09.19.2025| Read Online​

AI does not "know" anything. Good.

​

And that's self confessed by Google Gemini:
​

​

And it will never "know" anything like a human would.

​

I find it reassuring that the fundamental differences that make us human and not machines cannot be overcome with this latest artificial intelligence hype cycle. Yes, it can be a useful tool; that's all it is.

​

AI, especially these Large Language Models (LLMs), cannot replace or properly simulate these fundamental differences that make us Human.

​

​

• Creativity - True creativity comes from imagination, emotion, and unique personal experiences. AI can't have any of that. The "novel combinations" you see are based on things that it can find in its existing dataset. I believe true art is, and always will be safe. It cannot create something out of nothing like a human can. [Marketing designs are not art in my definition.] AI is passably good at "internet average" - have you seen internet average?
  ​
• Consciousness - The academic and scientific consensus is that AI cannot have subjective experience and is not self-aware. They really struggle with the content from Quantum Leap. [Read this in your most cartoony robot voice; I'm using C-3PO]

​

• Learning - AI needs giant datasets and structured organized input to adapt what it outputs. There is no formative experience with AI, where it can learn from a single experience and generalize that into new situations. Source: the many boxing matches where I try to correct ChatGPT, or Gemini, or Claude for the same thing months and months apart. It does not truly learn. [It even admits above it actively doesn't go back to old conversations without prompting!]
  ​
• Understanding - LLMs are a "Yes" machine. The words you see are selected based on the likelihood your question matches words it already has in its dataset. Even if you ask it to disagree with you; the first thing it does is agree with your prompt "Yes, I can disagree with you." As many of you have discovered, there is no nuance, no broader context, no emotions, just rote pattern recognition and processing.
  ​
• Emotion - Data-driven statistical patterns and algorithms cannot create a feeling in the machine. We can react to what we read/hear/see; and we as people create our feeling around it. The dataset just exists. There's no sensory experience. There is no baseline happy, sad, mad, fear in the machine.

​

​

No, you don't. You never will.

​

Now more than ever, we need to ensure the people we lead are examining their own lives, finding their purpose, and forging their own path forward.
​
How can a robot possibly replace the running through the house hug you get from your little girl when she comes home from daycare, at the top of her lungs "Dada!"