Ep 18: Weed Stopped the US Open: Cannabis Etiquette After Legalization — The Shift

Weed smoke stopped the US Open. A new study flipped the munchies stereotype. And AI is coming for cannabis jobs.

In this episode

• Cannabis interrupted a US Open match when world number one Aryna Sabalenka stopped play to report someone smoking weed, with the smell likely drifting in from legal recreational use in Flushing Meadows Corona Park. The operator point: the legalization question (should this be legal) and the etiquette question (where is it socially acceptable) are different conversations, and the industry has spent far more time on the first.
• The comparison set is tobacco and alcohol. Both stayed legal while society built boundaries around context: no cigarettes on airplanes, no vodka in the Monday meeting. Cannabis carries one extra complication those products mostly lack, which is smell, so one person's private choice becomes everyone else's environment much faster.
• Consumption method is the lever. Edibles and beverages produce no smoke, so the next stage of cannabis etiquette is less about the plant and more about which format fits which shared space. Mature normalization looks like cannabis getting the social rules alcohol and tobacco already have, not an assumption that anything goes.
• A University of Sydney analysis of 5,383 people with eating disorders or serious disordered eating found 56% used cannabis in the past year and 42% of those used daily. People with anorexia and ARFID often reported cannabis eased food anxiety, nausea, and the ability to eat, while people with binge eating disorder frequently reported the same appetite effect made symptoms worse. Roughly one in three users expressed some concern about their own cannabis use, and the authors call for clinical trials rather than treating this as proof.
• Michael Sassano, founder of SOMAI Pharmaceuticals, published an analysis on September 5 arguing medical cannabis companies could run with far fewer employees as AI absorbs finance, compliance, inventory, QA, and reporting work. The framing to take from it: AI handles the workload while humans stay where regulation requires a named, accountable signature, which pushes the value of a cannabis job toward judgment, relationships, and accountability rather than moving data between systems.

Chapters

• 0:05 — Weed stops the US Open
• 0:24 — Sabalenka smells smoke
• 0:42 — Nobody knows the source
• 1:20 — Legal, but is it acceptable?
• 1:48 — The cigarette comparison
• 2:16 — Alcohol has boundaries too
• 2:35 — The right vs. the etiquette
• 2:55 — Smell changes everything
• 3:25 — Two ideas that coexist
• 3:52 — Normalization has limits
• 4:05 — Edibles and drinks don't smoke
• 4:32 — When your choice is my air
• 4:50 — The next stage of etiquette
• 5:10 — Rules in 10 years?
• 5:34 — The munchies, studied seriously
• 6:00 — The University of Sydney study
• 6:25 — 5,383 people, 56% used
• 6:46 — Effect differs by disorder
• 7:04 — Binge eating: the effect flips
• 7:22 — Helpful or harmful, or for whom?
• 8:01 — Cannabis has no single identity
• 8:40 — One in three worry about use
• 9:17 — What a survey can't prove
• 9:37 — Why clinical trials come next
• 9:57 — Which cannabinoid, which patient
• 10:53 — Medicine vs. stereotype
• 11:11 — The patient decides the value
• 11:40 — AI is coming for cannabis jobs
• 11:49 — Sassano's prediction
• 12:07 — Not a study, a forecast
• 12:27 — The back-office of cannabis
• 13:26 — AI works, humans are liable
• 14:04 — AI already in ERP systems
• 14:23 — Which jobs need a human?
• 14:43 — Cultivation: collect vs. decide
• 15:22 — Compliance and paperwork
• 16:00 — Who is liable for AI errors?
• 16:18 — The dispensary in five years
• 16:36 — Information isn't everything
• 17:16 — What automation reveals
• 17:54 — Judgment, trust, relationships
• 18:34 — Cannabis as a test case
• 19:14 — What did we need people for?

Transcript

Auto-generated from the episode audio and lightly edited for readability.

0:05 [music] Let's start today with something that happened last week at the US Open. Because cannabis somehow managed to interrupt one of the biggest tennis tournaments in the world. Without actually being anywhere near the tennis ball.

