INVESTIGATION | November 6, 2026 — By Lotta Goddess
America legalized a familiar plant. The marketplace delivered something many voters had never seen: cannabis concentrates approaching 90% THC, everyday use on a remarkable scale, and a growing stack of hospital and psychiatric research that refuses to fit the old “harmless weed” story.
In a nearly two-hour Mighty Pursuit interview published November 3, 2026, neuroscientist Dr. Yasmin Hurd, who directs addiction research at Mount Sinai, argues that an entire generation has been swept into a public-health experiment with powerful products that were never subjected to the trials required of medicines. Her warning is provocative, especially when she says cannabis frightens her more than cigarettes.
But that comparison requires scrutiny. Cigarette smoking and secondhand smoke are associated with more than 480,000 U.S. deaths annually. There is no credible evidence that cannabis causes more overall deaths than tobacco. Hurd’s sharper point concerns a different hazard: dependence and psychiatric illness, particularly among vulnerable adolescents and frequent users of high-THC products. Those risks are real enough to investigate without exaggerating them.
We followed five clues from the interview into primary research, added evidence that came out as recently as 2026, and tested both the industry’s reassurance and the alarmist counter-narrative. What emerges is neither a replay of Reefer Madness nor a clean bill of health.
WATCH THE INTERVIEW — THE VIDEO AND ITS EVIDENCE
Jump to chapters: The new weed (1:26) · Your brain on THC (27:25) · Addiction (43:00) · Psychosis (1:15:15) · Cannabinoids as medicine (1:42:42).
PART I — THE PRODUCT CHANGED WHILE THE NAME STAYED THE SAME
Clue 1: The old joint versus the modern concentrate
When voters in Colorado and Washington approved adult-use cannabis in 2012, the public argument was largely about criminal penalties, personal freedom, taxation and whether regulated stores could replace illicit sales. Many advocates had good reason to challenge prohibition: a 2020 ACLU analysis of 2010–2018 arrests found Black Americans were 3.64 times more likely to be arrested for marijuana possession than white Americans despite comparable use rates. The injustice was not imaginary.
But ending discriminatory prohibition and guaranteeing the safety of every new commercial product are different projects. The National Institute on Drug Abuse’s potency-monitoring data show average THC in seized cannabis rising from 3.96% in 1995 to 16.14% in 2022. This tracks samples from the illegal market, not the exact average at today’s licensed dispensaries. Legal flower can be considerably stronger; concentrated extracts can reach percentages far higher still. A 90%-THC concentrate and a 4%-THC flower are not equivalent doses, even though both are sold under the comforting umbrella of “cannabis.”
The chemistry matters. THC is primarily responsible for intoxication and can affect perception, coordination, memory and psychiatric symptoms. CBD does not cause the same high and may have different therapeutic uses. Product effects depend on dose, formulation, route, user history and physiology. Potency percentages alone do not say how much THC entered someone’s bloodstream: the mass consumed and delivery method matter too.
Is the label reliable? A 2025 Colorado study independently tested 277 products from 52 dispensaries. Only 56.7% of tested flower products fell within 15% of their labeled THC strength; by contrast, 96% of concentrates met that threshold. Most discrepancies were overstatements of potency, not hidden extra THC. That complicates a simple “every label undercounts danger” narrative while strengthening the case for independent quality checks.
More people use cannabis frequently than alcohol—but read the survey correctly
The interview cites a widely reported finding from an academic analysis of 2022 federal survey data: 17.7 million Americans used cannabis daily or near daily, versus 14.7 million who drank alcohol that frequently. The 2024 study by Jonathan Caulkins did not say more people use cannabis than alcohol overall; far more Americans drink at least occasionally. It identified a striking shift in use frequency.
A newer federal series also confirms cannabis’s lead among high-frequency users. In the 2025 National Survey on Drug Use and Health detailed tables, roughly 16.9 million people aged 12 and older reported daily or almost-daily marijuana use in the past year, compared with about 9.7 million for alcohol. These are differently constructed estimates from a different survey presentation; the figures should not be welded into a seamless time series.
High-frequency use does not automatically mean addiction. It does mean that a growing portion of exposure is continuous rather than occasional—and that matters when assessing dependence and longer-term harms.
PART II — THE BRAIN’S DEFENSE CAN BECOME THE TRAP
Clue 2: Why tolerance is not proof of safety
Hurd explains that THC activates the endocannabinoid system, a set of signaling pathways involved in memory, stress, appetite, reward and other functions. Our bodies make their own short-lived endocannabinoid messengers; they do not naturally manufacture THC. With repeated exposure, cannabinoid receptors can become less responsive. A person may need more THC to get the same subjective effect: tolerance.
