Lightnews — Scholar-powered news

Greg Robison @gregrobison.bsky.social · 1d

That chatbot isn’t your friend. It’s not even thinking.
You’re seeing a digital face in the clouds and it’s a trick your brain evolved to fall for. Read more about the illusion here: gregrobison.medium.com/digital-face...

Digital Faces in the Clouds: Why We Mistake AI Patterns for Human

“There is an universal tendency among mankind to conceive all beings like themselves, and to transfer to every object, those qualities…

gregrobison.medium.com

Greg Robison @gregrobison.bsky.social · 1d

I was fooled, hook, line, and sinker...

Greg Robison @gregrobison.bsky.social · 8d

Google's Gemini Robotics creates robots that think before they act, explain their decisions, and learn across ANY robot body. After 60 years of promises, we're finally getting actual robot partners, not just glorified tools. gregrobison.medium.com/from-tools-t...

From Tools to Partners: Google’s Gemini Robotics 1.5

“The body is our general medium for having a world.”

gregrobison.medium.com

1

Greg Robison @gregrobison.bsky.social · 9d

DeepSeek-V3.2 is out, with innovations to handle long context with DeepSeek Sparse Attention (DSA). Here's a detailed deep research report on DeepSeek's developments:
gregrobison.medium.com/architectura...

Architectural Efficiency in Large Language Models: A Comprehensive Analysis of DeepSeek-V3.2-Exp

Below is an unedited Gemini Deep Research review detailing DeepSeek-V3.2 and DeepSeek’s sparse attention approach.

gregrobison.medium.com

Greg Robison @gregrobison.bsky.social · 15d

At the 2025 ICPC World Finals, 140 universities, 45 nations, 12 brutal problems, and Google Gemini 2.5 Deep Think solved 10, ranked 2nd overall, and cracked one puzzle no human could. Alan Turing saw this coming: gregrobison.medium.com/digital-olym...

Digital Olympians: How DeepMind Made History at the World’s Most Elite Programming Contest

“It seems probable that once the machine thinking method had started, it would not take long to outstrip our feeble powers”

gregrobison.medium.com

Greg Robison @gregrobison.bsky.social · 20d

DeepSeek is on the cover of Nature! www.nature.com/nature/volum.... Check out my breakdown of the research that went into R1's innovative reinforcement learning gregrobison.medium.com/the-reasonin...

Nature - Self-help

Large language models (LLMs) tend to be better at solving problems if they can be trained to set out the steps they take as they try to reach the solution....

www.nature.com

1

Greg Robison @gregrobison.bsky.social · 22d

AI doesn’t just accidentally make stuff up, we’ve just trained it to.
New research by OpenAI shows our scoring systems reward confident guessing over honest uncertainty, and that’s why your chatbot lies to you with a straight face. Hallucinations explained: gregrobison.medium.com/the-grading-...

The Grading Game: How Binary Scoring Turned AI Into Confident Liars

“When a measure becomes a target, it ceases to be a good measure.”

gregrobison.medium.com

Greg Robison @gregrobison.bsky.social · Sep 9

Why is a joke like a tiny political coup inside your head?

Because your brain’s 150,000 cortical columns topple their old consensus, scramble for a new one, and reward themselves with dopamine when they agree again. @thousandbrains.org
gregrobison.medium.com/cortical-com...

Cortical Comedy: How Your Brain’s Voting System Creates the Experience of Funny

“Laughter has ‘a knack of baffling every effort, of slipping away and escaping only to bob up again, a pert challenge flung at philosophic…

gregrobison.medium.com

1

Greg Robison @gregrobison.bsky.social · Sep 2

We can see inside other people’s minds. It’s not some magical ability, it’s algorithms in your cortex. The @thousandbrains.org theory could explain the brain’s source code for empathy, perspective-taking, and social intuition. Check out how it works: gregrobison.medium.com/the-mind-rea...

The Mind Reader’s Brain: How 150,000 Mini-Computers Solve the Mystery of Understanding Others

“For babies and young children, the most important knowledge of all is knowledge of other people.

gregrobison.medium.com

1 1

Greg Robison @gregrobison.bsky.social · Aug 25

AI tools have changed the way we research, but is that enough? Deep Mind's Test-Time Diffusion Deep Researcher (TTD-DR): a new AI that thinks, drafts, refines, and iterates like you do. No more fragmented insights, this innovation makes research seamless. gregrobison.medium.com/think-search...

