PsycheOS | Psyche Suite 🧠🧬 @PsycheOS

a new kind of engineer is showing up. they don't write prompts, they design how the AI works. it's called graph engineering, and the people who get it now are about to make everyone else look slow x.com/i/article/2080…

Ever heard of Regulatory Capture? The US government is NOT going to destroy the American AI Industry. The US government is owned by the American AI Industry who is ordering the US government and "Goons with Guns" to to kill off competition for the American AI Industry just like they did for automotive makers decades ago, which killed the American Automotive industry.

@ns123abc It sucks anyway so happy to hear it.

@Bagons1776 @ChinaSelect @zerohedge @amazon @RepMoolenaar Who cares? All our data goes to the NSA.

It would be like comparing a Japanese kei truck to a Scat Pack Challenger or Hellcat. They all have engines and wheels but the performance, memory b/w & ability to handle these workloads are in different leagues. Raspberry Pi 5 ~0.7–3 tok/s vs Jetson Orin Nano Super ~15 tok/s

Chips run the world but data defines reality. No matter how powerful or cost-effective the compute stack is, if the training data is biased, flawed, incomplete or intentionally misleading, then that "trillion-parameter" model on the sexy supercharged tech stack becomes a high-speed autobahn producing plausible falsehoods or distorted worldviews and gets outgamed by my 2005 Excel spreadsheet running VBA and vlookups (or hlookups if that's your cup of tea).

NVIDIA QUIETLY DROPPED A $249 BOX THAT REPLACES YOUR $200/MONTH OPENAI SUBSCRIPTION WITH $2 IN ELECTRICITY it's called the jetson orin nano super. smaller than a wallet, runs at 25 watts, does 70 trillion ai operations per second. runs llama 3, mistral, gemma and deepseek locally with no api fees and no data leaving your house a developer running automations and coding assistants pays $200 a month to openai. the same workload on this box costs $2 a month in electricity and breaks even in 10 weeks install ollama with one command. change one line in your code. point it at localhost instead of openai. everything else works identically 7 billion parameter models handle 80% of what people use chatgpt for. summarization, drafting, coding, document q&a, automation pipelines. total monthly cost drops from $200 to $22 cloud subscriptions keep getting more expensive and rate limits keep getting tighter. the people who set this up in 2025 are going to look very smart in 2027 bookmark this and read the article below

starmex @starmexxx

3 months ago

x.com/i/article/2058…

40 53 385 4.4M 2K

@fivepebbles__ @om_patel5 I enjoy watching Hylics struggle in the age of AI.

@AuroraMar1eL Many people knew this but the valley echo chamber and circle jerk is/was too powerful. The CIA AI Cartel thought they could push their bullshit that GPUmaxxing was the only moat that mattered. And then Qwen, Kimi, Deepseek and friends knocked their c0ck to their back pocket.

🚨BREAKING: HKUST just gave AI agents permanent memory that improves over time. No retraining required. Lessons from one model transfer to another. Up to 11 points better on the hardest benchmarks. > Every AI agent you use today starts each task completely blind. No memory of what worked last time. No memory of what failed. Every mistake gets repeated forever. > HKUST built XSKILL a dual memory system that accumulates two types of knowledge after every task: skills (what workflows to follow) and experiences (what specific mistakes to avoid). > The model itself never changes. The memory just gets smarter. > The part nobody expected: knowledge learned by Gemini transfers directly to GPT and o4 mini. No additional training. One model's lessons become another model's head start. → Up to 11.13 point improvement over the strongest baseline on hard benchmarks → Syntax errors cut nearly in half: from 20.3% to 11.4% after skills added → Cross-model transfer works: Gemini's knowledge improves GPT-5-mini and o4-mini → Zero parameter updates required at any point → Knowledge compounds: more tasks = smarter memory = better performance The fix is simple in principle. Skills stop the agent from wasting steps on errors it already made. Experiences tell it exactly which tool to pick in which situation. Together they turn a stateless agent into one that actually learns from its past. Every AI agent deployed today is leaving this on the table.

The entire RAG industry is about to get cooked. Researchers have built a new RAG approach that: - does not need a vector DB. - does not embed data. - involves no chunking. - performs no similarity search. It's called PageIndex. Instead of chunking your docs and stuffing them into pinecone, it builds a tree index and lets the LLM reason through it like a human reading a book. hit 98.7% on financebench. beats every vector RAG on the leaderboard. no embeddings. no chunking. no vector DB. 100% open source.

I’m unsure why consciousness would be exempt from scientific study. If it’s the "interface" or HUD/heads up display (dashboard), then it’s precisely a structured system that we can analyze, deconstruct and reconstruct. This allows us to learn what generates it, what shapes its contents and how it interfaces/relates to underlying mechanisms and systems (brain, regions of brain, psyche, etc.)

Maybe the issue is that we don’t force enough constraint/s on theories early on. If ideas are forced to be systemized, operationalized, and then required to produce falsifiable, discriminative predictions, this should stop these interchangeable interpretations (of the same intuitions/patterns sensed). And it may well be that "that" constraint is w hat generates a genuinely new structure/s by reducing error (eror reduction( in how we update or move between competing explanations.

Math + BCI make sense because they add constraints. But without clearer definitions of "consciousness" (and its dimensions/factors, thresholds, etc.), having even better data and formalisms may just have us see the same disagreements but in higher resolution and color.😀 In my view, the key is whether the theories can be made falsifiable in a way that actually separates them.

If all experience is already simulation (brain-generated), then "simulated vs real" collapses. What would appear to matter most is the structure and constraints of the simulation and factors/dimensions involved and converging for that to happen, rather than the substrate or label/s involved.

@pubity It just means you start a new company where you are not hiring people who are replaced by AI, you are an AI company where you are hiring some humans to assist the AI.

@fivepebbles__ @om_patel5 Keep crying harder - reality doesn't care about you. And neither do I.

Tencent has killed fine-tuning and RL with a $18 budget. Right now, if you want an AI agent to become an expert at a specific, complex real-world task, you have to use Reinforcement Learning. You let it try, fail, and update its internal parameters over and over again. This is the exact optimization technique (GRPO) that DeepSeek used to build their massive reasoning models. But there is a massive problem. Updating model weights is insanely expensive. It requires massive GPU clusters. And worst of all, when you train a model to be highly specialized at one thing, it often "overfits" and forgets how to be good at everything else. Tencent killed this bottleneck forever.. by building Training-Free GRPO. Instead of spending thousands of dollars to permanently alter the AI's brain, they asked a simple question: What if we just distill the experience of learning, and inject it as a memory? Here is how it works. They run the AI through the exact same trial-and-error process. But instead of updating the weights, they extract the "semantic advantage"—the actual logic of why one answer was better than another. They compress this winning logic into a "token prior”, a tiny package of high-quality experiential knowledge. Then, they just attach that knowledge directly into the API call. The results are staggering. Tested on DeepSeek-V3, this method required only a few dozen training samples to turn the AI into a specialized expert in complex math and web searching. It didn't just compete with models that were actually fine-tuned. It outperformed them. Zero parameter updates. Zero expensive training runs. Zero base-model amnesia.

@BurgenDobs @om_patel5 Science and software doesn't care about subtlety. But Semitic Spiritual Seed Oil for the Soul requires it.

@LaWRLDxx @om_patel5 Bayesian inference is "just saying numbers" to a moron. Heidegger or Kierkegaard is "just saying letters of the alphabet", to a moron. youtube.com/watch?v=dQWtl9…