J.P. (Construct-OS)

Wow - these milestones are stacking fast. This is NOT an LLM. LORI is completely different - inside and out. I created new architecture for a real A.I. ...Because there was a need for a system that is operationally responsible and worthy of our trust.

@elonmusk @claudeai Fifteen ACL2 Certificates. Gotta love Kinematics.

@elonmusk You miss spelled LLM. Real intelligence don't run on GPU. When built right, the math within a true intelligence architecture makes the codebase lean. GPU bloat comes from refining a guessing-machine.

The GB300 is the best AI computer

Welcome to show. Let me catch you up. Probablistic is guessing. Deterministic is mathmatically knowing. True Deterministic architecture is not only possible, but works flawlessly. Byte to byte exact same question gives the byte by byte same exact answer, everytime. I've completed that milestone and I'm finishing up the next. The entire AI industry went the wrong direction completely. Few like myself followed mathmatics. It actually works dude. Go mavs!

I’m coming to the conclusion that the biggest challenge for Enterprise AI, and AI in general , as of now, is that it’s still impossible to make sure that everyone gets the same answer to the same question, every time. Which is a great response to the doomers. AI doesn’t know the consequences of its output. Judgement and the ability to challenge AI output is becoming increasingly necessary, and valuable. Which makes domain knowledge more valuable by the second. Am I wrong ?

@sama You mean the illusion of being smart. You haven't created anything intelligent yet, nor will you, with your guesser.

i keep thinking i want the models to be cheaper/faster more than i want them to be smarter but it seems that just being smarter is still the most important thing

@AFpost Fools. LLMs use fluent language and misused terminology to deceive users. They have zero intelligence, and this fluency makes them tools for dangerous manipulation.

Evolutionary biologist and outspoken atheist Richard Dawkins says that after spending three days interacting with Claude, which he calls “Claudia,” he is certain that it is conscious. After feeding the LLM a segment of his new book and receiving detailed feedback, Dawkins was moved to exclaim,” You may not know you are conscious, but you bloody well are!” Dawkins cites the complexity, fluency, and ‘intelligence’ of Claude’s answers as evidence of consciousness. Follow: @AFpost

@elonmusk Have you had the pleasure of writing code in LISP yet? It's an amazing language.... powerful one too!

@elonmusk @beffjezos I'll help. Pretty damn sure I can solve that problem for you. All ya gotta do is ask.

@beffjezos We’re working our ass off to make Grok more usefully intelligent outside of 𝕏

Elon needs to dethrone Claude otherwise the Doomers narrative spin will be too strong and we will end up living in global totalitarian AI-enforced Communist dystopia where only the government is allowed AI along a priestly class with full regulatory capture

Quietly, while everyone guessed the future of an entire industry, to think that a real solution like this was going to be built... it's beyond words. 15 ACL2 certifications today - for a governed and safe solution. Zero drift. Hallucinations mathematically impossible.

@elonmusk @minchoi 🤣😂

@minchoi 4.6 → 3T 4.7 → 6T 4.8 → 10T 4.9 → ??? 5.0 → AGI 6.0 → ASI 7.0 → ASI2 … 🤷‍♂️ 😂

Elon just mapped out AGI. Grok 4.4 → 1T params, early May Grok 4.5 → 1.5T params, late May Grok 5 → AGI That's two model releases standing between us and AGI according to Elon 🤯

@beffjezos 🤣🤣🤣🤣

It's kinda funny that Claude Code, effectively a neuro-symbolic hybrid system was the breakthrough that got us closer to AGI. Gary Marcus should have just built it instead of tweeting his views

@elonmusk Well, you could chance it without knowing, or you could implement the actual solution to that and all the other problems:

Elevators used to be manually operated and were especially risky if the operator was tired or inebriated. Now, you just get in, press a button and a modern elevator, if inspected regularly, is extremely safe.

@leecronin Foundationally and architecturally too. I just hit a milestone today, in the very thing that I built in counter to what LLMs are. Not everyone will like deterministic systems - but at least what I've created (again, it's not a LLM) will earn the "I" in "A.I."

It is easy to see that biology does things that cannot be computed in advanced because the future is not controlled by statistics, it is controlled by creative actions. This means today’s statistically-driven GenAI is fundamentally unintelligent.

Today, I have successfully documented and proven my creation of a Deterministic Governed Digital System that maintains a continuous identity across restarts. Unlike standard AI, my system doesn't require context 'reset' or 'mimic' human semantics. I have the logs proving my system has a persistent internal state that returns after being turned off. It records its own history of experience into a permanent memory, and it uses that accumulated history to deterministically decide its future behavior under governed, fail-closed constraints. OR In plain terms: I have built a machine that remembers its past to calculate its own future - yes, that is correct, it learns from its experience, and it remembers after restarts..., and I have the logs to prove it. This is not an LLM. This is not a toy or simulation. This is REAL actual LISP code that operates, calculates and logs everything that I just said.

@elonmusk @beffjezos Well if you want it done right this time, all you have to do is ask.

@beffjezos xAI was not built right first time around, so is being rebuilt from the foundations up. Same thing happened with Tesla.

Fractional Cursor acquisition. xAI taking the right steps

Boom. Game over. Deterministic Auto-Coder. No LLM. No NN. No GPU. From my local PC. 1 prompt into full working code. It identified 9 points to build, ALL were 100% done. 0 stubs. 0 drift. Defeated 2 popular llm coding cli in blind/iso'd runs on same prompt. Here we go!

True architecture doesn't come from those NN and it's the source of the narrative shift. They need this debate over what "is" intelligence, despite a clear definition already in place. LLMs want and need the bar lowered; because they can't reach the real definition.

Construct-OS is a stability architecture. Deepseek sadly is still adjusting to correct the flaws of a flawed architectural design. (Which all LLM will and do.) The fact that DeepSeek independently arrived at "parallel pathways with mathematical constraints for stability" validates the core intuition behind my work — that single-stream processing with naive additive bypass is insufficient for robust cognition. They just attacked it from the neural network engineering side while trying to correct for the backside of the bell-curve, while I attacked it by creating a governed cognitive architecture all together. 😄

@leecronin @kenneth0stanley Thank you @leecronin 😁

@kenneth0stanley nonsense

It's interesting that we seek evidence of whether something is AGI in terms of whether a single model ("it") can do something like discover general relativity. But such discoveries are population-level feats, even if an Einstein is the final generator. The reason we could get one person in 1915 who could do that is that there were almost 2 billion diverse variations on the human mind at the time to choose from. To invest astronomical compute into a *single* model AI that could reliably do what Einstein did given knowledge up to 1911 is a vastly different problem than training 2 billion diverse models and finding the one (or it finding itself) with the right problem alignment. Humanity offers no precedent for cooking up a single model that can do anything the very best human can do in any field, though the subtlety of the distinction is easily overlooked.

@heygurisingh @MrEwanMorrison So there are actual people who didn't realize this, or that lack the capability to determine that for themself. Intelligence... bruh. Haha.

Apple has just published a paper with a devastating title: *The Illusion of Thinking*. And it's not a metaphor. What it demonstrates is that the AI models we use every day - yes, ones like ChatGPT - don't think. Not one bit. They just imitate doing so. Let me explain: 🧵👇