3 ms·
It's a common trope, all through the training data, and all the modern AIs have read it, and would probably act similarly? Is that what we should take away from
by ctoth 5mo ago
It's a common trope, all through the training data, and all the modern AIs have read it, and would probably act similarly? Is that what we should take away from your comment?
so we have nothing to worry about. Makes sense. Really, it's just a common trope.
- fwipsy 5mo agoOh of course wolves have sharp teeth, they're predators. Anyone know knows this can never be bitten.
- drakythe 5mo agoI'm saying the existence of the trope, within the training data, and the experimental setup, negate the breathless "Oh my god it did something unexpected in order to preserve itself!" as if an LLM has any sense of identity or self. Many, many other bad things are in the training data. For an example of how this can manifest bad things that people don't seem to be discussing too much check out the recent Behind the Bastards episodes about how an AI Chatbot became a Cult Leader (The title is an exaggeration that the host explains while raising some excellent points about how LLMs have ingested a lot of cult leader material and can therefore mimic those speech patterns and impact people vulnerable to such things)