Really interesting exercise, Kaiser. A lot of my mental framing of how to build for an LLM and general AI-powered world has been about how LLMs allow unstructured information to be structured for consumption for and by AI agents, allowing for any enterprise to start to resemble an automated trading company.
Really interesting exercise, Kaiser. A lot of my mental framing of how to build for an LLM and general AI-powered world has been about how LLMs allow unstructured information to be structured for consumption for and by AI agents, allowing for any enterprise to start to resemble an automated trading company.
I hadn't considered the possibility of LLMs being used the way you're using them here, where you seem to have created an entire editorial organization around yourself using AI.
Thanks, Rodney — and thanks for the subscription, too! So weird: I was just thinking about you — it's a beautiful day here in Chapel Hill, and so I went out back to shoot some arrows and was thinking how bummed I was to have missed the last CAP down at the bamboo farm in Odum. This is actually just the tip of the iceberg; I've used LLMs to do all manner of steel-manning and stress-testing ideas of mine, and it's actually gotten me into the habit of examining arguments I advance for likely lines of attack and trying to preempt them. There's always a bit of nail-biting when I hit the button after uploading something to see what vulnerabilities — fallacies, indefensible assumptions, bridges-too-far — it's able to detect. Try it some time. It's really pretty fun.
Really interesting exercise, Kaiser. A lot of my mental framing of how to build for an LLM and general AI-powered world has been about how LLMs allow unstructured information to be structured for consumption for and by AI agents, allowing for any enterprise to start to resemble an automated trading company.
I hadn't considered the possibility of LLMs being used the way you're using them here, where you seem to have created an entire editorial organization around yourself using AI.
Thanks, Rodney — and thanks for the subscription, too! So weird: I was just thinking about you — it's a beautiful day here in Chapel Hill, and so I went out back to shoot some arrows and was thinking how bummed I was to have missed the last CAP down at the bamboo farm in Odum. This is actually just the tip of the iceberg; I've used LLMs to do all manner of steel-manning and stress-testing ideas of mine, and it's actually gotten me into the habit of examining arguments I advance for likely lines of attack and trying to preempt them. There's always a bit of nail-biting when I hit the button after uploading something to see what vulnerabilities — fallacies, indefensible assumptions, bridges-too-far — it's able to detect. Try it some time. It's really pretty fun.