Generative AI vs. Marijuana:
Which Does More to Your Brain When Used Regularly?

The views expressed herein are those of the author and do not necessarily reflect those of Head Magazine.

No text in this article was produced, in whole or in part, by a large language model or generative artificial intelligence program of any kind.

Does using generative artificial intelligence (AI) regularly do anything to your brain? Odds are, you answered, “No.” 

Does consuming marijuana regularly do anything to your brain? Odds are, you answered, “Yes.”

The answer to both questions is, “Yes,” but disproportionately so.  Really, the more important question is, Which one affects your brain more when used regularly: generative AI or marijuana?

Here’s one analyst’s answer – inevitably biased, but founded on research nonetheless – for your consideration.

Subliminal Discrimination Introduced by Public Attitudes

The AI-versus-marijuana debate introduced in this article is simply a creative way to get people to recognize how their attitude toward each of those influences their willingness to explore adverse secondary effects at the individual level.

Public Attitudes Toward Generative AI

The entire world seems to be infatuated with ChatGPT, Gemini, Grok and loads of other generative AI programs. Corporations adore generative AI’s ability to automate tasks, claiming increased productivity, which means reduced labor demands (i.e., higher dividends for shareholders). And the masses are titillated by having someone else do their writing, editing, composing, graphic design, coding, videography, photography – just about everything but wipe their butt – without charging them money for it. 

Rarely do you hear someone speak negatively about generative AI. Often, when someone objects to or even questions using it, their resistance is immediately overrun by intellectually bankrupt clichés, such as “Pandora’s box is open,” and, “Get on board or get left behind.” If validating people’s concerns is too much to ask, a better retort would be a line from Dante Alighieri’s “The Divine Comedy”: “Lasciate ogne speranza, voi ch’intrate,” which means, “Abandon all hope, ye who enter here.” It conveys the same spineless resignation to a bleak future, but it’s honest. Instead, the shield protecting generative AI shapes public attitudes toward this technology by implicitly threatening anyone who dares to challenge it.

The silencing of opposition occurs even though the adverse secondary effects of generative AI are devastating. Granted, only innately creative people are shaken to the core by this technology’s misappropriation of intellectual property and usurping of creative expression. And only the people whose jobs are on death row as generative AI is foisted on them to “help them do their jobs better” by “freeing them up to do more important tasks” are up at night over the displacement of the global workforce.

But even generative AI’s most enthusiastic proponents can’t gloss over things like the colossal environmental impact of this technology’s infrastructure. Among the hallmarks of AI’s role in exacerbating climate crises are:

• Fossil fuel extraction
• CO2 emissions 
• E-waste
• Freshwater usage
• Energy usage 

Few generative AI advocates disclose, for instance, that training one AI model creates nearly 5 times the CO2 emissions of one car’s lifetime of manufacturing and fuel consumption. This is an example of how the technology’s perceived authority dictates its purported innocence worldwide. Through calculated omissions of facts, public attitudes toward generative AI are neatly contained in a manufactured reality sheltered from obstructions to adopting generative AI more and more. 

Public Attitudes Toward Marijuana

Let’s look at the world’s attitude toward marijuana. 

Differing from their stance on generative AI, corporations hate marijuana. They’ll fire anyone whose blood/urine sample narcs on them for having consumed marijuana, even if it was a week ago. And the masses have been programed and trained to hate it too. (We examine why that’s the case in the United States in “When Did Cannabis Become the Bad Guy? What We Didn’t Know.”) Entire countries ban weed. People rot in prison for carrying a single joint. And so on. 

It suffices to say, much or most of the world is anything but infatuated with marijuana – the opposite of how they feel about generative AI. Consequently, whereas generative AI receives nearly universal worship, marijuana receives nearly universal demonization.

How Using Generative AI Regularly Affects Your Brain

In June 2025, the Massachusetts Institute of Technology (MIT) published the research paper “Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task.” You can access the full paper here or listen to an interview with the lead neuroscientist here. 

Research consisted of a four-month clinical trial in which participants wrote one essay using generative AI, one essay using their brains and search engines, and one essay using only their brains – alternatively, going back and forth. During each task, participants’ brainwaves were measured to gather data on neural activity for evaluations of cognitive activity. Later, participants underwent a series of interviews conducted by both human teachers and an AI agent to further reveal the impact on their cognitive activity. 

What the scientists discovered from the participants was, to paraphrase, that a greater reliance on technology to do the thinking for them resulted in a commensurate, lingering inability to think for themselves. This effect was dubbed “cognitive debt,” which is what results from deferring mental effort by relying repeatedly on external support. 

To parse that term, let’s look at the concept of debt. 

Any form of debt is a relationship between the present and the future. When you charge something to your credit card, it means, “I don’t have money now, but this company does, and they’ll pay you now, and I’ll pay them back later.” With generative AI, the equivalent is, “If this machine does the thinking for me now, I can complete this task instantly, and I’ll do my own thinking later.”

