I got this text from a colleague yesterday that spawned a whole lot of thinking last night: “Everyone in federal policy land is basically using AI as their reason and rationale for more investment in every type of training, regardless of whether it’s directly relevant.”
I understand the skepticism. But I want to push back on the premise, because I think AI is always going to be relevant. Always.
We ran into this with the personal computer. Early adopters shoehorned it into “computer science” roles while the rest of the workforce deployed no PC skills training. Then we hit a tipping point and PCs were everywhere, in every job, every sector, every task. AI is on the same trajectory, and it is moving much faster. The question is not whether AI belongs in a given field. The question is how it shows up and whether we are building the domain-specific literacy to use it well.
That’s where I see education making two contradictory mistakes simultaneously, and it’s worth naming both.
The first is structural. Institutions are routing AI into Computer and Information Science programs as if CIS is the appropriate home for it. It is not. AI is not a discipline. It is infrastructure. It belongs in every classroom, every major, every technical program, the same way reading, writing, and math do. What that actually requires is domain-specific AI curriculum: how AI is used in biomanufacturing, in health sciences, in environmental monitoring, in advanced manufacturing. That content does not exist at scale yet, and we needed it yesterday.
The second mistake is happening in the same buildings, often in the same week. Administrators are convening working groups on how to detect AI use, build gates around it, and prevent students from using it to complete assignments. I understand the instinct. But that horse is out of the barn, folks, and no policy is going to put it back. Students are using AI. Professionals are using AI. Employers expect workers who can use AI. Every hour spent trying to enforce prohibition is an hour not spent building fluency.
These two positions cannot coexist. You cannot simultaneously treat AI as a CIS elective and a threat to academic integrity. Pick a lane. The right lane is integration.
At BCSI, we are building toward an AI in biotech credential series grounded in the same performance-based, employer-aligned model as our existing microcredentials. Not AI in the abstract. AI as it actually functions in a bioscience lab: data interpretation, process monitoring and optimization, documentation support, troubleshooting. Skills that can be demonstrated, assessed, and verified. But this work shouldn’t be done in isolation. If you are developing curriculum for bioscience, biomanufacturing, or adjacent technical programs and you are trying to figure out how to integrate AI in a way that is honest, practical, and employer-meaningful, get in my DMs.
The window for competitive advantage is real, but it is not permanent. Students who enter the bioscience workforce in the next ten years with demonstrated, domain-specific AI fluency will have a material edge over peers who do not. Employers will notice, and it will accelerate hiring decisions. But that window closes. The same thing happened with the personal computer: for a decade, proficiency was a differentiator. Then it became a baseline expectation so fundamental it stopped appearing on resumes. AI will follow the same arc. The students who move now do not just get a head start. They get to define what competent looks like before the field sets the bar for everyone else.