I agree with you. But do you remember when we could chat with a colleague about how much of a doofus our boss was without worrying that some automated system would possibly notify the boss about it? Nowadays big brother is always watching.
Maybe with very fast models you could request animation frames, e.g., frame 1) right foot at 12, left foot at 6; frame 2) right foot at 3, left foot at 9, etc.?
And instead of reporting tps, you would - of course! - report pfps (pelican frames per second).
I think it depends on what a memory system includes. Those that automatically inject relevant information into context are in my experience better than just md docs because agents often ignore, forget to read, or don't read md files in full.
It isn't ideal, but I am starting to write code (with AI tab completions) while waiting for LLMs. The tab completions are sometimes overeager and I wish I had more control over them, but at least I am not staring at "Thinking" all day. Having said that sometimes you have to monitor AI because, e.g., AGY CLI, often goes off the rails completely, including writing code outside of the "sandbox."
Thank you for explaining. Do you think there are still opportunities for stack optimizations to meaningfully speed up inference on single consumer-grade GPUs?
Or do you want it to speak to you too? I think this would have to be TTS on your phone. You can have ChatGPT speak to you but I don't see that feature in Codex.
Very nice! Would you have any recommendations for the leanest compatible "host"? Instead of adding this to my VSCode, I would rather use it as a separate app. Currently, I use a naked Zed install for Markdown because it launches faster than my system apps (and than Cursor, VSCode, etc.).
I like the idea. The demo music is very cool. The problem with using tone.js for such projects is, as I have learned the hard way, that it is hard to avoid issues like clipping; and AI agents are surprisingly bad at optimizing such apps.
I test nearly every new model that I can access on building a modular groovebox and they keep making the same mistakes. Even a 4-channel 16-step with a playhead can give them trouble.
> I spend more time refinding prd , epics and stories than I do elbows deep in code.
And do you enjoy this more than writing code? I used to look forward to writing code, solving these little optimization puzzles, learning, and staying sharp. Working with agents is dreadful in comparison. They lie, rarely learn, and I feel like a proctor.
Sure, you sometimes get to see something amazing, but usually I am just very annoyed by their performance and ever-changing but never-ending billing issues. First, with Claude Code, now with Codex, which was fine for a minute, but now I am out of tokens for the majority of time. (I don't have the income for those Pro INTx plans.)
Realistically, probably ads, but maybe not only that?
(AI start-up idea: one of our ads a day keeps dementia away! /s)