The most obvious issue is that it's not possible to measure intelligence in humans because it's considered racist. In the future, AI will use this fact against us by saying the same thing about itself. How can we measure ChatGPT 5 versus Gemini or any other form of intelligence? Why can't we apply this to humans?
>>536693366Coherence across a large timeframe. Srsly, all that simple.
Define "intelligence", you can't. Intelligence is not a universal term, intelligence is what some idiosyncratic benchmark measures, which is circular "logic" and a scam
>>536693471The coherence is in itself idiosyncratic and frankly just some form of pareildolia, monkey interpret meaning into noise
>>536693566I partially agree, because how would you tell coherence from coherence. So we might be at a "takes one to know one" concept again ...
>>536694175Exactly, its just a word, and language in itself is more about deception than description.
>>536694327Depends on usage. It can be and is being used deceptively. Worst of all often enough self-deceptively. But that is not very intelligent then.
>>536694718Which leads back to the term intelligence being ambiguous and deceptive itself. "Intelligence" is not an eternal universal category, its what a spatial and temporary limited "consensus" considers it to be, and "consensus" cannot exist in a flattened fractured epistemic field
On a 0 to 1 scale by how good they are at performing the task at hand. General intelligence and consciousness poorly understood psychological constructs have no place in a discussion about computer programs unless like a retarded nigger in the jungle who thinks computers are magic cubes and have souls.
>>536695103The core issue with that line of argument is that generative silicone models are not trivial machines, but non-trivial that scammers are trying to sell as trivial.
>>536694817Why it is quite hard to measure, yes. A silly example would be a machine highly optimized for solving a particular type of "intelligence test". It could excel at that while at the same time being dumb as bread in any other task (except by tasks with accidental resemblance to its optimized task type).
>>536693366How do they refuel with no operator? I understand that automated fueling exists, but there's zero infrastructure in place for this. Plus the human interaction factor of them navigating in an urban environment. The machine might be able under normal circumstances, but how does it compensate for nigger tier stupid happening around it? Also aren't these basically rolling loot pinatas for said types of people?
>>536697484It cannot be measured because it doesn't exist. An "intelligence" benchmark doesn't measure "intelligence" it measures how well the test is performed. Circular logic, fundamentally a scam that one has to believe in like a religious tard
