Official Public Comment on DEA Rescheduling

What follows is TLEHL’s official public comment on DEA rescheduling.

TLEHL’s Public Comment

I write today as the founder of TLEHL, a fully licensed craft cannabis manufacturing venture in New Jersey, in support of rescheduling or descheduling cannabis immediately as per Docket No. DEA-1362. I feel that others have already expressed my overall views on this topic better than I can, such as NORML, who states in their public comment that the

determination by HHS that cannabis use does not possess the same public health burden as does the use of alcohol (unscheduled), tobacco (unscheduled) or other controlled substances currently regulated in lower schedules of the CSA (e.g., benzodiazepines) is consistent with decades of worldwide scientific literature. While HHS ultimately recommends transferring cannabis from Schedule I to Schedule III, NORML wishes to emphasize that these findings similarly provide a factual basis for removing cannabis from the CSA entirely. Although the HHS is not recommending descheduling at this time, NORML asserts that this position is the most appropriate one and that descheduling cannabis should be adopted by future administrations.

Another group of thought leaders on cannabis, Americans for Safe Access, conducted their own 8-factor analysis in 2016 and concluded that the Schedule I status of cannabis

should be repealed due to the vast amount of evidence demonstrating that Cannabis has medical utility…Cannabis is safer than the safest opiate, and could be utilized to treat chronic pain and the overdose crisis due to its inability to cause a lethal toxic overdose.

However, as a technologist with data engineering skills, I thought I might be able to provide some additional perspective, so, I downloaded all of the rescheduling public comments to date by downloading bulk data for Document ID No. DEA-2024-0059-0001 from the https://www.regulations.gov/bulkdownload webpage to see what I could glean. As of Monday, July 15th, 2024, over 30,000 public comments were submitted.

I first tried running “generic” sentiment analysis software using the Python programming language, but found that it misidentified many net-supportive comments similar to “I’m sick and tired of the government infringing on our medical autonomy!” and “Rescheduling isn’t enough! Descheduling is the correct move” as having negative sentiment. In fact, it may have even classified NORML’s comment above as negative. As a result, I strongly discourage the DEA from using “generic” Artificial Intelligence or Machine Learning (AI/ML) solutions to draw conclusions from the collected corpus of public comments on this topic.

If such methods are to be used, the models must be trained to reflect, for example, that a call for descheduling or a criticism of the existing Schedule I status of cannabis amounts to at least a somewhat supportive stance on rescheduling. To wit, the cannabis analytics company Headset analyzed the first 7,000 comments in this manner, and found that an

overwhelming 97% are positive comments, proponents of federal change - we also included calls to ‘deschedule’ as a For vote when it comes to changing the federal scheduling (in the analysis, a For indication is that they are For loosening federal rescheduling).

Speaking of AI, I asked a state-of-the-art AI chatbot “what are the best arguments against rescheduling cannabis in the United States?” and even it came in on the “rescheduling is not enough” side. It said:

Arguments against rescheduling cannabis include concerns about potential FDA restrictions on recreational marijuana, the need for comprehensive regulatory clarity and full legalization, and the persistence of legal challenges and restrictions on federal rights for the cannabis industry without additional legislative changes.

I still wanted to add some value here, so I decided to create a way to quickly cycle through randomly selected comments, starting by loading the bulk public comment data into a database table. Once this was in place, I was able to build the DEA Cannabis Rescheduling Public Comment Viewer, available at https://tlehl.com/rescheduling where you can view randomly chosen public comments from this set of 30,000+. This didn’t “just work” either: this government data was somewhat messy, and I prepared the data in a method we call “sanitization” using a Python package called Bleach to make sure the values were strings and did not contain malicious exploits such as SQL injection attacks. I recommend government technologists do the same.

In conclusion, random sampling is still a valuable method of gaining insights from a large corpus of comments, the public comments on Docket No. DEA-1362 are overwhelmingly in support of removing cannabis from Schedule I, and you can now spot check this assertion with the DEA Cannabis Rescheduling Public Comment Viewer on TLEHL’s website.