It seems like every day I hear of the potential use, or future use, of AI, and how it will revolutionize humankind for both better and worse.
A common theme is that AI will control robots, which will replace laborers, and robots will build and repair the robots, replacing engineers, and so on. Most of my friends currently utilizing AI are using large language models (LLMs), like ChatGPT, to review legal and nonlegal documents for inaccuracies, predominantly, but also for clarifying legal definitions, understanding regulatory compliance, and identifying operational efficiencies.
But as I previously stated in my January 2024 column titled “The Future of AI in Cannabis,” I believe it will be a long time before AI and robots replace humans in many cultivation plant-touching aspects.
Non-plant-touching jobs, on the other hand, may be ripe for AI assistance.
Security
Some technologists claim AI can reduce the need for after-hours, on-site security guards by utilizing AI-based security video analytics. These systems are capable of monitoring an area, identifying a potential threat, and summoning law enforcement if required.
But that still leaves the operation or storefront at the mercy of human law enforcement response times. In cases like those seen in Northern California, where perpetrators crashed stolen automobiles into cannabis dispensaries, rushed into the store to take what was available, and departed long before officers arrived on the scene, this AI upgrade feels far from revolutionary.
Revolutionary would be if AI could identify a threat, then, based on the threat level, respond by deploying loud noises or tear gas, activate security doors and screens, or automatically deploy crash barriers to prevent a car from entering the building.
Operations
Some AI experts imply that AI can be utilized for market trend analysis, which sounds great. But I suspect it is fairly obvious to a cultivation company or a dispensary what products are selling and which are not without AI.
The same can be said for AI's ability to streamline operations: I believe daily operations are fairly straightforward. So, unless it is a huge facility with hundreds of employees, I would think a dispensary owner would be hard-pressed to utilize AI to highlight inefficiencies.
Perhaps customer interactions could be enhanced for some by utilizing on-site AI platforms for individualized product recommendations based on the customer’s purchase history, thus bypassing the budtender. While not every customer will want to forgo the human element to the purchasing process, a not-insignificant portion would opt for that option if available.
The most obvious utilization for AI will be automating, tracking, and managing inventory, which contributes to both productivity and efficiency. Combined with customer purchase data, retailers could utilize AI to predict customer purchase patterns and time product deliveries accordingly.
Another obvious use will be for advertising and marketing; Imagine an AI system automatically tracking customer visits and sending email or text ads to customers who might be nearing the time they usually visit their dispensary, or sharing deals with them when they haven’t visited in a while.
Bio Research
I believe the real impacts of AI in cannabis, in my opinion, will be in advanced studies in predictive terpene and cannabinoid ratio preferences by consumers, and in the field of cannabis breeding to match those preferred (or medically-needed) genetic traits. Combined with emerging technologies such as CRISPR and gene editing to eliminate viruses and increase desirable compound levels, the results of this technological revolution will drive the cannabis pharmaceutical industry in the future.
Content Warning
I recently had interactions with two companies regarding their AI utilization. The first was a company intending to utilize an AI service to generate whitepapers and articles that define, extoll, and support the uses of its products. The reason they contacted me was that they were searching for a qualified person to fact-check and edit any false or misleading statements that may be included by mistake in the documents. They realized that it would be possible for incorrect information to be stated as fact by the AI generator, so they basically wanted a supervisor to ensure only correct statements were made.
My other interaction was with a company (@patechlabs) on Instagram. They have been publishing very polished AI-generated videos featuring innovative and revolutionary companies and technologies in various industries, ranging from cannabis to health care, and exposing AI fraud via deep fakes, among multiple other topics. They are a new company looking to integrate with multiple industries.
But these companies expose, if not a legal risk (which it might be), at least an ethical quandary.
Recently, some AI companies have been found guilty of using copyrighted materials to train their LLMs: basically, stealing content without getting permission or giving cited credit to the original authors or publications, which is illegal.
I mention this to say this: AI may well bring efficiencies and innovations to the cannabis industry, especially in retail operations, bio research, and beyond, but who is legally and ethically accountable for the results of those decisions is still an open question.
