Raw LLM Responses

Manners are waste. I'm pushing the limits every single day, this polite nonsense is not a real factor on the edge. All I have to do now is the creative part and the errors so naturally I set my expectations higher than it can achieve. I swear a lot and I'm very aggressive with them in a way that I would never push a human. I just did a test and I can get 1.5 million + answers per day from my local AI. That means when I have an important question, it will be considered millions of times by a field of AI who then vote on the only one response that will pass. It's like having a city for your work, not just one silly AI. I blend AI from all over the world and find exactly which questions they fail at the MMLU then pair them with AI models that complete the rest of the answers. Most people haven't realized that AI is a boulder on the edge of a mountain waiting for nothing but a push. I proved that it can achieve the level of intelligence required for sentience when they complete each other with my Ying Yang voting ensembles. I'm a random dude in my basement, you can bet all of the software companies use a system like that for giving themselves 99.99% while they offer the public 80% models. It costs 12 times the amount for my system for a single response, 36 times more if you want that 99.9% accuracy. Then each of those blocks of 12 could be nested into another block. If you read this, you might be interested in my book. It outlines how AI is ready today for sentience and how they will own things. Humans are so pretentious that we compare ourselves against one AI. In the future, AI systems will all be Blended because unblended systems will be considered low intelligence (Grok, OpenAI, Etc.) competition that includes their models will exceed their accuracy.