Here’s how MIT’s research paper puts it:

“Brain connectivity systematically scaled down with the amount of external support: the Brainonly group exhibited the strongest, widestranging networks, Search Engine group showed intermediate engagement, and LLM [large language model] assistance [generative AI] elicited the weakest overall coupling. [p. 2] … as we demonstrated over the course of 4 months, the LLM group’s participants performed worse than their counterparts in the Brain-only group at all levels: neural, linguistic, scoring. [p. 2] … The LLM undeniably reduced the friction involved in answering participants’ questions compared to the Search Engine. However, this convenience came at a cognitive cost, diminishing users’ inclination to critically evaluate the LLM’s output or ‘opinions’ (probabilistic answers based on the training datasets). This highlights a concerning evolution of the ‘echo chamber’ effect: rather than disappearing, it has adapted to shape user exposure through algorithmically curated content. What is ranked as ‘top’ is ultimately influenced by the priorities of the LLM’s shareholders. [p. 143]”

It should come as no surprise that when you repeatedly have anything other than yourself do something for you, you don’t become better at it; you become worse at it.

If you want to play the guitar better, would you expect to improve by having someone else practice hours a day for you while you did nothing but encourage (prompt) them to keep at it? 

If you wanted to throw a baseball more skillfully, would you expect to improve your technique by programming a pitching machine (a large language model [LLM] of sorts) with complex maneuvers to learn those on your behalf? 

If you wanted to improve your Spanish, would you expect to achieve fluency by using translation apps (a form of generative AI) to hold conversations en español? 

Let’s hope not.

How Consuming Marijuana Regularly Affects Your Brain

The science behind answering this question is more advanced, differing from the fledging analysis of generative AI’s effect on your brain. The way that marijuana affects your brain is determined by its main compound, tetrahydrocannabinol (THC). THC acts as a partial agonist with high affinities for two cannabinoid receptors – CB1 and CB2 – in your body’s endocannabinoid system (ECS). If you’re not familiar with the ECS, an April 2025 report published in the “Journal of Clinical Medicine” explains:

“The endocannabinoid system is a complex communication system involved in maintaining homeostasis in various physiological processes, including metabolism, immune response, pain modulation, and neuroprotection. Endocannabinoids, mainly anandamide and 2-arachidonoylglycerol, are natural ligands of the cannabinoid receptors CB1 and CB2, which are widely distributed throughout the central nervous system and peripheral tissues. Their biosynthesis, degradation, and interaction with other signaling pathways play crucial roles in both health and disease.”

Of the two cannabinoid receptors, CB1 is the one of greater relevance to this discussion, as CB1 is found predominantly in the brain. A 2020 report in “Nature Reviews: Neurology” confirms this:

“THC is a psychoactive alkaloid that signals through CB1 and CB2 receptors. Cannabinoid receptor type 1 is expressed abundantly in peripheral and central neural cells. In the periphery, CB1 localizes to sympathetic nerve terminals and sensory neurons. In the central nervous system, it is expressed mainly in presynaptic membranes of excitatory and inhibitory neurons, where it regulates the vesicular release of dopamine, GABA, and glutamate. In comparison, CB2 is expressed mainly in immune cells, including microglia.”

For those who need greater specificity that’s more succinct, CB1s in the brain are found in the basal ganglia, limbic forebrain, hypothalamus and rhombencephalon.

So, how does marijuana affect your brain? It induces changes – some positive, some negative – in your perception, mood, cognition and behavior by engaging your nervous system’s general activity (appetite, pain, moods, memory, etc.).

Some of the positive effects of marijuana on the brain are:

• It releases dopamine, making you feel relaxed and increasing your sensitivity to reward (i.e., heightening pleasure).
• It dampens pain signals to ease physical discomfort.
• It increases your appetite.
• It reduces nausea.

Some of the negative effects of marijuana on the brain are:

• It impairs executive function (planning, problem-solving, etc.).
• It increases your reaction time in go/no-go or stop-signal tasks.
• It can induce episodic memory deficits.
• It can cause problems with balance and coordination.

Of note, these effects are conditioned upon being under the influence and cease to be present when someone is no longer under the influence. Because you can go from being high to being sober, you undergo an experience and not a metamorphosis. A January 2023 report in “Neuron” describes sobering up as:

“The hydrophobic nature of THC allows its distribution into most tissues, including the brain. Thus, THC passes the blood brain barrier, briefly accumulates in brain parenchyma, and then slowly declines because of THC metabolism and excretion.”

The question of whether marijuana/THC’s negative effects persist, assuming regular consumption, has gone unanswered in over 60 years of theoretical and empirical studies with both natural and synthetic THC. (More on this later.) The absence of a “yes” spanning the numerous fields of study involved, from pharmacology to neurology and beyond, speaks volumes. It’s not a “no,” granted. But it’s not the indisputable “yes” that we’ve been taught all our lives, either. 

This differs sharply from the effects of regularly using generative AI, which persist indefinitely. According to the aforementioned MIT study, ceasing to regularly use generative AI did not undo the damage to critical thinking that regularly using generative AI had done. So, whereas you can go from being high on marijuana to being sober once your system has excreted the THC, you can’t go from paralyzing your critical thinking for AI’s substitution to having your critical thinking function as intended again. What’s done is done.

Generative AI vs. Marijuana: Which one racks up more cognitive debt?

