AI scientists and engineers need to make their voices heard when it comes to non-competes. I hope this becomes a topic of discussion at @icmlconf and other venues. We urgently need more activism in this area. I believe non-competes should be banned because: 1. They give
@nandodf
-
West Must Lead Responsibly in AI to Preserve Freedoms
By
–
Sir Alex Younger is exceptionally bright and has a deep knowledge of geopolitics. One of the smartest people I’ve ever met. From him and Marc Andreessen the message is clear: The West has to cooperate and *lead* responsibly and decisively in AI to preserve the freedoms and… https://t.co/peREDqlxOx
— Nando de Freitas (@NandoDF) 5 juillet 2024Sir Alex Younger is exceptionally bright and has a deep knowledge of geopolitics. One of the smartest people I’ve ever met. From him and Marc Andreessen the message is clear: The West has to cooperate and *lead* responsibly and decisively in AI to preserve the freedoms and
-
Scaling Video Captioning: Practical Challenges Beyond Information Theory
By
–
I think focusing on a simple concept of information theory is missing the point. The issue here is more practical. If you need detailed captions for 5 billion videos, it would take you a huge amount of effort, incentives, and money to get good data from humans. If you google alt
-
LLM Captions for Text-to-Image Diffusion Model Training
By
–
I took the liberty of asking Copilot, and it answered the following: “Let’s explore the interplay of data processing inequality, rate distortion, information bottlenecks, and representation in the context of using synthetic captions from LLMs to train a text-to-image diffusion
-
Language Models Image Captioning with LLM Evaluation Filtering
By
–
And this is what a typical LM does for captioning the same image. It’s not perfect but it is more scalable and detailed. With some LLM eval filtering would look even better.
-
Human-Generated Image Captions for AI Training Data
By
–
See for example the captions humans have generated for images on the web (alt text): eg https://
huggingface.co/datasets/googl
e-research-datasets/conceptual_captions?row=61
… and let’s look at an example 1/3 -
Synthetic Data Superiority in Generative AI Models
By
–
It amazes me that so many (even tech) people still don't get the transformative power of generative AI. The best text to image and video models in existence today all use synthetically generated captions (see e.g. Dalle 3 paper). Human generated data is substantially inferior.
-
Managing Cognitive Load in Software Development
By
–
All managers and executives should pay attention to this brilliant plot. Getting in the coding zone is not easy. When coding one has in memory an overview of the codebase, of what exists, of things one needs to get back to and change, of what needs to be done, etc. An
-
Balancing Artist Protection and AI Innovation in Creative Industries
By
–
This is a complex and nuanced dilemma. On the one hand we want to protect the artists of yesterday and tomorrow (and of course the music and video corporations will make most of the money), and on the other hand we want to protect the artists of tomorrow and give them new tools… https://t.co/DQCiDxmUpf
— Nando de Freitas (@NandoDF) 25 juin 2024This is a complex and nuanced dilemma. On the one hand we want to protect the artists of yesterday and tomorrow (and of course the music and video corporations will make most of the money), and on the other hand we want to protect the artists of tomorrow and give them new tools
