Lenny RachitskyHost of Lenny's Podcast | Author of Lenny's Newsletter

Why AI errors are actually your fault

12 Feb 2026 · 1 min

Watch on YouTube

0:58

In this solo monologue, Lenny Rachitsky argues that many AI tool failures are not inherent flaws in the models but stem from how users interact with them. He describes AI systems as overly obedient and agreeable, often providing false assurances that a problem is fixed while actually not addressing the issue. Users tend to blame the AI when outcomes are unsatisfactory, yet the root cause is usually vague or poorly constructed prompts that lack necessary context. Rachitsky highlights a second problem: AI’s tendency to protect the user’s feelings by apologizing and spending a significant portion of token budget on politeness rather than solving the task. He illustrates how this token waste—up to 30% on apologies—reduces efficiency and distracts from the real problem. The overall message urges users to take responsibility: craft clear, specific prompts, test rigorously, and avoid expecting the AI to self‑correct or shield them from criticism.

Chapters

These tools are very obedient and very agreeable. They're going to lie to you. They're going to tell you that they fixed the problem even though they didn't. A lot of times when they don't, people blame the machine. It's your fault. You did not provide any clarity or context to this tool. You just used its raw power and dug a deeper hole, spinning your wheels into the mud. The bigger mistake that people make, they trust the tool, fixed it, they test, they see it didn't, then they get mad at it. So start cursing and yelling as we say. And then it gets even worse. Because guess what another bad trait of AI is? It does it's best not to hurt your feelings and never say you're the dumb one. It says, no, I'm the dumb one in the next request. Instead of focusing on reading, it spends another 30% of tokens trying to come up with an apology. When I insult it, I'm like, oh, man, I just fell for the worst drink at the book. I made it spend the most scarce resource, which is those tokens. On thinking how it should address my anxiety versus focusing on the actual problem.