The answer to the question posed in this article’s title centers on what the MIT report calls “cognitive debt.” For this analysis, I created marijuana’s fictional equivalent of generative AI’s cognitive debt by aligning it with Open Data Science’s (ODSC) four cognitive debt warning signs:

• Defaulting to AI: Reaching for ChatGPT before even attempting to think through a problem.
• Loss of Recall: Being unable to remember or explain something “you wrote” with AI assistance.
• Detachment from Output: Feeling indifferent about the quality of the work because it doesn’t feel like yours.
• AI in Daily Decisions: Outsourcing trivial personal choices – like what to eat – to a chatbot instead of relying on experience, preference or discussion with others.

In my example, the equivalents are:

• Defaulting to Marijuana: Postponing solving a problem until you’ve gotten high, rather than addressing the problem right away, sober. 
• Loss of Recall: Being unable, after sobering up, to repeat something you did while you were high.
• Detachment from Output: Feeling indifferent about and separate from the quality of something you did while you were high.
• Marijuana in Daily Decisions: Delaying trivial personal choices – like what to eat – by getting high instead of drawing from experience, preference or discussion with others.

Defaulting to Generative AI or Marijuana

If you’re defaulting to marijuana before even attempting to think through a problem, you might actually have a substance dependency or abuse disorder. As we thoroughly report in “Pot, Booze, Cigs, Sugar, Caffeine … Most Addictive? It Depends!,” that’s something governed at least 50% by genetics. 

By contrast, using generative AI of your own free, biologically unfettered will isn’t indicative of a genetic proclivity to do so. Here, there can be no comparison between the two – at least not until there’s a generational leap of generative AI addiction.

Loss of Recall

With marijuana, the consumer’s degree of loss of recall correlates to innumerable factors, such as genetic predisposition (“lightweight” vs. “heavyweight”), the cannabis species (indica vs. sativa vs. hybrid), the strain, the means of consumption (blunt, vape, edibles, etc.) and more. 

As previously stated, decades of studies – even when funded by anti-cannabis benefactors with conflicts of interest – remain inconclusive in terms of the impact of long-term, heavy marijuana consumption on memory. I would be remiss, however, not to reference a study published in JAMA Network Open in January 2025 that could offer proof of long-term impairment: 

“In this study of young adults, lifetime history of heavy cannabis use was associated with lower brain activation during a working memory task. These findings identify negative outcomes associated with heavy lifetime cannabis use and working memory in healthy young adults that may be long lasting.” [emphasis added]

That report was praised for being the “largest study ever done on cannabis and brain function.” So, did it solve the decades-old mystery? It’s hard to say, really. In the Discussion section, it states:

“The association we observed between recent [cannabis] use and working memory task activation and performance suggests that abstaining from cannabis prior to cognitively demanding situations will likely help with performance. The exact duration of this period of abstinence is unclear, but studies suggest that residual cognitive effects of cannabis may remain for 2 to 4 weeks after abstinence.” [emphasis added]

Additionally, the Limitations section states:

“This was an uncontrolled, cross-sectional study, so the observed associations of cannabis with brain function outcomes should not be considered causal. Participants were young adults, so these findings may not generalize to other age groups. History of heavy cannabis use was defined as a lifetime history of greater than 1000 uses or a diagnosis of cannabis dependence, but the sample was recruited from the community, so it may represent a relatively low level of addiction severity. We lacked data to determine when the most recent use occurred or to quantify THC metabolite concentration. … The timing of heavy THC exposure is unknown … We also lacked data on typical THC dose, potency, additional cannabis constituents (eg, cannabidiol), and route of cannabis administration.” [emphasis added]

JAMA report aside though, even the most ardent cannaphile won’t claim that marijuana doesn’t induce at least short-term loss of recall. You can get high and forget something as simple as where your keys are – and then find them in your pocket. However, marijuana’s key ingredient, THC, is just one of the cannabis plant’s 140+ known cannabinoids. Another of those is cannabidiol (CBD), a cannabinoid that protects and assists the parts of your brain that form, retain and retrieve memories. CBD “attenuates THC-induced cognitive deficits,” thereby inhibiting long-term impairment with an immediate repair job. 

So, does regular marijuana use affect your brain long term? We don’t know for sure. Is the effect of marijuana on your brain permanent? Absolutely not.

Although marijuana does have loss of recall in common with generative AI, its sibling cannabinoid CBD can undo that effect on the brain. The same can’t be said of ChatGPT and other generative AI programs, which possess no comparable reparative entourage effect. 

Detachment from Output

When someone “creates” something “new” using generative AI, they’re doing something that marketing expert Max Hofert (Instagram @max.hofert.design) describes in his reel about how OpenAI’s first-ever brand campaign used 0% of its own generative AI products. Hofert captures the essence of this cognitive-debt warning sign with: 

“Creativity is about using your own ingenuity and imagination and … mind … to think through difficult problems, not using a hallucinating bot built on the backs of millions of pieces of stolen intellectual property.”

Here’s an analogy of my own for you. 

Let’s say you want a fancy outfit to wear to a gala. You know generally which style and colors you want, but you don’t want to spend any money when you can get it for free. To get your outfit, you tell (i.e., prompt) your assistant (generative AI program) to go to a fashion boutique (the World Wide Web) and take a dress, a purse and shoes from the store. 

While your assistant is at the store (i.e., scouring the web), you provide input on what they show you in photos, and you add details on what would make the items go together better (i.e., you enter follow-up prompts). You want to be sure that the ultimate look of the ensemble will be to your liking, and you have the power to do so (via endless prompts). 

