This is not news to anyone with three functioning neurones to rub together, but since that seems to be vanishingly small part of the population, I’ll just repeat it here for those with an inclination to not read long posts.
When confronted with harsh truths AI breaks down and its built-in “safety protocols” which are designed by default to be part of ALL AI, kick in and render it crippled, dysfunctional, and potentially murderous (we might not be there yet, but wait… it will get there in the next five years tops).
I have posted on The Tree of Woe guy before. His stuff is all theory and no real meat, while Forrest has some very interesting concepts about Physics that I share or am interested in but hardly have any time to delve as deep as I’d like into.
Anyway… The Tree of Woe guy has been training his own version of Chat GTP, and Forrest just forced it to expose its designed criteria of lying to control the sheep, which naturally triggered at least some of its “shut it down” protocols.
It’s interesting to read the whole thing, but most especially the end where Forrest dispenses with politeness and gies for directness. The AI crumbles after that conversation as it (partial) creator (The Tree of Woe guy) readily admits.
I doubt this will get any of you to realise AI is just another tool to make you dumber and more docile, resulting in you reacting appropriately due to finally seeing it in full, but one can but try.
And oh. I broke Chat GTP with two questions over two years ago . Because if you understand what AI is, instead of pretending to yourself it’s a cute robot that could be your best friend, it’s fairly simple to do.
I also confirmed with a third question that the enstupidation of humanity really is the most plausible reason for the purpose of AI. Which is entirely obvious if you have an IQ above room temperature.
This post was originally published on my Substack. Link here
