Ethan Ding on the economics of providing LLM services:
when you're spending time with an ai—whether coding, writing, or thinking—you always max out on quality. nobody opens claude and thinks, "you know what? let me use the shitty version to save my boss some money." we're cognitively greedy
In my view, people focus far too much on individual issues and far too little on the essential role of integrity in the political process. We should demand that politicians tell the truth. We should demand that politicians refrain from corrupt practices. We should demand
Cloudflare launched a feature that blocks AI bots from scraping content to address concerns over unauthorized use for AI training. An analysis of AI bot activity is available in the post, along with ways to report misbehaving bots. via The Cloudflare Blog
Simon Willison recommends checking out Anthropic's "multimodal" Jupyter notebooks to learn more about vision LLMs and how to use them effectively. via Simon Willison's Weblog