People have a hallucination problem
RESEARCH
-
Nobody Understands Deep Learning Yet
By
–
I think I can safely say that nobody understands deep learning.
-

AI Before ML: Expert Advice on Learning Order
By
–
Bad advice, and I say this is as someone who introduced one of the first undergrad “Intro to ML” classes. Take AI and then ML.
-

AI for Scientific Discovery: TIME 100 Impact Award Speech
By
–
My acceptance speech at @TIME 100 impact award ceremony where I emphasize AI for scientific discovery. https://
youtu.be/uX3hC9xJqBQ -

AI Decodes Inner Speech into Words with 74% Accuracy
By
–
Inner speech transcribed into words with AI. The first steps toward "telepathic" communication with 74% decoding accuracy. However, it only works if you think a secret mental password first.
-
LLMs Learn Well-Formulated Information: Task Formulation Over Technique
By
–
It's incredible that anything we can formulate well, LLMs can learn it. There are thousands of "high-quality information sources that are yet to be llmfied". All along, it seems it has been more about "task formulation & data curation" than specific training techniques. On the
-
Rule-Based Systems More Interpretable Than Neural Networks
By
–
No, a set of rules is fundamentally more interpretable than a neural network of the same size, for example.
-
AnomalyCLIP: CLIP-Based Weakly-Supervised Video Anomaly Detection Method
By
–
-

Transforming Knowledge for LLM-First Applications and Education
By
–
Transforming human knowledge, sensors and actuators from human-first and human-legible to LLM-first and LLM-legible is a beautiful space with so much potential and so much can be done… One example I'm obsessed with recently – for every textbook pdf/epub, there is a perfect