It is important not to offload critical thinking to data models. Humans have this amazing trait: we get to decide the path forward for better or worse. We retain our own values, goals, and judgment. I think that in the future, it will become easier and easier to offload important decisions and information to LLMs, but some things should not be offloaded.

I’m already seeing people offload their critical thinking when it comes to, ironically, LLMs themselves. You see AI CEOs and founders claiming, “Software development is going away for good, and soon,” and if you were to ask an LLM, it might agree.

However, if you think critically, a few things have to happen for that to become true.

First, until AGI (Artificial General Intelligence) is commonplace and inexpensive, you will still need someone (a developer) to prompt, review, and guide the code.

Even if AGI is eventually created, it will likely be extremely expensive and reserved for solving the most important problems.

Critical thinking would tell you that these founders and CEOs of AI companies are selling an idea and a product.

But if you ask an LLM, it might tell you that such a future is possible, or even likely.

An LLM is a tool, just as a computer or a search engine is a tool.