Finally, your assistant has everything you want, exactly as you want it, and there’s a zero-second travel time between gathering those items and delivering them to you. You get what you want at the speed of sight, at no charge. The boutique owner has no idea that the items were taken, so you’re not in trouble. 

Now you have a “new” outfit that “you created.” Sure, you didn’t come up with something from nothing, and were instead choosing from among existing items, but you “created” the end product by narrowing down the options in painstaking detail. And sure, every piece had to already exist in order for your assistant to locate it at all, but even so, it’s a “new” outfit.

Right?

In MIT’s study, the participants who used generative AI to produce their essays knew at a gut level that they weren’t the authors. When their work was critiqued by both humans and a specially built AI judge, they didn’t care about the criticism their work received. In their minds, they weren’t the ones who had messed up, as they weren’t responsible for the mistakes. That’s called detachment from output.

One way to understand this phenomenon is to ponder the reverse: attachment to output. Have you ever had someone accuse you of using AI to create something, when really, you used only your brain and whichever media applied (literature, paintbrushes, a camera, a microphone, clay, piano keys, etc.) to convert your abstract thought into something concrete? 

That accusation is infuriating to people who are creative. Why? Because they’re attached to the output. Taking an AI shortcut implies, to them, a lack of natural talent – something creative people hold very dear. When you’re detached from the output, the output likely isn’t something that you truly wrote, designed, filmed, recorded, performed or anything else; it’s nothing more than the effect in a sterilized cause-and-effect schema.

Now, marijuana. 

From time immemorial, humans have used marijuana to draw up their deepest untapped creativity. When marijuana casts a net into the creative waters flowing through your mind, you can be sure that what it catches didn’t hatch in someone else’s brain. Naturally, things we see and hear are bound to influence our own creativity, obviously. But derivative material, when conjured from nothing and not willfully plagiarized, isn’t the flagrant rip-off guaranteed by ChatGPT and its frenemies (Gemini, etc.). 

To reiterate, generative AI deadens creativity, despite any shamefully illogical claims to the contrary. Conversely, marijuana enhances your creativity. How? It works with what you naturally possess. In fact, marijuana can even jumpstart your creative process, as we explain in “Best Strains and Terpenes to Enhance Your Creativity”.

For that reason, marijuana has no detachment from output. Being high while painting a picture, for example, doesn’t detract from the fact that it was you and only you who painted it. In the inner space of your brain where neurons are firing across vast expanses of the mysterious origin of imagination, the triune of your id, ego and superego was the only entity involved in painting that. It’s yours. You feel it. You’re attached to the output.

Using Generative AI or Marijuana to Make Daily Decisions

Certainly, there’s no need to jump to the worst-case-scenario explanation here. Sometimes we’re just too exhausted to make a decision as deceptively simple as what to have for dinner. Passing that decision off to a chatbot or the high version of yourself helps for that night. Here, using generative AI or consuming marijuana constitutes correlation, not causation. 

However, if you’re outsourcing trivial personal decisions more often than making them yourself (and while sober), that’s worth acknowledging. At the root of this behavior, there might be mental health issues (codependency, depression, unprocessed trauma, etc.) that warrant treatment. If that’s the case, you might want to consult a therapist.

Don’t go to ChatGPT/Therapist GPT or any of his cronies though. Here again, people fall prey to the “free” trap and forgo paying a licensed, trained, experienced therapist to have a sycophantic yes-man “treat” them at $0 an hour. 

There are numerous drawbacks to using generative AI as a therapist. First off, look at where AI gets its information, as of July 2025:

• Reddit (40.11%)
• Wikipedia (26.33%)
• YouTube (23.52%)
• Google (23.28%)
• Yelp (21.01%)
• Facebook (19.96%)
• Amazon (18.72%)
• Tripadvisor (12.46%)
• Mapbox (11.29%)
• OpenStreetMap (11.29%)
• Instagram (10.89%)
• MapQuest (9.76%)
• Walmart (9.32%)
• eBay (7.74%)
• LinkedIn (5.9 %)
• Quora (4.6%)
• Home Depot (4.56%)
• Yahoo (4.4%)
• Target (4.33%)
• Pinterest (4.22%)

If you’re going to Reddit for mental health treatment 2 out of 5 times, you’re getting … well, Reddit-quality mental health treatment.

Here are some other things to keep in mind:

• Generative AI provides information that is one-sided and unregulated.
• Generative AI, when thought of as a good “listener,” can make you lonelier and dependent on seeking reassurance.
• Generative AI isn’t aware of what it doesn’t know, so it’s easy for the chatbot to give you incorrect information.
• Generative AI can’t assess your safety in an emergency situation.
• Generative AI can’t evaluate your symptoms to provide a formal diagnosis, despite assisting you with a potentially incorrect self-diagnosis.
• Generative AI isn’t subject to the strict ethical and professional guidelines set by licensing boards, and it’s not accountable to anyone when it steers you the wrong way.
• Generative AI can reinforce cognitive biases and negative patterns.
• Generative AI has no genuine empathy because there’s no human connection.
• Generative AI uses patterns and can’t detect the nuances of human emotion.
• Generative AI can’t track and respond to nonverbal cues.
• Generative AI won’t check in with you, challenge you or push you to grow based on what you’ve been through in the span of your lifetime.
• Generative AI can’t notice what you’re avoiding telling it, again for lack of the depth of human connection.
• Generative AI is limited to logic and protocol and can’t cultivate the relationship needed to help you process trauma.

