Hooray awesome! 😀 And thanks for the O3 price drop too – super cool.
@jeremyphoward
-
ASI Development Speed and Exponential Capability Growth Discussion
By
–
I think jacks “compounding exponentials” wording rather suggested ASI was coming fast – doesn’t sound like bicycles 😉 I don’t feel like your notes reflect a compounding exponential growth in capability
-
DeepLizard Course Partnership for AI Learning
By
–
The deeplizard course is a great partner for our course 🙂
-
Customer Control and AI Model Deployment Decisions
By
–
I'm just a customer. I have no way to influence how these models are served to me.
-
Learning Diffusion Principles: FastDiffusion Course and Research Papers
By
–
Our fastdiffusion course is a good way to learn diffusion principles more generally, which I guess is a good place to start. Other than that, reading the papers is the only option afaik.
-
Service Provider Model Speed and Hardware Improvement Considerations
By
–
Nothing so sophisticated – this is simply showing how long the current service providers take to serve their models. As a customer, I like things to be faster. Perhaps in the future they might get better hardware and make them faster, who knows…
-
Why Faster AI Models Matter for Practical Applications
By
–
Because it's useful to have models that are faster.
-
Fast AI Models: Optimus Alpha and Grok 3 Beat R1
By
–
BTW on the next page, there's some interesting fast-yet-decent models, including the mysterious Optimus Alpha (is this the upcoming OpenAI open source model?) and Grok 3 Beta, which are both super fast! (~40x faster than R1!) (GPT 4.5 tldr: slow, very expensive, & wrong.)
-
DeepSeek Speed vs Opus Performance: Claude Code Model Comparison
By
–
Adding time spent to the results is pretty interesting. As good as it is, deepseek is *really* slow (we had to stop using it at http://
Answer.AI for this reason). Opus is fast, which makes it a great Claude Code option if you can afford it. Gemini 03-25 will be missed -
Random Forests for Reliable Feature Importance Measurement
By
–
Random forests are pretty much impossible to mess up. That's important for measuring feature importance, because you want your measurements to be reliable.