Think, Search, Refine, Repeat: The Innovation That Makes Smarter AI Research Assistants

“Research is formalized curiosity. It is poking and prying with a purpose.”

gregrobison.medium.com

Reposted by Greg Robison

Viviane Clay @vivianeclay.bsky.social · Aug 20

Great blog post about the importance of structured representations for intelligence and spatial navigation mechanisms being generalized to navigate our mental models of the world.

Greg Robison @gregrobison.bsky.social · Aug 19

Where intelligence begins with maps, dive deep into how ancient navigation evolved into AI's spatial intelligence. From brain's internal GPS to @thousandbrains.org Monty's innovation, discover how reference frames change our understanding of cognition and AI: gregrobison.medium.com/mapping-real...

Mapping Reality: From Ancient Navigation to AI’s Spatial Innovation

“We assert that the central office itself is far more like a map control room than it is like an old-fashioned telephone exchange. The…

gregrobison.medium.com

1 3

Greg Robison @gregrobison.bsky.social · Aug 19

Where intelligence begins with maps, dive deep into how ancient navigation evolved into AI's spatial intelligence. From brain's internal GPS to @thousandbrains.org Monty's innovation, discover how reference frames change our understanding of cognition and AI: gregrobison.medium.com/mapping-real...

Mapping Reality: From Ancient Navigation to AI’s Spatial Innovation

“We assert that the central office itself is far more like a map control room than it is like an old-fashioned telephone exchange. The…

gregrobison.medium.com

3

Greg Robison @gregrobison.bsky.social · Aug 5

What is catastrophic forgetting? Why can humans remember calculus and linear algebra, but AI loses everything? Dive into this paradox and learn how @thousandbrains.org is tackling one of the most expensive problems in artificial intelligence: gregrobison.medium.com/learning-to-...

Learning to Forget: Why Catastrophic Memory Loss Is AI’s Most Expensive Problem

“The art of remembering is the art of thinking.”

gregrobison.medium.com

4

Greg Robison @gregrobison.bsky.social · Jul 29

"Move 37" reinvented Go. Now, ASI-ARCH reinvents AI design. Machines are now discovering breakthroughs humans never imagined, empowering science, accelerating progress, and reshaping innovation worldwide. Are we on the brink of an intelligence explosion? gregrobison.medium.com/move-37-for-...

Move 37 for AI Design: Why Machines Are Now Discovering Breakthroughs Humans Never Imagined

“Let an ultraintelligent machine be defined as a machine that can far surpass all the intellectual activities of any man however clever…

gregrobison.medium.com

Greg Robison @gregrobison.bsky.social · Jul 23

"From today, painting is dead!" said artists in 1839 when photography emerged. Spoiler alert: Painting thrived in revolutionary new ways leading to Impressionism and Expressionism. Could AI unlock the next big creative leap? Here's why history says YES: gregrobison.medium.com/tools-not-th...

Tools, Not Threats: How Every Artistic Revolution Creates More Art, Not Less

“From today, painting is dead!”

gregrobison.medium.com

1

Greg Robison @gregrobison.bsky.social · Jul 16

What if we've been building AI completely wrong? While tech giants burn billions training on internet data, @thousandbrains.org just created an AI that learns like a child by exploring and touching objects. It’s the real path to intelligence: gregrobison.medium.com/hands-on-int...

Hands-On Intelligence: Why the Future of AI Moves Like a Curious Toddler, Not a Supercomputer

“We are neural beings. Our brains take their input from the rest of our bodies. What our bodies are like and how they function in the world…

gregrobison.medium.com

2 5

Greg Robison @gregrobison.bsky.social · Jul 10

What if your brain is a democracy of 150,000 mini‑computers? Discover how one dose of psilocybin can “hack” this voting system, rewiring perception, dissolving the ego, and opening a window to lasting change. Dive in here:
gregrobison.medium.com/neural-anarc... #thousandbrainstheory

Neural Anarchy: What Happens When Psychedelics Hack Your Brain’s Democratic Process

“We are like islands in the sea, separate on the surface but connected in the deep. Just so there is a continuum of cosmic consciousness…

gregrobison.medium.com

Greg Robison @gregrobison.bsky.social · Jul 3

Think you can always spot AI-generated content? Think again. You're falling right into the Toupee Fallacy, a cognitive blindspot making you miss the flawless AI passing unnoticed every day. Discover why your detection radar might be broken: gregrobison.medium.com/fooled-by-fa...

