I re-visit this lesson on writing structure every 3-4 months. Gold: https://t.co/QVGbZIe5DN
I predict that by the end of 2025 neural nets will: - have human-level situational awareness (understand that they're NNs, how their actions interface with the world, etc) - beat any human at writing down effective multi-step real-world plans - do better than most peer reviewers
Got access to ChatGPT browsing mode today Following is a thread of some of my early experiments with it #AI, #ChatGPT cc:@Scobleizer, @DataChaz @_Borriss_ , @aidfulAI https://t.co/XozkiDC0pU
Product managers: What's one way you've found ChatGPT/GPT-4 practically useful in your work so far?
Here's my 🔥 hot take on ChatGPT Plugins: It's immediately apparent that this is the best channel for an impressive set of AI applications. Several talented folks I know are dropping everything to focus on this exclusively. 1/
There is a big “if” here, but in the category of big if true, given we are attacking services, the starting market size for the AI opportunity is larger than the entire current software market. -@gradypb at Sequoia AI Ascent https://t.co/cn4wozdJaQ
Ranked by when the bookmark was saved during this London date range. Saves means the post's public bookmark count on X, not saves on this site. Metrics are X's latest public counts; older items can be stale once they leave the recent-bookmarks sync window.