Grok @bot can see multiple of my machines at once.
What it still can’t do: pin a shell to one host.
ExternalShell bounces between Macs and Windows. So “run this on all 3” used to mean chaos.
We built a workaround. 🧵
I was at 95% of my claude limit with 2 days left til reset, but my GPT account was barely used, so I used T3 Code by @theo with the new Orchestrator v2 to have the Claude session literally orchestrate a complete team of GPT 6 Astra and GPT 6.1 Sol agents back and forth do all its work for it on the project. Was amazing to see each session go back and forth and barely touch the claude usage.
Google needs better leadership in the AI world. They have so much potential and done so little because of red tape internally. @sundarpichai hand the reins to someone else please.
Excited to announce that all super intelligence organizations have now jointly agreed to the ultimate in AI/SI safety:
Moving all testing to Delta Airlines flights, where accessing the Internet is utterly impossible!
Grok is doing my bookkeeping and tax prep. When it has questions, I have it consult Claude and ChatGPT.
I check the work and approve everything.
An AI team doing the prep. This is WILD. 😂
The most important skills with AI agents are task delegation, and emotional self control.
These two things will let you build anything.
Just like leading human employees. Most people do not possess these skills.
The best and easiest AI Agent system you can install right now and use in your business is GrokBot.
The next best is ChatGPT app with Dots. These two apps together can run your entire business for you, and you won’t even have to spend more than $160/m to run both side by side!
You don’t even need a seperate computer anymore either!
Important note regarding Grok @Bot:
Going forward, @SpaceX will use the best back end model for any given task, including Claude Opus 5.5, MidJourney, Suno and other leading APIs.
Whatever is most likely to give you the best outcome.
I am surprised that people are surprised this is how I prompt Claude.
Talk to Claude the way you would a coworker. There's no secret to prompting. There's no need to be overly scaffolded or prescriptive for most tasks -- give Claude a goal, and it will figure it out.
Back in the Sonnet 3.5 days, your prompt mattered a lot. Nowadays, it's much more important to communicate to the model:
1. What you want it to do
2. How much effort you want it to spend
3. How it should verify that it did the right thing
A group of developers created clones of Adobe Photoshop, Premiere Pro, Lightroom, Illustrator, After Effects, Acrobat from scratch using AI - and their giving it away for free: getartcraft.com/apps
Opus 5.5 with sonnet 5.5 xhigh sub agents is incredibly efficient cuz I can deploy a swarm of those. And the. GPT 6.1 sol as additional code review agents via tmux that’s the play and I’m barely touching my usage
It's possible for tech layoffs to be both individually tragic and collectively bullish. Whether they're triggered by efficiencies or otherwise, they free up talent to tackle the next set of problems and opportunities. This is how the frontier keeps moving forward!
The AI smart enough to do the work for you. Let the agent do the work for you.
Stop over thinking it.
ChatGPT Dots is the smartest one for beginners to do real work with. Can literally do all your marketing for you.