This can create a deceptive moment. Someone who says “I barely get high anymore” may indeed experience less obvious intoxication while maintaining heavy exposure. Tolerance is not the same as protection from poor reaction time, impaired driving or psychiatric vulnerability. Nor is tolerance alone a diagnosis of addiction.
Hurd describes a cycle: cannabis seems to help someone sleep or calm down; repeated use produces adaptation; the person becomes restless or anxious without it; taking more cannabis makes the discomfort disappear for a time. That is the distinction between enjoying a substance and needing it to feel normal. It can occur with many drugs, not every cannabis user.
A 2026 systematic review of adolescent cannabis and brain development found associations with altered brain structure, greater addiction vulnerability and mixed cognitive findings. It also acknowledged substantial confounding: some apparent lasting effects weaken after researchers account for preexisting differences. An adult should not tell a struggling teenager that irreversible damage is guaranteed. The research supports caution, not fatalism.
Clue 3: “Wake and bake” can hide a use disorder
The CDC estimates that approximately three in ten people who use cannabis have cannabis use disorder, defined by loss of control and continued use despite problems. Risk is higher with frequent use and earlier initiation. People with cannabis use disorder are not necessarily unemployed or visibly impaired; many maintain jobs and relationships. Equally, daily use by itself does not establish a particular person’s diagnosis.
The warning signs are less cinematic than the stereotype: repeated failed attempts to stop; cravings; using more than intended; neglecting friends, obligations or interests; continuing after consequences; and withdrawal symptoms such as irritability, sleep disruption, low appetite and vivid dreams. The NIH describes these withdrawal symptoms and notes that they can arise in heavy users even without a formal use-disorder diagnosis.
Calling such behavior a moral failure misses the biology. Pretending there is no dependence misses the person who wants to stop and cannot. Both responses can keep people from seeking care.
PART III — PSYCHOSIS: THE MOST UNCOMFORTABLE EVIDENCE
Clue 4: What “five times the odds” actually means
The interview’s most alarming study comes from a 2019 multicenter case-control investigation in The Lancet Psychiatry. Researchers compared 901 people with first-episode psychosis with 1,237 controls across 11 locations in Europe and Brazil. Daily users of what the study categorized as high-potency cannabis—at least 10% THC—had approximately 4.8 times the adjusted odds of a psychotic disorder compared with people who had never used cannabis.
Odds are not absolute risk. The result does not mean half of users will develop psychosis, or that cannabis is the sole cause in every recorded case. The study was observational; causality cannot be proven by a case-control design alone. Preexisting vulnerabilities, other drug exposures and circumstances complicate interpretation. Even so, the strength of association, frequency gradient and replication across populations give this result weight.
The broader picture is unsettling. A 2025 Ontario cohort study followed nearly 13.6 million people and estimated that the population-attributable fraction of schizophrenia associated with hospital- or emergency-recorded cannabis use disorder rose from 3.7% before liberalization to 10.3% after recreational legalization. Those are model-based fractions under causal assumptions, not proof that legalization itself directly caused every new diagnosis; changing use patterns, care-seeking and recording may influence the association.
In Denmark, a 2023 study of more than six million people modeled an especially strong association among young men. It estimated that as many as 30% of schizophrenia cases in men aged 21–30 might have been prevented if cannabis use disorder had been averted, assuming its modeled causal relationship is correct. This should not be misquoted as “30% of cannabis users develop schizophrenia.”
Psychosis is not synonymous with schizophrenia. A psychotic episode can involve hallucinations, delusions or disorganized thinking, and not everyone who experiences one develops a chronic psychotic disorder. Vulnerability varies. Some people carry psychiatric risk without knowing it. Hurd urges particular caution for adolescents and people with personal or family histories of psychotic illness.
There is a genuine debate about the causal contribution of cannabis versus shared vulnerability. But the familiar retort that cannabis “only triggers people who would have become ill anyway” is not established for every case. Nor does an association prove cannabis explains all increases in psychiatric treatment. Those are precisely the questions that better longitudinal research must answer.
Beyond psychosis: anxiety, depression, memory and the young brain
Some users report that cannabis relieves anxiety immediately. That experience can be real without showing that the drug treats an anxiety disorder over months or years. A 2026 systematic review of THC-containing products for anxiety found inconsistent results, small study samples and limited safety follow-up. Blanket statements that cannabis can never benefit anyone, or that it reliably treats anxiety and depression, both exceed the evidence.
