Originally Posted by Matt Finley
I am always amused by the assuredness with which AI answers. Then you say, but if that’s true, then why is this the case. AI then apologizes, says how you are right to question the first answer, and around we go again.
Yes, I agree and have observed this over-confidence and then apologetic outputs from publically available LLMs many times.

But "However #1" is how many times have we experienced over-confident output from humans that when they are proven wrong do not apologize or even acknowledge their faulty thinking?

"However #2" : In the big scheme of things AI technology is still only an infant; we ain't seen nothin' yet.

We find ourselves in the rather comical situation of expecting large language models to respond fluently, intelligently, responsibly, instantly and accurately to all sorts of esoteric questions and demands that no individual human could ever hope to match. People get irritated when these [free to the user] AI systems fail to answer flawlessly while simultaneously debating wheather they are intelligent at all . . . comical.

Consider the fact that we are running out of intelligence tests that humans can pass reliably that AI models cannot. Think of the now quaint example of Gary Kasparov being beaten at chess by a relatively primitive computer by today’s standards. By those benchmarks, not only is AI intelligent, but I would argue that that there is no individual person on this forum or on the planet that can match the reasoning and general knowledge of today’s publicly available AI bots. And my prediction, based on the massive scaling that continues in this space [around the world] that soon an AI’s intelligence will exceed that of all humanity, both living and dead.

Personally, I see no intellectual hurdle to accept that our technologies exceed our innate capabilities:
In physical strength: think earth mover
In computation: think simple hand calculator or spreadsheet
In land speed: think automobile
In water speed: think speed boat
In air speed: think fighter jet
In underwater duration: think submarine
In space duration: think satelite
In dexterity: think surgical robots
In high temperature, low temperature, caustic/toxic environments, and on and on.

Why is intelligence, or even creativity any different?

To those that might disagree, are you game for going up against DeepMind’s AlphaProof LLM at the next International Mathematical Olympiad?
Or what about a friendly programming competition between you and ChatGPT? wink

https://arstechnica.com/ai/2025/11/deepminds-latest-an-ai-for-handling-mathematical-proofs/

https://soundcloud.com/user-646279677
BiaB 2025 Windows
For me there’s no better place in the band than to have one leg in the harmony world and the other in the percussive. Thank you Paul Tutmarc and Leo Fender.