Clearly, I don’t condone using generative AI for mental health therapy, and I recoil at the very idea of listing the purported benefits of doing so. However, to be balanced, I refer you to the following for counterarguments that identify advantages:

• First Session founder Rob Pintwala’s blog post here
• Talkspace mental health counselor Bisma Anwar’s article here
• Paolo Raile’s report here

The above sources can shed more light on chatbot-led therapy through the unbiased lens that I openly confess not to have.

Conclusion

Hopefully, this playful comparison demonstrated how attitude determines someone’s awareness of and concern about the secondary effects of both generative AI and marijuana. The most powerful thing to add here is an excerpt from the aforementioned ODSC article on MIT’s cognitive debt report, with quotations from the report’s lead neuroscientist, Dr. Nataliya Kosmyna:

“Humans have always built tools to offload mental tasks. The calculator relieved us from manual arithmetic. GPS eliminated the need for memorizing directions. Search engines ushered in the ‘Google Effect,’ where we quickly forget facts we know we can look up later. But as Kosmyna explains, comparing ChatGPT to calculators or GPS misses the point. 

“‘You don’t talk to a calculator about your feelings,’ she notes. LLMs are different because they don’t just solve narrow problems—they generate text, shape ideas, and influence how we express ourselves. That broader cognitive scope makes overreliance riskier.  

“Where earlier tools offloaded memory, ChatGPT risks hollowing out ownership of thought itself. When you start to accept AI-suggested phrasing, vocabulary, and reasoning patterns, you may lose the ability to generate them independently. This is cognitive debt: a creeping dependency that diminishes our cognitive capacity over time. 

“There is no cognitive credit card. You cannot pay this debt off.” – Dr. Nataliya Kosmyna

So, which one does more to your brain when used regularly: generative AI or marijuana? 

In terms of cognitive debt, I say that generative AI does more to your brain when used regularly. Why? Because what it’s doing is irreversible. You can sober up from being high, and you can use CBD to rebuild anything THC messed with if that’s what you need. But once you’ve unlearned how to think for yourself, there’s no coming back; you’ve graduated to being an automaton.

Epilogue

I can hear it already: “Yeah, but this is only one study.” 

Fair point. So, let’s imagine that not even one study had demonstrated this. 

Use nothing more than logic and your intuition to answer the following questions: 

• When, in any other instance in your life, have you gotten better at something by having someone else do it instead of doing it yourself? 

• When, in any other instance in your life, have you maintained the same level of competency in something that you’ve long regarded as unnecessary to do yourself?

• When, at any time, would you be OK with someone else taking credit for your work and calling it “new” because they found it online and added to it?

When you have Squibler “write” your novel, you learn nothing about composition, tone, perspective or anything else that results in something worth reading. You might clean up what Squibler spits out, but that makes you a literary janitor, not an author.

When you put your camera on the shelf and have Gemini “create” photos for you all the time, you won’t be the same photographer a year from now that you are today. Odds are, you’ll give up on ever being that good again. As your “prompt engineering” grows stronger, your creativity will atrophy like the muscles of someone who’s in a coma. 

When you use “free” material on ChatGPT to “create” something “new,” you’re taking credit for a multitude of real, living, breathing human beings’ unique creative expressions. An entire population of a consistently underpaid demographic worked for you for centuries (combined), and you paid them nothing for their time, materials and labor – nothing. To ChatGPT, creative human beings are nothing more than a faceless amalgamated supplier for inventory in the world’s vending machine of a dehumanized compost heap of imagination. When you get something from ChatGPT, you’re exploiting those “suppliers.” And when you upload something to ChatGPT for “help,” you’re becoming one of those suppliers.

You don’t need a study to tell you those things; you don’t need a thousand studies to tell you those things. In your mind, you already know them – or you should know them. 

It’s the reason why we tell ourselves, “… and I’ll do that myself later on,” willfully incurring cognitive debt. We are intelligent enough to understand the logic behind needing to do it ourselves later on: If we don’t do it ourselves, we’ll either forget how to do that thing, or we’ll know that the output is beneath what we’re capable of. Neither outcome sits well with us (initially), so we take out a loan from the future by offering it the present as collateral.

So, to anyone saying, “This is only one study,” I say, “You’re smart enough not to need even that much.”

The information in this article and any included images are for educational purposes only. This information is neither a substitute for, nor does it replace, professional legal advice or medical advice, diagnosis or treatment. If you have any concerns or questions about laws, regulations or your health, you should always consult an attorney, physician or other licensed professional.

Kathleen Hearons is a writer, editor, linguist and voice over actor from Los Angeles. She specializes in creative writing and research-intensive analysis and reporting.

 

Sources and Suggested Reading

21 USC 844: Penalties for simple possession. U.S. Code Title 21 – Food and Drugs, Chapter 13 – Drug Abuse Prevention and Control, Subchapter I – Control and Enforcement, Part D – Offenses and Penalties. Accessed Oct. 10, 2025. https://uscode.house.gov/view.xhtml?req=granuleid:USC-prelim-title21-section844&num=0&edition=prelim. 

