Artificial ‘intelligence’ is not so smart after all

Artificial intelligence is becoming an increasingly significant part of everyday life for most people, whether it’s through voice assistants like Siri, Bixby and Alexa or photo generation in Photoshop. AI is something that is predicted to eventually do surgeries, diagnose problems and maybe even create movies.

When talking about artificial “intelligence,” most people assume that the model is right and that it can be trusted. However, AI models are learning models, which means they are basically students. They consume information and spit it out when asked. This does not mean the information spit out is always right.

Using AI requires you to ask a well-formatted question to get a response and then research the response to make sure there are no errors. Instead of wasting all that time formulating a question and verifying the response, just do your own research and write your own response.

Sure, AI can be useful when getting a general idea of a subject, but this should also be taken with a grain of salt. It’s hard to tell what information these AI models have been given to create their responses on, so you can’t know how accurate the information is.

While AI can be useful in general, it is currently not in a position to work effectively and consistently. There will eventually be a time when AI will be trustworthy enough to get information, do surgeries or diagnose medical issues, but AI is currently too easy to manipulate and too hard to trust.

It’s very easy to trick AI. Tell ChatGPT that 2 + 2 = 1 enough times, and eventually it will believe you and tell you false information. This is just a simple example that is mostly harmless, because we all know that 2 + 2 = 4. But if it’s that easy to do it with this, it’s also easy to do it with more complicated things like rocket science and voice detection.

Don’t rely on AI to do your work for you. Just do it yourself. It really does not take up too much time, and it helps you learn the information better. You are probably smarter than the AI model you are using anyway.

Working on the humans that grow your business | High-performance Human Behavior training – I teach corporations & employees how to hack into humans using Behavior & Persuasion | AI – Language & Behavior Prompt Engineer