AI is evolving fast. There is growing co-emergence optimism and fear amongst legacy tech and tech adjacent businesses (and the broader world) about what AI will enable and displace. My thoughts evolve but so does the technology. I’ve compiled a few helpful resources that helped me gain better understanding about this paradigm shifting domain.
1. Intro to LLM (https://youtu.be/zjkBMFhNj_g?si=cf-Kffa4FAI8_OSb) by Andrej Karpathy; great overview on the general anatomy of LLM, What it can do / what it cannot do yet (from a systemic perspective), the shadow operating system created by the LLMs, cat-and-mouse LLM security concerns
2. Deep Learning Evolution Blog post by Richard Ngo; Slightly dated blog post by Richard, who has a great deal of ethos in the AI research space. He talks a lot about the evolution curve of deep learning / self improvement capability of AI and what it can mean to the broader white collar / research fields. As a broad stroke, he makes some great high level points. However, I don’t agree with everything he says (i.e. understanding the capability without industry nuance typically creates a bias towards over-indexing the tech dominance vs. the incumbents). For example, it’s reductive to say ChatGPT passing a bar exam means that it has the necessary skills to actually practice law as a whole (Yes, it may get there but it requires drawing of an extrapolative linear (or exponential) progression to contemplate such reality).
can taemin read!? 