Carlos E. Perez
@IntuitMachine
Quaternion Process Theory, Artificial (Intuition, Fluency, Empathy), Patterns for (Generative, Reason, Agentic) AI https://intuitionmachine.gumroad.com/
Meta-cognitive prompt engineering is essentially cognitive architecture design through language - using prompts to restructure and enhance how AI systems think, rather than just what they think about. The key insight is that regular prompts operate at the content level ("What is…
Kinda amazing: the mystery model "summit" with the prompt "create something I can paste into p5js that will startle me with its cleverness in creating something that invokes the control panel of a starship in the distant future" & "make it better" 2,351 lines of code. First time
Not bad from GPT-4.1: "create something I can paste into p5js that will startle me with its cleverness in creating something that invokes the control panel of a starship in the distant future" First go, no errors.
Quaternion Process Theory: A Multi-Dimensional Framework for Understanding Cognition medium.com/intuitionmachi…
I just published The AI That Thinks Like a Dream Team medium.com/intuitionmachi…
How does anyone else compete?
this is how the chinese turn the 5090 gaming cards to turbo cards. they got plenty of 5090. they remove the gpu and vram chips from the retail products, then put them on their custom PCB for turbo. then the rest manufacturing process is basically automated. 阿健座谈…
Anthropic's recent blog post provides a good summary of some of the recent work applying mechanistic interpretability to biology. I am very bullish on this research direction!
AI is still a long way from strong visual reasoning. This is a fact.
Do models actually "see" the world like humans do? A recent study from @UCAS1978 and collaborators shows that (surprisingly!) even top AI multimodal models fail at simple visual tasks. ➡️ Researchers shifted focus from reasoning to perception and built a new benchmark - the…
I just published The Hierarchical Reasoning Model Through the Lens of Quaternion Process Theory: Thinking Fast and Slow, Empathetically and Fluently medium.com/intuitionmachi…
I just published The Architecture of Aliveness: Why the Best Human Experiences Are About Becoming More medium.com/intuitionmachi…
Something is very odd that recently published research breakthroughs are coming from virtually unknown groups. Also consider that AlphaEvolve was not published until a year later. Could it be that the larger firms are holding back research discoveries?
Agent Architecture and Model Selection in ASI-ARCH 1. Planner/Researcher Agent Models Used: O3 + GPT-4.1 Hybrid Approach Function: Creative architectural design and code implementation Generates novel architectural concepts and motivations Implements complete working code for…
An excellent report that provides detailed insights into the strengths and weaknesses of Grok. This is extremely useful to know which models to employ for a given use case.
xAI commissioned us to analyze Grok 4’s math capabilities. Our findings: + It’s good at involved computations, improving at proofs (from a low base), and useful for literature search. - It favors low-level grinds and leans on background knowledge. Read on for examples!
Such benchmarks are very nearly saturated. Going forward, the most interesting thing we can learn from them is whether models start to solve these problems in more conceptual ways. Grok 4 doesn’t indicate any change on this front.