I was catching up on my #GenAI news for the day, and came across an article where a student failed an assignment because she used Grammarly to check her own creative work, and Turnitin flagged it as AI-created work.
This is not the first time this happens, but in a future not so distant, we could have one of the instances be the last. I’m removed from the evidence and the facts of this case, so my thoughts here are around the scenario, which sadly enough is not isolated. Here are my thoughts.
AI, and more specifically Generative AI (GenAI) applications are becoming a strong current in the business and academic application mainstream. If I were a product manger for SaaS products that are repeatedly in the middle of situations like this one, I’d be making a beeline (had to use Grammarly for the correct new spelling) to my counterparts in other app companies and discuss a way to solve the problem; a way to collaborate.
For instance -
Could an application attach a metadata tag in the content it creates to assert the form of origin? To identify edits on materials, indicating the level and nature of correction? Say for example a tag that could not be edited by the user, and that is automatically attached to the content based on the user’s actions.
Could the sending and receiving applications have a unique “handshake” that exchanges the materials and the metadata tags with acknowledgements on both ends such that there is positive control and a chain of custody?
Could the applications combine to warn the user that some content may be questioned by humans, or conversely, that the content has been certified “AI Free”?
The answer to all of the above is yes!
The practical reality is that work cycles and budgets are limited, and some of these features albeit possible and practical to implement, would not be as profitable as others.
These comments are no judgement on any app. They are actually missed opportunities that can be capitalized at all levels:
Positive user experience (Ops/CS)
Positive press (PR/Marketing)
Incremental revenue streams (Ops/Sales)
Additional channel sales (Sales/BD)
Sometimes these issues can be solved or mitigated by technology business consultants who beyond the implementation of a software application, focus first on the business/operational implications and on how to benefit the process, not mess it up.
Imagine how UNG or any other higher learning institution could benefit from avoiding the negative press proactively, by implementing well designed process improvement recommendations that could, ironically, use technology like AI to avoid the problem in the first place.
Again - a doable do, if we are willing to look for solutions.
I mean, if we can hit a rock hurtling through space at ludicrous speed, millions of miles away from earth’s position, a decade from today; we should be able to collaborate as product managers and technologists to solve problems like the above in “yesterday would be fine” fashion. And to state the obvious for those looking at the bottom line, yes - a solution that is responsibly priced for user to benefit, and for creator to profit.
Product managers and integrators; what do you think?
Link to the article here:
Commentaires