Six Lines
AI Is Great at Taking the Blame
It’s pretty clear that we’ve reached the point where the one thing that AI is actually the best at is taking the blame for things that aren’t really its fault. Examples abound. Here are a few.
If you’re running a company into the ground and you are going to have to do something drastic that may have negative repercussions, then why not reframe those actions as a result of a rapidly changing technology that has revolutionized your business? What do you have to lose? I think the best example in this space is Jack Dorsey’s massive, 40% layoffs at Block.
If you’re trying to understand why the new college graduates are having trouble getting a job, then why not gin up a story about AI overlords destroying our future? This sells. This gets clicks. But a boring story about the pervasive disruption resulting from tariffs killing supply chains and from the uncertainty of an economy run by a geriatric toddler who’s bankrupted six businesses won’t. Seventy-seven million people don’t want to hear how their support of a racist with a penchant for sexual assault wasn’t actually justifiable based on his ability to run an economy. Many of these people are small business owners who have taken it on the chin because of Trump’s policies and they still don’t want to recognize the truth. Talking about economic uncertainty and tariffs is a boring story that’s tedious to research and requires a basic understanding of macro economics to understand, whereas the AI story only requires having seen several popular movies in the last decade or so.
If you’ve just bombed a school and killed over 175 girls under the age of 12, then why not blame a new-fangled AI-based targeting system? After all, killing children is horrific and everyone knows new technology breaks all the time. Why go out of your way to admit that you’ve made a catastrophic mistake when there is such a good excuse available?
Of course, not every example is obvious. The data centers used to power AI do actually pose real environmental challenges. But those challenges aren’t fundamentally new, and AI cannot justifiably take the blame for decades of poor environmental policy. Deepfakes aren’t created without a human prompting the AI, but AI safety systems can make generating deepfakes much more difficult. These nuances often end up as finger-pointing exercises in the inevitable lawsuits that follow.
Even with those less-clear scenarios, AI is great at taking the blame.