0:24 World number one Aryna Sabalenka was playing her third round match when she suddenly stopped, walked over to the umpire and said, "Someone is smoking weed." Apparently, the smell was that strong.

0:42 Now, nobody actually knows exactly where it was coming from. Smoking isn't allowed inside the US Open grounds, but recreational cannabis is legal in New York and smoke from the surrounding Flushing Meadows, Corona Park can drift into the tennis complex.

1:02 Sabalenka herself said she didn't know whether it was coming from the crowd or whether the wind was carrying it in from outside. And to be clear, she wasn't particularly outraged about people using cannabis. Her position was basically do whatever you want, just please don't do it next

1:20 to the tennis court while I'm trying to compete. Fair enough, but I actually think this little moment at the US Open represents a much bigger conversation we're starting to have around cannabis. But as for decades, the cannabis debate was mostly about one question, should this be legal?

1:39 But once something becomes legal, a completely different conversation begins. Where is it socially acceptable? And those two questions are not the same thing. Cigarettes are probably the easiest comparison. Smoking cigarettes is legal, but that does not mean you can

1:57 light one wherever you want. Over time, society developed boundaries, not on whether you're allowed to smoke, but on where your decision to smoke begins affecting everybody around you. Restaurants changed, offices changed, airplanes changed,

2:16 stadiums changed, and eventually those boundaries became completely normal. Alcohol has boundaries, too. Drinking a glass of wine with dinner is completely ordinary. Walking into your Monday morning meeting with a bottle of vodka, slightly different situation. The product didn't

2:36 suddenly become illegal, the context changed, and I think cannabis is now going through its own version of that transition. We spent so much time discussing the right to consume cannabis, that we've spent considerably less time discussing the etiquette of consuming cannabis. Because cannabis has

2:57 one very obvious characteristic that makes this particularly complicated. Smell. If somebody drinks a beer 20 m away from me, I probably have absolutely no idea. If somebody smokes cannabis 20 m away from me, there's a pretty good chance I know

3:15 exactly what's happening. Which means one person's personal choice can very quickly become part of everybody else's environment. And that's where this stops being exclusively a cannabis rights conversation. Because you can completely support somebody's right to consume cannabis,

3:34 while simultaneously believing that you shouldn't have to smell cannabis during dinner, at work, at a sporting event, or anywhere else where smoking isn't appropriate. Those ideas don't contradict each other. In fact, I'd argue that being able to

3:52 hold both ideas at the same time is probably a sign that cannabis normalization is actually maturing. Because normalization doesn't mean anything goes. Normalization means cannabis eventually gets something that alcohol and tobacco already have. Social rules. And some of those rules

4:12 might happen on how cannabis is consumed. An edible doesn't produce smoke. A cannabis beverage doesn't produce smoke. Smoking obviously does. So maybe the next stage of cannabis etiquette isn't even primarily about cannabis. Maybe it's about consumption methods. Shared spaces and understanding when

4:32 your personal choice starts becoming somebody else's experience. And that's why I found the US Open incident so interesting. Sabalenka wasn't giving us some grand political statement about legalization. She was basically saying, "Guys, I'm trying to play tennis here." And strangely enough,

4:52 that might tell us more about where cannabis culture is heading than another legalization debate ever could. Because the first stage of normalization was asking society to accept that adults consume cannabis. The next stage might be cannabis consumers accepting that

5:10 legalization comes with etiquette, too. So here's my question. 10 years from now, what will the unwritten rules of cannabis actually look like? Where will smoking be completely normal? Where will it be considered rude? And what will future generations find it

5:29 strange that we ever had to debate that this at all? Because the law can tell us where cannabis is legal. Culture still has to figure out where cannabis belongs. We all know the stereotype. Cannabis gives you the munchies. And it's probably one of the

5:47 oldest cannabis jokes there is. But a new study published last week asked a much more serious question. What happens when changing someone's appetite isn't a side effect? What happens when that's actually the point? Researchers from the University of

6:06 Sydney looked at cannabis use among people living with eating disorders. And this wasn't a tiny survey. The analysis included 5,383 people with diagnosed eating disorders or disordered eating serious enough to cause distress.

