Developer & technical writer exploring AI and robotics. Contributor @freeCodeCamp.
127.0.0.1
Joined May 2025
- Tweets42.7K
- Following491
- Followers13K
- Likes132K
Mari retweeted
For the first time in the history of AI model launches, OpenAI’s Astra launch video is so good that two months later, the entire industry still hasn’t figured out how to top it.
The level of creativity behind this video needs to be studied.
For the first time in the history of AI model launches, OpenAI’s Astra launch video is so good that two months later, the entire industry still hasn’t figured out how to top it.
The level of creativity behind this video needs to be studied.
Mari retweeted
Dear @elonmusk, I have an idea. Please hear me out.
People on X have been complaining about Grok Bot usage limits. Personally, I find it very helpful, especially the X integration, and I know other users do too.
You mentioned that serving costs were unsustainably high relative to revenue. So alongside paid top-ups, could we also look at making the existing allowance go further?
Anthropic recently made status updates cheaper for Claude Code’s helper agents running in the background. It reported that sessions using those agents consumed about 5% less of the usage limit.
Could similar improvements help Grok Bot users get more work done with their allowance?
The systems may work differently, but reducing background overhead seems worth exploring.
Dear @elonmusk, I have an idea. Please hear me out.
People on X have been complaining about Grok Bot usage limits. Personally, I find it very helpful, especially the X integration, and I know other users do too.
You mentioned that serving costs were unsustainably high relative to revenue. So alongside paid top-ups, could we also look at making the existing allowance go further?
Anthropic recently made status updates cheaper for Claude Code’s helper agents running in the background. It reported that sessions using those agents consumed about 5% less of the usage limit.
Could similar improvements help Grok Bot users get more work done with their allowance?
The systems may work differently, but reducing background overhead seems worth exploring.
Mari retweeted
Wait… how does he make sure a mistake by one agent doesn’t get passed down to all the others?
I've changed how I work over the last few weeks.
I now have one lead agent for each area of focus, running all the time, and a chief of staff agent that oversees them. Each lead manages its own sub-agents that write code, test it, chase reviews, and coordinate with team members.
Everything I need to decide or review comes into one prioritized queue, a simple local web app. Each product review includes screenshots and a video demo of the feature. I clear the queue in batches, mostly by voice and keyboard shortcuts. The leads also bring back ideas I hadn't considered.
Working from one queue, instead of switching between dozens of chats, has cut a lot of mental overhead. The system keeps learning my preferences and distilling my judgment into a "Brian brain," so its recommendations and first drafts get better before they reach me. The goal is fewer corrections per change.
Every change still goes through the same human review and security checks. It is eliminating tasks, not human judgment.
What an exciting time to push the frontier of productivity, especially as we build the economy for agents (agentic finance). We should all keep pushing in this direction.
Wait… how does he make sure a mistake by one agent doesn’t get passed down to all the others?
I've changed how I work over the last few weeks.
I now have one lead agent for each area of focus, running all the time, and a chief of staff agent that oversees them. Each lead manages its own sub-agents that write code, test it, chase reviews, and coordinate with team members.
Everything I need to decide or review comes into one prioritized queue, a simple local web app. Each product review includes screenshots and a video demo of the feature. I clear the queue in batches, mostly by voice and keyboard shortcuts. The leads also bring back ideas I hadn't considered.
Working from one queue, instead of switching between dozens of chats, has cut a lot of mental overhead. The system keeps learning my preferences and distilling my judgment into a "Brian brain," so its recommendations and first drafts get better before they reach me. The goal is fewer corrections per change.
Every change still goes through the same human review and security checks. It is eliminating tasks, not human judgment.
What an exciting time to push the frontier of productivity, especially as we build the economy for agents (agentic finance). We should all keep pushing in this direction.
While you’re here, I recently wrote an article on how to make your AI assistant more useful without connecting more tools.
Give it a read! Thank you! 🫶
