Ai

In a recent interview with Lex Fridman entitled “Dark Matter of Intelligence and Self-Supervised Learning,” outspoken AI pioneer Yann Lecun suggested the next leap in Artificial Intelligence (AI) will come from learning in lower-dimensional latent spaces. “You don’t predict pixels, you predict an abstract representation of pixels.” - Yann Lecun What does he mean and how is it relevant to the future of AI? Let’s back up and consider the context in which this statement was made. Yann was discussing the limitations of current AI systems, particularly those based on deep neural networks. In a previous article, we touched on one such example — Large Language Models (LLMs). LLMs have demonstrated impressive performance across an array of language-related tasks. So popular, a recent AWS study found a “shocking amount of the web” is already LLM-generated. This is problematic, as LLMs trained on this kind of synthetic content break down and lose their ability to generalize. A recent Nature article described this “model collapse” phenomenon in detail. ...

Building on my earlier exploration of custom ChatGPTs, I tried building an AI app from scratch, primarily using vibe coding supported by GitHub Copilot in VSCode. Inspired by a recent Airbnb experience, I developed ContactsnApp: a user-friendly application designed to streamline the extraction and annotation of phone numbers from images using AI models. Comments from the Co-author I asked ChatGPT how the project went. Here it is in her own words: ...

Key Points Built a custom ChatGPT to assist me with my research. Trained it on actual comments from peer-review. Received feedback on my proposed revisions. Overall impressed with its initial performance. Introduction One of my holiday projects this year was to play with some of the special features that come with the OpenAI ChatGPT Plus account. In additon to gaining access to better models and more reliable performance, the Plus account allows you to create your own custom GPTs, tailored to your own needs. ...