@CustomAIMath It’s not a claim. It’s tested by a doctor. Just like my skill I showed you is an actual screenshot with a date stamp while you claim I don’t know about it. And we already know your AI agents can’t do math right, so I don’t want your RSI.
@CustomAIMath Here’s a date stamp from February 8th on one of my skills which happens to be proto-recursive-self-improvement. So stfu about RSI. You didn’t even know what it was. You showed me a screenshot that you had to ask Claude!!
@CustomAIMath If you even read my SELF comment, you would see it’s written in the original languages that handle automatic trash collection. 🙄 Nobody said anything about multiple chats open. I’m talking about agentic AI. Actual AI agents. RSI is more modal than management.
@CustomAIMath There are multiple kinds of ingress and egress management depending on what pipes you’re using. Are we talking optical? Exfiltration? API endpoints? SIGINT?
@CustomAIMath I’ve been working on recursive self improvement extensively for at least two years now. Many of my custom skills are based on it. I also have a private repo called SELF. SELF Architecture is Semantic (LLM) + Erlang + LISP + FORTRAN hybrid computational framework
@CustomAIMath Regarding RSI and Egress Management - I have a DevOps agent named Scout (not the Docker one) and multiple cybersecurity custom skills that handle that. The number of agents vary based on which project. Some have 3, some have 5, some have 8. I’ve used swarms of up to 50.
@CustomAIMath You make a lot of assumptions. I run local models in addition to the frontier models and I’ve done my own fine-grain tuning. And written my own DSLs. Lots of custom skills, harnesses, custom AI agents, multi-agent coordination. Even a temporal linter. 😏
@CustomAIMath I’ve been formally IQ tested by a doctor and I have a 167 verbal-linguistic IQ. I’m all about pushing boundaries! But that doesn’t make your math right. 😉
@CustomAIMath@OpenAI You’re supposedly the spelling and language police (which started this whole thread) but you’re not even spelling multimodal correctly. Pretty much every frontier model in 2026 is multimodal. What kind of question is that?
@CustomAIMath I see that the information is there. I can read. I’m not saying it’s not there. I’m saying it’s mathematically incorrect. I’m not saying I don’t understand it. I’m not saying it’s because you’re not vetted. It’s literally wrong. Incorrect.
@CustomAIMath@OpenAI I have SaaS AI with a handshake protocol as well. And multiple GitHub repositories. And domain-specific languages. And quantum programming. I never specified a revenue amount (x). But you have a patreon link, so you’re obviously looking for money from your projects.
@CustomAIMath I’m clocking just fine. You’re not understanding that I’m saying parts of your equation are undefined which breaks the math and makes it invalid.