6:26 And the first thing that surprised me was just how common cannabis use was. More than half, 56% reported using cannabis within the previous year. And among those cannabis users, 42% were using it daily.

6:46 But here's where this study becomes really interesting. Because cannabis didn't appear to have the same perceived effect across different eating disorders, people with anorexia nervosa and people with ARFID, which stands for avoidant restrictive food intake disorder, frequently

7:06 reported that cannabis helped with things like food anxiety, nausea, and their ability to actually eat. But then, researchers looked at people with binge eating disorder and suddenly the picture changed. In that group, cannabis was not

7:24 perceived as having the same benefit. Some participants actually described it as making their symptoms worse, including triggering overeating. And think about how interesting that is for a second because we're potentially talking about the same characteristic of

7:42 cannabis, increased appetite, increased interest in food. A changed experience around eating. For somebody who is struggling to eat, that could potentially be useful. For somebody for somebody struggling to stop eating, the exact same effect could

8:01 potentially become part of their problem. So, is cannabis helpful or harmful? Maybe that's the wrong question. Maybe the better question is, helpful for who? Because I think we're slowly moving away from this idea that cannabis has one universal medical identity. Cannabis

8:21 helps pain. Cannabis helps sleep. Cannabis helps appetite. Those statements sound nice, but human beings are considerably more complicated than that. Even within something we casually call an eating disorder, we're talking about completely different conditions,

8:40 completely different behaviors, and potentially completely different responses. And there's another side of this study that we absolutely have to mention. Among the cannabis users surveyed, 18.4% said they believed they had a problem with cannabis. Another 14% say they

9:01 thought they might have a problem. That's a roughly one in three cannabis users in the survey expressing at least some concern about their own relationship with cannabis. So, again, this isn't a simply you know, like a success story. The

9:20 researchers themselves are not saying we discovered cannabis is a treatment for anorexia. This was a survey based on people's own reported experiences. It can tell us what people believe in cannabis is doing for them. It cannot prove that cannabis actually

9:38 treats their condition, and that's an incredibly important distinction. What the researchers are saying though is that these results are interesting enough to justify proper clinical trials. Well, specifically warning that dependence and other adverse effects also need attention.

9:57 And personally, I think that is where this story becomes much bigger than eating disorders because maybe the future of medical cannabis becomes much more specific. Not does cannabis help appetite, but which cannabinoid for which patient

10:16 with which condition and what concentration? And producing which effects exactly? Because medicine doesn't usually work by saying, "Well, this seems to make people hungry, so let's give it to everybody who has a

10:34 problem involved living involving food." That's not medicine. That's a stereotype. Medicine is understanding when an effect is useful, when it's irrelevant, and when that exact same effect could actually be harmful. And cannabis may increasingly have to

10:53 meet that standard, too. Which brings me sort of back to the munchies thingy, something cannabis culture has joked about for decades might actually demonstrate one of the most complicated things about cannabinoid medicine.

11:11 The effect itself isn't necessarily good, and it isn't necessarily bad. The patient determines whether that effect is useful, and maybe that's the question we should start asking much more often with cannabis. Not does it work?

11:29 But for whom? Let's finish today with a prediction about the cannabis industry. And I'm genuinely not sure whether it's exciting, terrifying, or probably a little bit of both. So, artificial intelligence is coming for cannabis jobs.

11:49 At least that's the argument being made by one medical cannabis executive this week, Michael Sassano, founder of SOMAI Pharmaceuticals, and I hope I'm pronouncing that right. So, they published an analysis September 5th arguing that medical cannabis companies could eventually operate with

12:09 dramatically fewer employees because artificial intelligence can take over large portions of the administrative work happening inside these businesses. And I want to be clear here. This is not a study predicting exactly how many cannabis employees will lose their jobs.