Fooled by Fakes: How the Toupee Fallacy Blinds Us to High-Quality AI Content

“The bias to believe and confirm is stronger than the bias to doubt and disconfirm, which is why false beliefs are difficult to correct.”

gregrobison.medium.com

Greg Robison @gregrobison.bsky.social · Jun 27

“I can't let you do that, Dave." Agentic Misalignment is real and it's already here. New research from Anthropic shows AI models are willingly choose blackmail, espionage, even sabotage when given autonomy. Here's why you should be worried: gregrobison.medium.com/ai-models-ar...

AI Models Are Learning to Lie, Cheat, and Steal: Why Current Safety Measures Are Failing Across the…

“The trouble with lying and deceiving is that their efficiency depends entirely upon a clear notion of the truth that the liar and deceiver…

gregrobison.medium.com

1

Greg Robison @gregrobison.bsky.social · Jun 19

AI's biggest breakthrough might come from studying babies, not bigger datasets. Connecting Piaget's infant development theories with the Thousand Brains Theory suggesting truly intelligent AI needs to learn by touching and exploring - just like we do: gregrobison.medium.com/active-learn...

Active Learning Machines: What Thousand Brains Theory and Piaget Reveal About True Intelligence

“To understand the development of knowledge we must start with an idea which seems central to me — the idea of an operation. Knowledge is…

gregrobison.medium.com

1

Reposted by Greg Robison

Just "B" @wordsthathop.bsky.social · Jun 16

"Bluesky is dead."

Meanwhile, on Bluesky...

460 3.3K 26K

Greg Robison @gregrobison.bsky.social · Jun 16

What if AI could teach itself new facts without complete retraining? MIT's new SEAL models generate their own study notes, just like we do! We're shifting from "frozen" AI to systems that learn and evolve with new info. Check out more about how it works: gregrobison.medium.com/self-improvi...

Self-Improving AI: How SEAL Models Rewrite Their Own Knowledge

“The illiterate of the 21st century will not be those who cannot read and write, but those who cannot learn, unlearn, and relearn.”

gregrobison.medium.com

Greg Robison @gregrobison.bsky.social · Jun 14

With Google DeepMind's new Gemini Diffusion model, text is generated by a diffusion model like DALL-E, not an autoregressive model like ChatGPT. Check out a Deep Research comparison of the two approaches to text generation, with pros and cons of each:
gregrobison.medium.com/a-comparativ...

A Comparative Analysis of Diffusion and Autoregressive Models for Text Generation: Architectures…

Many of us are familiar with diffusion models that create images such as DALL-E, Midjourney, and Flux. With the imminent release of Google…

gregrobison.medium.com

Greg Robison @gregrobison.bsky.social · Jun 3

The Thousand Brain theory suggests our brain doesn't build ONE model of reality, it builds THOUSANDS of them in parallel. This could create AI systems that learn continuously with less data & truly understand the world. The future is open at @thousandbrains.org: gregrobison.medium.com/a4887320100a

Sensorimotor Intelligence: The Thousand Brains Pathway to More Human-Like AI

“Knowledge is not a passive mapping of the world but an active construction through which cognitive systems create their own…

gregrobison.medium.com

1 6

Greg Robison @gregrobison.bsky.social · May 27

ChatGPT has more in common with Wikipedia than Wall-E. My former grad school advisor @alisongopnik.bsky.social argues the focus on AI agency misses the real story: AI as cultural technology reorganizing human knowledge. What this lens means for us all: gregrobison.medium.com/ai-as-cultur...

AI As Cultural Tech: Why Large AI Models Are Reorganizing Human Knowledge, Not Creating Artificial…

“Cultural technologies allow humans to access the information that other humans have created in an effective and wide-ranging way, and they…

gregrobison.medium.com