80% of scenarios, 80% of the time
I'm starting to think generative AI is a proxy for revealing skills gaps. Regardless of field, if you are a high performing, critical thinking, strategic and effective individual then AI is a time saving augmentor of your existing capabilities.
Your high degree of skill lets you use AI tactically. You quickly iterate with prompts, better orchestrate, evaluate outputs, and immediately identify when you've been received a quality output. Alternatively, you will quickly revert to the traditional non-AI way of doing things if needed while offloading the more discreet and deterministic tasks to AI.
And all that is great... If you're great at what you do.
On the flip side, AI has also allowed for us all to easily fake that we are skilled in areas where we've not spent any time honing our craft. To our untrained eye, AI has given us passable-quality outputs that are in need of a trained eye to call out the risks, challenges, issues, or problems. At best, despite our lack of skill, we may sense something is off - but can't put our finger on it. And because we're relying more and more on AI validating AI, this is leading to normalizing mediocrity. In may cases speed is often more important than quality. Good enough is always better than too late: not all new apps need to have a unique brand and aesthetic. In fact a ubiquity in design, while boring, builds faster familiarity, reduces friction.
AI satisfies utilitarian goals: Ask AI to write a story with metaphors and it produces they are obvious, contrived and forced. It won't fully hit the the essence of what you have in mind. That may not matter if you aren't looking for nuance. Ask AI to write code, marketing materials, a business doc, a strategy, a presentation, an email and it all comes out good, maybe even great (to the trained eye). To the untrained eye, it looks complete.
That's the worry I have every time I turn to AI. It helps me in many ways and in areas that aren't my area of expertise. And it looks good enough that I may not spot the potential pitfalls. It let's me walk deeper into the forest than I could without it, without raising the question as to whether I should be going this far on my own.
As much as we have unlocked through AI we need to continuously remind ourselves to value the expertise of others. For our own disciplines, we need to continuously train and build upon our core expertise by broadening our experiences and relevant skillset. AI is raising the bar of expectations and will shine a brighter spotlight if you're underperforming. The fix isn't to avoid AI, or to use more AI to compensate. The fix is to get better at getting better and to continue to invest in your growth.
