"AI Safety" - A Misnomer at Best

Posted by Louis Jones • Aug 10, 2023 8:43:26 PM

There is much swirl and prognostication around the impact that AI will have on society.  Like most future trends, there is much track to be laid down before anyone can say with confidence what its true impact maybe - it is still very much being defined.  To the point of this Forbes article, there is a move afoot among the big tech firms to self-regulate (which also mean self-defining) what AI safety principles may look like in the face of mitigating AI risk.  But without data/privacy rules or regulations, it may be hard to put the appropriate boundaries around AI, because, at the end of the day, good AI practices rely on good data.
While AI, in its current state of discussion, revolves around large language models and the ability to generate “synthetic” results, AI in more practical terms has been around and in use in our businesses for some time.  AI has driven algorithms, optimized results/outcomes, and led to more intelligent pathways between marketers and customers, we have relied on AI principles at a more mundane operational level.  As the Generative AI gains rapid popularity and attention, there is a lot of talk about how it will change everything we do.  Please hold!  There are many lanes where this technology may indeed be impactful (and some, not so much), but those lanes have yet to be adequately vetted and defined.  New aspects of AI will reveal themselves in time, we must have the discipline and the wisdom to know when and where it can add value.
So with all the banter and the tech elite sitting with President Biden - thank you!  At the very least it establishes the importance of having a stated concern for the responsible use of AI technology that serves as a critical baseline for improvement.  It also sets the tech firms on a path to more progressive integration of AI into their different visions for its role in their products and applications as the threat of regulations loom in the distance.  For the time being, as marketers, we must learn to test, modify, and refine AI applications in a useful and gainful manner.