#Deep-Learning
3 notas · última:
Raphaël Millière on the Limits of Deep Learning and AI x-risk Skepticism
Raphaël Millière argues that current scaling results do not justify claims about human general intelligence, and that today’s models still fall short of broad generalization and human-level compositionality.
News Feature: What are the limits of deep learning?
PNAS profile of 2019 surveys where deep learning breaks: adversarial spoofing, data inefficiency, opacity, and the missing common sense that symbolic AI once tried to hand-code.
A Primer on Large Language Models and their Limitations
Johnson and Hyland-Wood's LLM primer argues hallucination is the wrong word: LLMs produce Frankfurtian bullshit, indifferent to truth, alongside catastrophic forgetting, model collapse, and jailbreak risk.