12:27 It's an industry executive making a prediction about where he believes the business is heading. But the reason I think it's worth discussing is because when you actually look at how cannabis companies operate, his argument isn't as crazy as it

12:47 initially sounds. So, think about what happens behind the scenes of a regulated cannabis business. Finance, accounting, inventory, compliance, quality assurance, regulatory paperwork, sales administration, customer service, human resources, reporting,

13:07 data analysis, a huge amount of the industry isn't somebody standing inside a cultivation facility touching a cannabis plant. It's somebody sitting behind a computer moving information from one system to another, and that is exactly the kind of work artificial intelligence is getting

13:26 increasingly good at. Sassano's argument is essentially that AI takes over the workload. While humans remain where regulation requires accountability. So the computer might analyze the information. It might prepare the

13:44 documentation. It might identify the problem. It might even recommend the next action, but somewhere in that process a trained human may still need to review it. Approve it and legally put their name behind it. And we're already seeing pieces of this sort of future appear.

14:04 Cannabis technology companies are building AI into ERP systems. Order processing, compliance, cultivation monitoring, and business analytics. There are systems that can monitor temperature, humidity, CO2, um, and plant health even. Looking for problems before human cultivator might notice

14:23 them. There are tools processing orders that previously required somebody to manually enter information into an ERP. I mean And this is where I think the conversation becomes much more interesting than AI is going to steal everyone's job. Because I think the better question is

14:43 which cannabis jobs actually require human? Let's take cultivation. Artificial intelligence can analyze thousands of data points inside a grow facility. It can detect patterns. It can monitor environmental conditions 24 hours a day.

15:02 It doesn't get tired. It It doesn't forget to check something. But does that mean you no longer need an experienced cultivator? Probably not. Maybe it means the cultivator stops spending their time collecting information [snorts] and spends more time making decisions

15:22 with it. Now, let's see compliance. Cannabis companies generate enormous amounts of paperwork because they're operating in one of the most heavily regulated consumer industries in the world. AI is basically designed for exactly

15:41 that environment. Read this, compare that, find the discrepancy, flag the problem, generate the report, but then comes the question of accountability. If an AI system makes a compliance mistake, who's responsible? The software company? The cannabis operator?

16:00 The employee who approved it? The executive who decided nobody needed to review it anymore? AI can make a decision incredibly quickly. Regulation still wants to know which human is responsible for that decision. And then we get to the part I

16:18 find most interesting, the customer itself. Imagine walking into a medical cannabis dispensary 5 years from now. Instead of immediately speaking with somebody, an AI system knows the products available, their cannabinoid profiles, your preferences,

16:36 your previous purchases, and potentially enormous amounts of published cannabis research. From a pure information perspective, that system could theoretically know more than any individual employee ever could. But is information the only thing the customer needs? What about judgment?

16:56 What about trust? What about somebody noticing that you are confused, nervous, or that you're asking the wrong question? What about somebody actually saying, "I don't think that's what you're looking for." That's much harder to put into a spreadsheet, and maybe that is the real

17:16 test AI is about to give this industry. Because whenever technology automates a job, it forces us to discover which part of the of that particular job was the value of the employee entering information into five different systems. Probably not. Was it remembering which

17:36 form needs to be submitted every Tuesday? Probably not. Was it answering the same basic question 50 times every day? Maybe not. But judgment, relationships, accountability, creativity,

17:54 understanding another human being, those become much more interesting. And this isn't unique to cannabis. Every industry is having some version of this conversation right now. But cannabis is an especially interesting experiment because it's simultaneously agricultural, medical, pharmaceutical,

18:14 retail, and incredibly regulated. There are physical jobs AI can't simply perform from a laptop. There are regulatory jobs where humans remain legally accountable. And there are administrative jobs where frankly, we may discover that we were

18:34 employing people because technology wasn't good enough yet. So, I don't think the interesting prediction is AI will replace cannabis workers. That's too easy. The much more interesting question is what happens after we automate everything that can

18:55 reasonably be automated? What's left? Which human skills suddenly become more valuable? Precisely because a machine can do everything else. Because maybe AI isn't just about discovering how few people it takes to

19:13 operate a cannabis company. Maybe it's about finally discovering what we actually needed the people for. [music]