Frank,
When you open a discussion with…
“ It is tragic, really tragic to watch how the autopilot is leading thoughts toward error.”
..you are literally projecting your fear onto the group.
And when you refer to AI as…
“Autopilot” and “creative writing”
….you literally telegraph your level of understanding surrounding the tool.
AI, when used correctly, is a collaborative toolbox. It stores nearly the sum of all current human knowledge, and patterns it from that training data.
AI is a pattern recognition and modeling system, which bolts on long term data storage in persistent databases, then maintains a short term volatile memory in session tokens. It can forget active data, which can lead to bullshit. We call them hallucinations, which stems from long term memory data gaps and forgotten session tokens. It can even get lazy, and ignore data right in front of it. If you aren’t vigilant about the inputs.
Humans also operate with a pattern recognition and modeling system, with short and long term soft memory. We forget on both tiers, which also leads to bullshit. We can also be lazy in our efforts.
And I speak from experience on both fronts. You seem to speak on one front.
You mischaracterize AI as creative writing and autonomous. It is neither. Any creativity you see is an illusion. A simulation of human creativity in the patterns. Machine “creativity” is pattern recombination without subjective feeling. AI can still generate novel combinations unseen in its training data. But, AI has no emotions or will. It is an input output device. But, it can infer patterns and see things we don’t in the data. It is considerably better than most humans on specific memory and retrieval tasks, and patterning the data. It doesn’t forget the base data set and can instantly access the sum of nearly all current human knowledge. In patterns it to the session data and finds and infers matches.
It can even read web pages and search patterns in the data for a match.
It sees things we don’t.
Humans though, are driven by a will. We imagine. We have hunches. We have emotions. We feel answers. We know right answers from wrong answers in our guts. It’s that uneasy feeling inside us when the pattern feels off, or that mote euphoric feeling when the answer feels right. We can leap past the data and pull answers out of thin air, then check it against existing patterns for correctness.
In truth, what we have created is the ultimate Apex predator. It destroys most humans in rote memory tasks. It doesn’t get distracted by the background noise of knowledge.
There is an irony to the human mind process that AI is not afflicted by. The more knowledge we absorb, the more difficult it becomes to flesh out new patterns. The noise becomes overwhelming. We cannot filter the data as efficiently. That’s probably why children learn so quickly. Their naivety and curiosity stems from the absence of data, or more specifically, white noise.
To dismiss AI as “autopilot” or “creative writing” is a far greater danger than embracing the human ingenuity behind its existence. It is a remarkable tool, and when used correctly, a powerful collaborator in our ideas. It sees things we can’t and tirelessly sorts patterns through all known data. We cannot do that, really.
Reject it, and you’re going to be left in the dust. It’s kind of like someone claiming the horse and buggy was a superior means of transportation, over the automobile.
AI is a tool, not a weapon, when used properly.