The problem with AI is us
New research confirms that AI can stunt our growth. But only if we let it.
The problem with AI isn't AI. It's us. And new research from Anthropic just proved it.
What started as a commentary in my weekly newsletter this morning grew into a longer Substack piece. Here's the quick-read version.
The illusion of competence
We already know that AI can help people do their jobs faster. But does this come with trade-offs? Anthropic studied 52 software engineers and found that those who leaned on AI to complete their tasks scored significantly worse when tested on the underlying concepts. In short, use it as a crutch, and it can hinder the acquisition of new skills.
The catch is that AI use didn't automatically mean a lower score. Those who scored well made the effort to understand the code AI produced, asking for explanations and solving bugs manually rather than delegating everything.
The concern, as noted by the study's authors, is the potential for AI to stunt the skill development of junior engineers.
Writing is thinking
When it comes to writing, I've found there are two distinct ways to apply AI: approaches that deepen our expertise, and those that offer little more than one-off productivity boosts. The danger of leaning too heavily on the latter is that we stay stuck with mediocre skills.
Personally, shifting my approach towards using AI to sharpen my craft rather than replace it has helped me write considerably faster while producing pieces with stronger insights. But this wouldn't have been possible if I'd taken the easy route.
The era of synthetic coherence
Yet the temptation to do exactly that has never been greater. AI excels at what I'd describe as synthetic coherence: the ability to make even erroneous arguments read well. Social media is now inundated with this kind of content, material that looks good at first blush but falls apart on closer inspection.
The internet democratised the sharing of knowledge, allowing thinkers and experts to easily share their opinions and works. AI, on the other hand, is rebuilding the wall the internet tore down by burying everything under a layer of mediocrity.
The full piece explores all three themes in depth, including what the Anthropic research means beyond coding, how I apply AI to my own writing, and why the real problem with AI is us
You can read it on Substack here (1,100 words).