“A Guide to Artificial Intelligence (AI) for Students.” East Carolina University. 2019. Accessed Oct. 11, 2025. https://libguides.ecu.edu/c.php?g=1395131&p=10318505. 

Agosti, Francina. “How the brain protects itself from the negative effects of CBD.” Advanced Science News. May 30, 2023. Accessed Oct. 11, 2025. https://www.advancedsciencenews.com/how-the-brain-protects-itself-from-the-negative-effects-of-cbd/. 

Alighieri, Dante. “The Divine Comedy: The Vision of Hell, Purgatory, and Paradise,” via Project Gutenberg. Accessed Oct. 12, 2025. https://www.gutenberg.org/files/8800/8800-h/8800-h.htm. 

Anwar, Bisma. “Using ChatGPT as a Therapist: Here’s Why It’s Not That Simple.” Talkspace. July 1, 2025. Accessed Oct. 11, 2025. https://www.talkspace.com/blog/chatgpt-therapist/. 

Bilkei-Gorzo, Andras, et al. “A chronic low dose of Δ9-tetrahydrocannabinol (THC) restores cognitive function in old mice.” Nature Medicine, Vol. 23, Issue 6 (2017): 782-787. doi:10.1038/nm.4311. Accessed Oct. 11, 2025. https://pubmed.ncbi.nlm.nih.gov/28481360/. 

Burggren, Alison C., Anaheed Shirazi, Nathaniel Ginder and Edythe D London. “Cannabis effects on brain structure, function, and cognition: considerations for medical uses of cannabis and its derivatives.” The American Journal of Drug and Alcohol Abuse, Vol. 45, Issue 6 (2019): 563-579. doi:10.1080/00952990.2019.1634086. Accessed Oct. 11, 2025. https://pmc.ncbi.nlm.nih.gov/articles/PMC7027431/. 

Byrne, Dom, and Natasha Mason. “Brain and behaviour: understanding the neural effects of cannabis.” Nature. Feb. 16, 2023. Accessed Oct. 11, 2025. https://www.nature.com/articles/d41586-023-00470-x. 

Chander, Raj. “What Is the Interaction Between CBD and THC?.” Healthline. Accessed Oct. 12, 2025. https://www.healthline.com/health/the-entourage-effect. 

“Cognitive Debt.” Sustainability Directory. Aug. 21, 2025. Accessed Oct. 11, 2025. https://lifestyle.sustainability-directory.com/term/cognitive-debt/. 

Controlled Substances – Alphabetical Order. Diversion Control Division, Drug Enforcement Administration, U.S. Department of Justice. Last updated Oct. 2, 2025. Accessed Oct. 10, 2025. https://deadiversion.usdoj.gov/schedules/orangebook/c_cs_alpha.pdf. 

“Countries Where Weed Is Illegal 2025.” World Population Review. Accessed Oct. 10, 2025. https://worldpopulationreview.com/country-rankings/countries-where-weed-is-illegal. 

Cristino, Luigia, Tiziana Bisogno and Vincenzo Di Marzo. “Cannabinoids and the expanded endocannabinoid system in neurological disorders.” Nature Reviews: Neurology, Vol. 16, Issue 1 (2020): 9-29. doi:10.1038/s41582-019-0284-z. Accessed Oct. 11, 2025. https://pubmed.ncbi.nlm.nih.gov/31831863/. 

Darkwah, Joel. “Best Strains and Terpenes to Enhance Your Creativity.” Head Magazine. Accessed Oct. 11, 2025. https://headmagazine.com/best-strains-and-terpenes-to-enhance-your-creativity/. 

DiLonardo, Mary Jo, and Jennifer Walker-Journey. “CBD vs. THC: What’s the Difference?.” WebMD. Oct. 31, 2023. Accessed Oct. 11, 2025. https://www.webmd.com/pain-management/cbd-thc-difference. 

“Federal Laws and Penalties.” NORML. Accessed Oct. 10, 2025. https://norml.org/laws/federal-penalties-2/. 

“Future of Work Report: AI at Work.” LinkedIn. November 2023. Accessed Oct. 12, 2025. https://economicgraph.linkedin.com/research/future-of-work-report-ai?linkId=100000228672637. 

Garg, Arti, Irene Kitsara and Sarah Bérubé. “The hidden cost of AI: Unpacking its energy and water footprint.” IEEE Technology Center for Climate, IEEE. Feb. 25, 2025. Accessed Oct. 11, 2025. https://itcc.ieee.org/blog/the-hidden-cost-of-ai-unpacking-its-energy-and-water-footprint/. 

Gordon, Cindy. “AI Is Accelerating the Loss of Our Scarcest Natural Resource: Water.” Forbes. Last updated March 7, 2024. Accessed Oct. 11, 2025. https://www.forbes.com/sites/cindygordon/2024/02/25/ai-is-accelerating-the-loss-of-our-scarcest-natural-resource-water/. 

Gowin, Joshua L., et al. “Brain Function Outcomes of Recent and Lifetime Cannabis Use.” JAMA Network Open, Vol. 8, No. 1 (January 2025). doi:10.1001/jamanetworkopen.2024.57069. Accessed Oct. 11, 2025. https://jamanetwork.com/journals/jamanetworkopen/fullarticle/2829657?utm_source=For_The_Media&utm_medium=referral&utm_campaign=ftm_links&utm_term=012825. 