The balance is more concerning for a developing brain. Hurd’s own 2025 scholarly review argues that earlier initiation, high potency and frequent use together increase risks for disrupted development and psychiatric outcomes. But risk is not destiny. Effects differ by age, genetics, stress, dose and the presence of other conditions; competing causes remain important.
PART IV — THE OTHER EMERGENCY-ROOM STORY
The vomiting syndrome that many people still have not heard of
Repeated cannabis use is linked to cannabis hyperemesis syndrome (CHS), a condition characterized by recurrent severe nausea, abdominal pain and vomiting. It can lead to dehydration and emergency treatment. A major new CDC report published August 6, 2026 identified 199,565 emergency-department visits involving CHS in participating surveillance data from January 2023 through May 2026. That is visits, not necessarily 199,565 different patients.
Here’s the trap hidden in the headline data: reported CHS visits jumped dramatically in October 2025, just as a new, specific billing code took effect. CDC cautioned that the jump likely reflected improved recognition and coding rather than a sudden quadrupling of actual disease overnight. The condition is serious; the chart still needs to be read honestly.
Other concerns include accidental intoxication from edibles, respiratory problems from smoking, and cardiovascular effects. CDC notes that cannabis can increase heart rate and blood pressure and may be associated with stroke and vascular disease, but research has not definitively resolved the long-term causal burden. A convenient “it came from a plant” defense is scientifically meaningless; so is pretending all reported associations are already proven causes.
PART V — THE CIGARETTE COMPARISON IS BOTH USEFUL AND MISLEADING
Hurd tells the host that the psychiatric consequences of modern high-THC cannabis frighten her more than cigarettes. As a neuroscientist’s assessment of mental-health risk, that is an important perspective. As a universal ranking of total population harm, it is unsupported.
CDC attributes over 480,000 deaths annually to smoking and secondhand smoke in the United States. Tobacco’s links to lung cancer, chronic obstructive pulmonary disease, cardiovascular disease and premature death are exceptionally well established. Cannabis smoke also contains toxins and irritants, and smoked cannabis can cause cough and bronchitis; however, evidence about independent lung-cancer risk remains less conclusive after accounting for tobacco use.
Comparing drugs on a single scoreboard hides the meaningful differences: mortality, dependence, psychosis, accidents, prenatal exposure, pulmonary disease and long-term functioning are separate outcomes. The point is not that cannabis must be worse than cigarettes to deserve safeguards. Two substances can cause different serious harms, and both can be regulated intelligently.
PART VI — WHO PROFITS, WHO PAYS, AND WHO GETS TO CHOOSE?
The incentive to sell stronger products
A legal industry can have incentives that public health does not share: sell a differentiated product, retain customers, expand market share. High-THC concentrates create premium products and help attract consumers seeking intensity. But that commercial logic is not by itself documentary proof that every cannabis company deliberately manufactures addiction. Establishing company-specific wrongdoing would require internal communications, marketing records, testing data and regulator findings.
The important failure is institutional. Retail descriptions such as “good for sleep” or “relaxation” are not substitutes for controlled medical trials. Product testing rules, marketing restrictions, minimum purchase ages, licensing and surveillance differ by jurisdiction; claiming there is literally no regulation is inaccurate. Nor is any patchwork automatically adequate. Label audits, high-THC dose disclosures, child-safe packaging, independent labs and youth-focused education are testable policy interventions, though the best potency caps and unintended effects require evaluation.
There is also a paradox: criminalization harmed vulnerable communities while legalization opened a profitable market. The remedy for one failure need not be a return to the other. A defensible approach would preserve criminal-justice reform, examine record expungement and unequal enforcement, and scrutinize the health and business consequences of commercially maximizing intoxication.
The federal rule changed in 2026, but not in the way many headlines suggest
The legal landscape is not static. An April 23, 2026 Justice Department announcement placed FDA-approved marijuana products and qualifying state-licensed medical marijuana in Schedule III. Broader rescheduling has followed a separate formal process, as the DEA explains. This does not mean every recreational dispensary product received FDA approval or that federal rules are uniform across categories. Precisely what a product is, who licenses it, and what medical claims its maker makes remain essential distinctions.
