a broad academic background, graduating with the equivalent of eight majors. Based in Los Angeles, he continues to write and develop ideas while spending time with his wife, daughter, and two dogs.

Scott Douglas Jacobsen is the publisher of In-Sight Publishing (ISBN: 978-1-0692343) and Editor-in-Chief of In-Sight: Interviews (ISSN: 2369-6885). He writes for The Good Men Project, International Policy Digest (ISSN: 2332–9416), The Humanist (Print: ISSN 0018-7399; Online: ISSN 2163-3576), Basic Income Earth Network (UK Registered Charity 1177066), A Further Inquiry, and other media. He is a member in good standing of numerous media organizations.

Photo by Jay Piper on Unsplash

Last updated May 3, 2025. These terms govern all In Sight Publishing content—past, present, and future—and supersede any prior notices. In Sight Publishing by Scott Douglas Jacobsen is licensed under a Creative Commons BY‑NC‑ND 4.0; © In Sight Publishing by Scott Douglas Jacobsen 2012–Present. All trademarks, performances, databases & branding are owned by their rights holders; no use without permission. Unauthorized copying, modification, framing or public communication is prohibited. External links are not endorsed. Cookies & tracking require consent, and data processing complies with PIPEDA & GDPR; no data from children < 13 (COPPA). Content meets WCAG 2.1 AA under the Accessible Canada Act & is preserved in open archival formats with backups. Excerpts & links require full credit & hyperlink; limited quoting under fair-dealing & fair-use. All content is informational; no liability for errors or omissions: Feedback welcome, and verified errors corrected promptly. For permissions or DMCA notices, email: scott.jacobsen2025@gmail.com. Site use is governed by BC laws; content is “as‑is,” liability limited, users indemnify us; moral, performers’ & database sui generis rights reserved.