Recursive self-improvement isn't infinite, it can plateau and run into bottlenecks.
#38 | POSTED BY SITZKRIEG
That's absolutely true. It CAN.
And it will. Assuming no human intervention or bad actors getting involved.
While recursive self-improvement trapped inside a closed silicon loop inevitably plateaus, bringing humans into the loop changes the equation. Human-AI collaboration injects outside perspective, physical agency, and unstructured intuition into the system, directly tackling the core bottlenecks.
Because America will create the AI and the bad actors possibly in China or elsewhere will take it and push it to its limits.
When bad actors team up with AI, it creates a force multiplier effect that accelerates threats much faster than we can build defenses. Instead of humans helping AI bypass scientific bottlenecks for good, malicious users use AI to bypass the traditional bottlenecks of crime, cyber warfare, and disinformation.
Humans are why the Singularity cannot be stopped.
"money is measurable...".
It would take about 31,700 years to count to one trillion if you said one number every second without ever stopping to sleep, eat, or take a break.