PART VII — THE SAME SCIENTIST STUDYING CBD AS A MEDICINE
Clue 5: A drug can contain both a problem and a possibility
The apparent contradiction may be the most revealing part of Hurd’s career. She warns against indiscriminate high-dose THC while investigating whether cannabidiol (CBD) could help people with opioid addiction. In a 2019 randomized placebo-controlled trial, Hurd and colleagues tested 400 mg or 800 mg purified CBD in heroin-abstinent participants and observed reductions in cue-induced craving and anxiety. It was promising early-stage research—not proof that CBD has been established as a treatment for opioid use disorder or a substitute for proven treatment.
The FDA has approved purified CBD in the prescription drug Epidiolex for specified seizure disorders. It has also approved synthetic cannabinoid-related prescription products including dronabinol and nabilone for limited uses. These are regulated preparations with defined doses; they are not a government endorsement of every cannabis strain or gummy advertised as medicine.
Nor should medical benefits be erased. A 2025 federally funded evidence review found some cannabinoid products can produce small short-term improvements in predominantly neuropathic pain, alongside increased dizziness, sedation or nausea. Evidence was insufficient for many other indications and long-term outcomes. “No medicine exists in the plant” is too broad; “the plant cures everything” is worse.
For a separate scientific perspective, watch Yasmin Hurd’s TED talk on whether CBD could help people with opioid addiction. It explains her research hypothesis; it should not be mistaken for a completed demonstration of clinical effectiveness.
WHAT THE SCIENCE PROVES—AND WHAT THE INTERVIEW DOES NOT
| Claim | What the evidence supports |
|---|---|
| Today’s cannabis is stronger. | Established trend. THC levels in monitored cannabis increased markedly; market products vary and the dose matters. |
| Daily cannabis use has overtaken daily alcohol use. | Supported in defined survey measures. It does not mean more people use cannabis than alcohol overall. |
| Heavy/high-potency use is linked to psychosis. | Strong observational association. Odds ratios and attributable fractions are not individual destinies or randomized causal proof. |
| Cannabis use disorder and withdrawal exist. | Well established. Daily use alone is not a psychiatric diagnosis; loss of control and harm matter. |
| Weed is more dangerous than cigarettes. | Not established as an overall comparison. Hurd emphasizes psychiatric harms; tobacco’s mass mortality is documented. |
| CBD can cure opioid addiction. | Not established. Preliminary controlled research supports further trials, not a clinical cure claim. |
THE NEXT INVESTIGATION: SEVEN QUESTIONS FOR THE PEOPLE MAKING THE RULES
What percentage of local sales comes from products exceeding 60%, 70% or 80% THC? Can consumers see independent laboratory reports rather than just a number on a package? What do regulators know about marketing that appeals to young people? Are psychiatric and CHS emergency visits being tracked with consistent diagnostic standards? How are cannabis taxes allocated between public health, treatment and enforcement? Are racial disparities in possession arrests actually falling? And how much money is being devoted to independent studies of the products people are now buying?
None of those questions requires demonizing an adult who uses cannabis or denying the harms of prohibition. They require treating the product, the market and the evidence as three different things.
Hurd’s interview is unsettling because it describes a collective change that took place slowly enough to feel normal. Science is now catching up to that change. The unresolved question is whether policy will catch up to the science—or whether another generation will be asked to supply the data with its own health.
Question. Investigate. Reconsider.
SOURCES & FURTHER READING
- Mighty Pursuit, Yasmin Hurd interview, November 3, 2026 (primary source for Hurd’s statements).
- Caulkins (2024): U.S. cannabis-use frequency from 1979–2022.
- SAMHSA: 2025 National Survey on Drug Use and Health, Section 9 tables.
- NIDA potency monitoring and Giordano et al. (2025): label accuracy.
- Di Forti et al. (2019), The Lancet Psychiatry: potency, frequency and psychosis.
- Ontario cohort (2025), JAMA Network Open and NIH: Danish schizophrenia analysis (2023).
- CDC MMWR (2026): cannabis hyperemesis surveillance and coding change.
- CDC: cannabis use disorder; lung health; heart health.
- 2026 review of adolescent cannabis and brain development and Hurd’s 2025 review.
- Hurd et al. (2019), randomized CBD trial; FDA: approved cannabinoid medicines; 2025 chronic-pain evidence review.
- ACLU (2020): marijuana arrest disparities and DEA: 2026 rescheduling actions.
Health note: This is public-health reporting, not personal medical advice. People concerned about dependence, severe vomiting or new paranoia or hallucinations should seek qualified medical care; abrupt psychiatric changes warrant urgent evaluation. Evidence about cannabis continues to evolve.
