@bryan_johnson@bryan_johnson I’ve had Long Covid since 2020. After Mayo and years of searching and self advocacy, my labs pointed to immune dysfunction as the primary pathology driver. A protocol from Dr. Yong Zhao at Throne Biotech, first flagged to me by @TonyRobbins, changed my life.
@Shpigford Will you still use Claude Cowork MAX sub at 2000/mo? 20,000/mo? Don’t think for a second that $200/mo is going to last much longer. Pricing will continue to rise and token allowance will continue to drop. Local is the only way
My son is due in March.
When he arrives, his cord blood is going to be collected.
And if it works, that cord blood is going to help heal me from long COVID.
Here's what I'm doing and why. 🧵
Here's how it works:
Your white blood cells leave your body and pass through a device coated with cord blood stem cells.
They get re-educated. Reset to how they worked before disease.
Then only your cells go back in. The cord blood stem cells never enter you. (via @thronebio)
@ThroneBio If you have long COVID and you are looking into stem cell therapy, full blog post dropping soon on @LongCovidStore.
And if your family is expecting a baby: cord blood collection has to happen before birth. The window is narrow.
More to come.
@ThroneBio This isn't fringe. Dr. Zhao has a decade of peer-reviewed published work. Active trials on https://t.co/YjPTWBHezM.
The science is real. The LC application is still being established.
I'm one data point in that process. I'll share everything. Good and bad.
@XFreeze one person with a 4-agent swarm debating and fact-checking each other is now a more effective team than most 10-person startups
solo founder era just entered a completely different phase
@AlexFinn@IamEmily2050 What qwen version and quantization do you recommend for the Studio m3 Ultra 96gb? This will be my primary local model with one smaller model running prompt routing. Wish I had more headroom but this is what I am currently working with. Thanks!
I preface any brain dump with /prompt. That tells Claude Code to take my confused mess and to structure it in a way that will execute the ideas more coherently. If I type /prompt-refine I can iterate before it moves ahead.
Anthropic recently leaked the anatomy of the perfect prompt for Claude.
Your Claude outputs will never look the same after this.
Literally handed to us on a silver platter: