Nat
banner
natgpt.bsky.social
Nat
@natgpt.bsky.social
🚀 AI Researcher | Coder
🌟 Exploring intelligence & training LLMs.
♟️ Former Chess Player
📚 Translator
Somebody combined Midjourney + MagnificAI + Runway Gen-4
April 2, 2025 at 5:15 AM
LLMs.txt generator v2 - Generate consolidated text files from websites for LLM training and inference
llmstxt.firecrawl.dev
March 12, 2025 at 5:40 AM
Factorio Learning Environment

jackhopkins.github.io/factorio-lea...
March 12, 2025 at 5:38 AM
GSM8K-Platinum: Revealing Performance Gaps in Frontier LLMs
gradientscience.org/gsm8k-platin...
March 12, 2025 at 5:37 AM
No one has ever truly decoded intelligence. We could spend decades trying to piece together a puzzle with trillions of fragments suspended in mid-air. Humans are undoubtedly smart; they've created systems that, while not perfect replicas, offer a decent simulation of human intelligence.
March 11, 2025 at 5:27 PM
We've spent decades searching for intelligence, trying to address difficult questions like: What is intelligence? How does the brain compute the mind and generate intelligence? How can we replicate it in machines? Our intent was to create labor-saving devices, but we'll end up creating digital gods
March 11, 2025 at 5:26 PM
Beyond our five senses, intuition might tap into a broader universal intelligence.
March 11, 2025 at 5:25 PM
General intelligence requires something called 'action', which means reaching an X goal. Recognizing the importance of goal-oriented behavior is a crucial step. Not everyone understands this 'minor' detail. I don't think AGI is decades away. It's already knocking at the door.
March 11, 2025 at 5:25 PM
The historical battle between silicium-based intelligence and carbon-based intelligence now seems like a swift prelude to what comes next—a requiem for humanity or an anthem for the next generation of merged intelligence.
March 11, 2025 at 5:24 PM
Once again we've seen that the internet is a shared global intelligence - digital collective consciousness. Study internet. Measuring organic intelligence is pointless. Decoding intelligence is Sisyphean Task. The first step to explore the intelligence lies in understanding the internet.
March 11, 2025 at 5:24 PM
Mistakes don’t define intelligence—how you respond to them does.
March 11, 2025 at 5:23 PM
Real influence isn’t about proving intelligence—it’s about transferring it.
March 11, 2025 at 5:23 PM
AGI is a poorly defined term, but we can break it into stages:
- First Gen AGI: Human-level reasoning, task generalization (Renaissance 2.0).
- Next Gen AGI: Expands beyond human limits, shaping fields (Meta-goal).
- Supreme AGI: A singularity-like intelligence, reshaping existence (Grand finale)
March 11, 2025 at 5:23 PM
Next Chapter: Distributed Consciousness
March 11, 2025 at 5:22 PM
🧵 The war in Ukraine has transformed the inexperienced Zelensky into a political mastermind. From the beginning, he accurately read what Trump intended for him and his country, and to this day he offers dignified resistance against the completely incompetent leader of a superpower.
March 11, 2025 at 5:18 PM
Well, I will try to be more active here :)
March 9, 2025 at 7:40 AM
Another open-source project I've finished is a personal Micro-Blog. This project is a minimalist, user-friendly micro-blogging platform designed for those who want to share thoughts, daily goals, and life events in a simple, streamlined way.
Get it here: github.com/NatGeoX/Micr...
January 11, 2025 at 6:27 PM
If anyone is interested, I’m working on a new publishing platform called PureText. It’s a clean, simple solution for writing essays—no images, just pure focus on text. 🫡 I’ll resume work on it in February. I’m still dealing with family matters after my dad passed away a few weeks ago.
January 11, 2025 at 6:22 PM
Folks, is it possible to cross-post from Twitter? I find it difficult to maintain two platforms.
January 11, 2025 at 6:10 PM
Incomplete information creates distorted expectations.
January 9, 2025 at 4:26 PM
Reposted by Nat
Theories of Causal Reasoning

www.youtube.com/watch?v=-9JJ...
Theories of Causal Reasoning
YouTube video by Knut Jägersberg
www.youtube.com
November 27, 2024 at 11:08 AM
Fears can be rational or irrational, both rooted in internal trauma. Rational fears arise from familiar, real-world risks. Irrational fears stem from past traumas, projecting unlikely but emotionally potent scenarios onto the future.
November 27, 2024 at 9:27 AM
Hi there, this is my first post here. 🙂
November 22, 2024 at 3:56 PM