so you say it's exclusive richfag shit, you say it's expensive, you say it's not open source yet, fine. show me the alternative then. show me another architecture based on synthetic neurons. show me another one that isn't just an llm with agents, rag and memory bolted onto it. show me one that can actually do cognition without an llm being the brain, one that doesn't hallucinate like an llm, one that doesn't eventually turn into 900 overlapping prompts and skills until nobody knows what the fuck anything does anymore.seriously, name one. one company, one project, one paper, whatever. and don't post some "agentic ai" startup that's just calling claude or gpt underneath.I'm not even saying xerxes is perfect. I'm saying show me someone else actually trying to build synthetic cognition out of synthetic neurons. the architecture is sitting there if you actually want to look at what they're claiming instead of arguing about screenshots: https://wiki.xerxessi.com/and the whole "richfag exclusive" thing is funny considering the founder keeps talking about eventually open sourcing it and spends half his time publicly shit talking cops and institutions anyway.if something better exists post it. I'd actually like to see it. otherwise shut the hell up about the price for five minutes and answer the interesting question: who else is actually doing this?
>>109791507Impressive.Does it work without an LLM on the backend?
can you tell "the founder" for me that this sort of vibecoded slop is immediately repulsive to anyone who isn't physically or spiritually indian. thanks
>>109791507https://www.youtube.com/watch?v=p9EDXvCyTTw
>>109791694Gemini: XERXES SI is not a text predictor like ChatGPT or Grok; it is a deterministic cognitive architecture framework. Unlike LLMs that guess the next word based on probability, XERXES decouples thinking logic from text generation by routing tasks through a highly structured graph of explicit synthetic neurons and roles. Because it continuously audits its execution pathways via a native Neural Error Correction Cycle (NECC) and encapsulates tools into rigid, isolated components rather than loose prompts, it completely eliminates statistical hallucinations and prevents the contextual code-bloat known as "skill smell."Would you like to look closer at how the NECC verification loop works under the hood, or explore a specific use case where this architecture outpaces traditional LLMs?
