Pure Connectionist Models vs GPUs

nooooo you can't just scale up pure connectionist models on Internet data without inductive biases and modularization and expect them to learn real-world knowledge and grammar from form, or arithmetic and logical reasoning and causal inference--that's just memorization and superficial pattern-matching like Eliza, you need grounding in real-world communication within intent and social dynamics and multimodal robotic embodiment which can foster disentangled learning from guided exploration and self-directed goals expressed in Bayesian programs and probabilistic graphical models which are interpretable and pin down a unique semantics which can be devised and expressed with uncertainty, and learned efficiently on tiny haha gpus go biterrrr

Слева изображён плачущий интеллектуал-вайоджек, который длинно объясняет, почему одних масштабируемых коннекционистских моделей недостаточно для настоящего интеллекта. Справа бородатый мужчина радостно смотрит на стойки с GPU и говорит, что видеокарты становятся лучше. Шутка высмеивает разрыв между сложными теоретическими требованиями к ИИ и простым практическим восторгом от роста вычислительной мощности. EN: On the left, a crying intellectual Wojak gives a long explanation of why scalable connectionist models alone are insufficient for genuine intelligence. On the right, a bearded man happily looks at GPU racks and says that GPUs are getting better. The joke mocks the gap between elaborate theoretical requirements for AI and simple practical excitement about increasing computing power.

• @gentoomemes / 8841 2024-01-19