4 ms·
On the first hand, all communication is purely syntactic All formal communication is syntactic, but not all communication is formal, and not all experiences ar
by chongli 4y ago
On the first hand, all communication is purely syntactic
All formal communication is syntactic, but not all communication is formal, and not all experiences are acquired via communication.
Take the word “pizza”. If you’ve ever seen, smelled, touched, or tasted pizza, then you’ve acquired semantic information about pizza that could not possibly be acquired by syntax alone.
Same goes for informal communication such as eye contact, touch, hugs, kisses, handshakes, sex, etc. These are experiences people share that cannot be acquired through any other means.
Why does this matter for the AI debate? Suppose I ask a master pizza chef for tips to improve my pizza recipe. This chef has presumably many years of pizza-making and tasting experience. The chef can make recommendations to improve the flavour of the pizzas I make at home that are informed by semantic contents; actual knowledge of what effect the ingredients and methods have on the taste.
An AI can do none of that. It can only take the written words (syntactic inputs) of many pizza chefs and synthesize a model for pizza recipes. From that it can send me a bunch of words which, when followed, may or may not improve my pizza recipe.
It has no knowledge whatsoever, not least of which is experience of making and eating pizza. It cannot tell a good recipe from a bad one, unless its training set was specifically labeled to do that. Furthermore, models in general are highly vulnerable to adversarial inputs which would make it trivial for someone to construct a bad pizza recipe that gets misclassified as a good one.
Lastly, the model would be no help at all if I were to suggest a novel topping to add to the pizza. Whereas the pizza chef, if they had the experience of tasting that topping separately from pizza, would be able to imagine how it might taste on a pizza and make an informed recommendation. Furthermore, the pizza chef might even try that novel topping on a pizza and get back to me, something a language model could never do.
- MrScruff 4y agoThis is true. But if the AI was multimodal and we had some sensors that could measure the chemical composition of an object that it’s told is labelled ‘pizza’. Presumably that would count as semantic information?
- chongli 4y agoYes. But then it’s no longer an LLM. And it still runs into the issues of taste and preference which would have to be included in its training set. At some point we’re no longer talking about AIs, but androids that are built to simulate the complete range of human experiences. And this was Searle’s original argument. You can’t build intelligence by building strictly symbolic machines (computers). You have to endow them with semantic content by some other means.
- mcguire 4y agoI'm tempted to ask if you acquire semantics via nasal chemoreceptors or via absorption in the small intestine (and if the latter, I'm going to get a bottle of ketchup and take another pass at Kant). But I won't. I will note that your semantic conception of pizza is likely very different from mine, or from my friend's, who doesn't like cheese. You may have a semantic conception of pizza, but I have no way of knowing that.