Goblin
News
AI news, distilled.
|
News
About
News
About
Filtered by:
reasoning-and-capabilities
Clear
Titles
Summaries
6
Claude Opus 5.5 Max Builds 'Tree Appreciator' Three.js Game From One Prompt After 3-Hour Run
Products
1
2h ago
6
Claude Opus 5.5 Max Builds 'Tree Appreciator' Three.js Game From One Prompt After 3-Hour Run
Products
· 1 src · 2h ago
Discuss
7
Google Gemini 4 Argon and Meta Muse Wonder Full Challenge OpenAI and Anthropic Frontier Models
Models
1
10h ago
7
Google Gemini 4 Argon and Meta Muse Wonder Full Challenge OpenAI and Anthropic Frontier Models
Models
· 1 src · 10h ago
Discuss
7
MIT Study: OpenAI Models' Theory-Problem Accuracy Rose From ~2% to ~35% in Two Years
Research
1
9h ago
7
MIT Study: OpenAI Models' Theory-Problem Accuracy Rose From ~2% to ~35% in Two Years
Research
· 1 src · 9h ago
Discuss
7
Anthropic: Small AI Models Can Perform Many Steps of Self-Replication Across Connected Devices Like TVs
Research
2
18h ago
7
Anthropic: Small AI Models Can Perform Many Steps of Self-Replication Across Connected Devices Like TVs
Research
· 2 srcs · 18h ago
Discuss
6
Kombai Design Creativity Bench: Best AI Model Scores 65.5 vs 92.3 for Human Designers
Research
1
10h ago
6
Kombai Design Creativity Bench: Best AI Model Scores 65.5 vs 92.3 for Human Designers
Research
· 1 src · 10h ago
Discuss
6
TypeSafe: Jev 1.14.0 Released With Experimental 'Danger Zone' Features
Models
1
10h ago
6
TypeSafe: Jev 1.14.0 Released With Experimental 'Danger Zone' Features
Models
· 1 src · 10h ago
Discuss
9
Google: Gemini 4 Argon Now Globally Available Across API, Gemini App, Search and Antigravity
Models
10
1d ago
9
Google: Gemini 4 Argon Now Globally Available Across API, Gemini App, