As our reliance and addiction to AI escalates, there are some stories worth sharing.
Right now, one of the biggest frustrations among judges is AI: People representing themselves in court, without lawyers, are building their entire defences on AI outputs. And it doesn’t often end well.
If we rely entirely on AI outputs, it’s no longer only the AI that’s hallucinating.
That’s like discussing a retail fit-out with a colleague, and they start mixing plasterboard and steel framing with React, Python and payment APIs in the same sentence.
The same thing is happening in every industry. To get a 10x efficiency gain, you need to guide the AI properly, and that starts with building a solid knowledge foundation in whatever you’re applying AI to.
That lets you write precise prompts and spot errors in what comes back.
After all, what AI is is an extremely advanced autocomplete. It gives you what it predicts that you want, based on very advanced mathematical models.
⚠️ Without the knowledge required, without you even knowing it, it’s easy to end up:
- Presenting an AI-built legal defence to a judge without ever speaking to a lawyer.
- Following a self-diagnosed health plan without consulting a doctor.
- Drawing up plans for your future house without an architect, which is probably already an expensive lesson for some people.
- Shipping technical solutions without technical foundations, exposing your app to critical security and data issues.
- And the list goes on.
🚀 If you want to build tech solutions efficiently and effectively, the fastest route is understanding what you’re actually building.