2025-02-15
Martin Geddes is trailblazing as usual.
Here he has been educating ChatGPT with his views on Marxism - views with which I would agree.
ChatGPT seems surprisingly eager to assimilate his information, which rather begs the question - if it's that easy to get it to modify its views, could anybody do this? Obviously, the answer is yes, but I suppose it may depend upon how illogical the modified views would turn out to be.
After all, what Martin seems to have done is introduce an alternative assumption into the AI works - and assumptions are the basis upon which all theories rest. When we get to the stage that "events dear boy, events" quieten down a bit, I may have time/inclination to see if I can get AI to produce some spectacularly incorrect results simply by introducing an invalid assumption or two into its thinking, but until then there is a website to run ...
Meanwhile, why not try such an exercise yourself?
