5 ms·
A vocal minority of researchers are essentially human chauvinists --- they "want to believe" that LLMs can't "really" perform this or that part of cognition eve
by quotemstr 2y ago
A vocal minority of researchers are essentially human chauvinists --- they "want to believe" that LLMs can't "really" perform this or that part of cognition even though the evidence is blinding that they can. (Anyone who genuinely believes that LLMs can't reason at all has never used an LLM.) These researchers start with their conclusion and work backwards to an argument, making their work seductive but useless.
- rurban 2y agoSo you are saying that LLM do can reasoning? Logical reasoning is something completely else than likelyhood in word completion. A pure LLM will never be able to do reasoning, you need a hybrid. Use the LLM for classification and completion and a logic system for reasoning
- incrudible 2y agoThe problem is in being able to discern reasoning from patterns that happen to exist in the training data. There are plenty of tricks you can play on an LLM by subverting the expectations it must necessarily have due to its training data. A human might fall into the same trap, but can then reason themselves out of it, whereas an LLM tends to double down on its mistake.