Sublime
An inspiration engine for ideas
I gave a talk in 2018 on the incompatibility of technological progress and privacy at the @Stacks conference in Berlin. Since then the advancements in AI have made this argument ever more relevant. The talk is short (15 minutes)
in the satirical Pretraining on the Test Set Is All You Need paper, the author trains a small 1M parameter LLM that outperforms all other models, including the 1.3B phi-1.5 model. This is achieved by training the model on all downstream academic benchmarks.
Sebastian Raschka • Ahead of AI #12: LLM Businesses and Busyness
More than a century ago, the father of standardization, Frederick Taylor, argued that standardization was necessary because people were cheaper and easier to rearrange than machines. But a century later, machines are cheaper and easier to rearrange than people.
Todd Rose • Dark Horse: Achieving Success Through the Pursuit of Fulfillment
With the release of Codex, however, we had the first culture clash that was beyond saving: those who really believed in the safety mission were horrified that OAI was releasing a powerful LLM that they weren't 100% sure was safe. The company split, and Anthropic was born.
Jeremy Howardtwitter.comHow AI Coding Agents Will Change Your Job
youtube.comdeep learning, is fabulous at