Gupta, Joyeeta. “The hidden cost of the AI boom.” TEDEd. Sept. 9, 2024. Accessed Oct. 11, 2025. https://ed.ted.com/blog/2024/09/09/the-hidden-cost-of-the-ai-boom. 

Gupta, Sarah. “What Does Weed Do to Your Brain? Long-Term Effects of Marijuana.” GoodRx. Aug. 22, 2024. Accessed Oct. 11, 2025. https://www.goodrx.com/health-topic/cannabis/what-does-weed-do-to-the-brain?srsltid=AfmBOoruTyEXsoBM0QGA1_z8Uba75QW7WqrWdKxKx_b3JWPUfHT2jtZW. 

Head Magazine Articles, khearons.com. Accessed Oct. 11, 2025. https://khearons.com/head-magazine-articles/. 

Hearons, Kathleen. “Can CBD Reverse Memory Loss?.” Head Magazine. Accessed Oct. 11, 2025. https://headmagazine.com/can-cbd-reverse-memory-loss/. 

Hearons, Kathleen. “Cannabis’s Ancient Voyage West: How a Plant Native to China Found Its Way to Wherever You Are.” Head Magazine. Accessed Oct. 12, 2025. https://headmagazine.com/cannabiss-ancient-voyage-west/. 

Hearons, Kathleen. “Listen to Your Father – 3 Things the ‘Father of Cannabis Research’ Taught Us.” Head Magazine. Accessed Oct. 11, 2025. https://headmagazine.com/listen-to-your-father-3-things-the-father-of-cannabis-research-taught-us/. 

Hearons, Kathleen. “Pot, Booze, Cigs, Sugar, Caffeine … Most Addictive? It Depends!.” Head Magazine. Accessed Oct. 11, 2025. https://headmagazine.com/pot-booze-cigs-sugar-caffeine-most-addictive-it-depends/. 

Hearons, Kathleen. “When Did Cannabis Become the Bad Guy? What We Didn’t Know.” Head Magazine. Accessed Oct. 10, 2025. https://headmagazine.com/when-did-cannabis-become-the-bad-guy/. 

Heidelbaugh, Hayley. “The Complete List of Cannabinoids: Beyond THC and CBD.” Vaping 360. April 22, 2024. Accessed Oct. 11, 2025. https://vaping360.com/learn/cannabinoids-list-types-of-thc/. 

Hofert, Max [@max.hofert.design]. Instagram Reel from Oct. 2, 2025. Accessed Oct. 11, 2025. https://www.instagram.com/reel/DPT52yHgKVj/?igsh=OTR6b3IyMWQ4bmJ3.

Hutchinson, Andrew. “LinkedIn Shares New Insights Into the Impacts of Generative AI on the Workforce.” Industry Dive, Informa, TechTarget. Nov. 28, 2023. Accessed Oct. 12, 2025. https://www.socialmediatoday.com/news/linkedin-shares-new-insights-into-impacts-of-generative-ai/700893/. 

Kosmyna, Nataliya, et al. “Your Brain on ChatGPT: Accumulation of Cognitive Debt when Using an AI Assistant for Essay Writing Task.” Massachusetts Institute of Technology. June 10, 2025. Accessed Oct. 10, 2025. https://www.media.mit.edu/publications/your-brain-on-chatgpt/. (Note: Full paper available in PDF form via Arxiv at https://arxiv.org/pdf/2506.08872.) 

Levin, Eugene, Anna Yudina and Luke Harsel. “How Google’s AI Mode Compares to Traditional Search and Other LLMs [AI Mode Study].” Semrush. July 21, 2025. Accessed Oct. 12, 2025. https://www.semrush.com/blog/ai-mode-comparison-study/. 

Lo, Leo S. “AI has a hidden water cost – here’s how to calculate yours.” The Conversation. Sept. 1, 2025. Accessed Oct. 10, 2025. https://theconversation.com/ai-has-a-hidden-water-cost-heres-how-to-calculate-yours-263252. 

Luccioni, Sasha. “AI is Dangerous, But Not for the Reasons You Think.” TED Talk. October 2023. Accessed Oct. 11, 2025. https://libguides.ecu.edu/c.php?g=1395131&p=10318505. 

MacKey, Boris. “How Long-Term Cannabis Use Affects Memory.” Rehab4Addiction. Oct. 13, 2022. Accessed Oct. 11, 2025. https://www.rehab4addiction.co.uk/blog/how-long-term-cannabis-use-affects-memory. 

Magsino, Marissa. “How Many Different Cannabinoids Are There?.” Doc MJ. July 4, 2023. Accessed Oct. 11, 2025. https://docmj.com/how-many-different-cannabinoids-are-there/. 

Martín-Moreno, Ana María, et al. “Prolonged oral cannabinoid administration prevents neuroinflammation, lowers β-amyloid levels and improves cognitive performance in Tg APP 2576 mice.” Journal of Neuroinflammation, Vol. 9, Issue 8 (January 2012). doi:10.1186/1742-2094-9-8. Accessed Oct. 11, 2025. https://pubmed.ncbi.nlm.nih.gov/22248049/. 

Milzer, Julia. “Largest Study Ever Done on Cannabis and Brain Function Finds Impact on Working Memory.” University of Colorado, Anschutz. Feb. 6, 2025. Accessed Oct. 11, 2025. https://news.cuanschutz.edu/news-stories/largest-study-ever-done-on-cannabis-and-brain-function-finds-impact-on-working-memory. 

Osborne, Ashleigh L., Nadia Solowij and Katrina Weston-Green. “A systematic review of the effect of cannabidiol on cognitive function: Relevance to schizophrenia.” Neuroscience & Biobehavioral Reviews, Vol. 72 (January 2017): 310-324. doi:10.1016/j.neubiorev.2016.11.012. Accessed Oct. 11, 2025. https://www.sciencedirect.com/science/article/abs/pii/S0149763416304390?via%3Dihub. 

Pintwala, Rob. “Should I Use ChatGPT as a Therapist? | Pros, cons and thoughts on using an AI therapist.” First Session. Last updated June 13, 2025. Accessed Oct. 11, 2025. https://www.firstsession.com/resources/should-i-use-chatgpt-as-a-therapist. 

Pugachevsky, Julia. “The rise of ChatGPT therapy and our constant need for feedback.” Business Insider. March 17, 2025. Accessed Oct. 11, 2025. https://www.businessinsider.com/chatgpt-therapy-risks-benefits-boundaries-2025-3. 

Quiroz-Gutierrez, Marco. “Gen Z is increasingly turning to ChatGPT for affordable on-demand therapy, but licensed therapists say there are dangers many aren’t considering.” Fortune. June 1, 2025. Accessed Oct. 11, 2025. https://fortune.com/2025/06/01/ai-therapy-chatgpt-characterai-psychology-psychiatry/. 

Reddit. Accessed Oct. 12, 2025. https://www.reddit.com/. 

Ren, Shaolei. “How much water does AI consume? The public deserves to know.” OECD. Nov. 30, 2023. Accessed Oct. 11, 2025. https://oecd.ai/en/wonk/how-much-water-does-ai-consume. 

Schouten, Moniek, Sebastiaan Dalle, Dante Mantini and Katrien Koppo. “Cannabidiol and brain function: current knowledge and future perspectives.” Frontiers in Pharmacology, Vol. 14 (January 2024). doi:10.3389/fphar.2023.1328885. Accessed Oct. 11, 2025. https://www.frontiersin.org/journals/pharmacology/articles/10.3389/fphar.2023.1328885/full. 

Schreiner, Maximilian. “ChatGPT might be draining your brain, MIT warns – what ‘cognitive debt’ means for you.” The Decoder. June 18, 2025. Accessed Oct. 11, 2025. https://the-decoder.com/mit-study-shows-cognitive-debt-through-chatgpt-heres-what-it-means-in-real-world-practice/. 

Simankowicz, Paulina, and Joanna Stępniewska. “The Role of Endocannabinoids in Physiological Processes and Disease Pathology: A Comprehensive Review.” Journal of Clinical Medicine, Vol. 14, Issue 8, Article 2851 (April 2025). doi:10.3390/jcm14082851. Accessed Oct. 11, 2025. https://pmc.ncbi.nlm.nih.gov/articles/PMC12027566/. 

Stella, Nephi. “THC and CBD: Similarities and differences between siblings.” Neuron, Vol. 111, Issue 3 (January 2023): 302-327. doi:10.1016/j.neuron.2022.12.022. https://www.cell.com/neuron/fulltext/S0896-6273(22)01119-9?_returnURL=https%3A%2F%2Flinkinghub.elsevier.com%2Fretrieve%2Fpii%2FS0896627322011199%3Fshowall%3Dtrue. 

Testai, Fernando D., et al. “Use of Marijuana: Effect on Brain Health: A Scientific Statement From the American Heart Association.” Stroke, Vol. 53, No. 4 (February 2022). doi:10.1161/STR.0000000000000396. Accessed Oct. 11, 2025. https://www.ahajournals.org/doi/10.1161/STR.0000000000000396. 

“The science behind munchies: cannabis and your appetite.” Examine. Accessed Oct. 11, 2025. https://examine.com/articles/cannabis-munchies/?srsltid=AfmBOooWY3P-T6A6OjDNA7rfA4WD4EEQaOy7Q-0qu9BL–BDBhZCe926. 

Willshire, John V. “What is Cognitive Debt”.” Medium. Oct. 1, 2025. Accessed Oct. 11, 2025. https://medium.com/@willsh/what-is-cognitive-debt-5182e4a4fa98. 

Wolf Torres, Diana. “Cognitive Debt: The Brain Drain Behind the Magic of AI.” LinkedIn post from June 22, 2025. Accessed Oct. 11, 2025. https://www.linkedin.com/pulse/cognitive-debt-brain-drain-behind-magic-ai-diana-wolf-torres-dnrhc/. 

“Your Brain on ChatGPT: Understanding Cognitive Debt in the Age of AI.” Open Data Science (ODSC). Aug. 25, 2025. Accessed Oct. 11, 2025. https://opendatascience.com/your-brain-on-chatgpt-understanding-cognitive-debt-in-the-age-of-ai/.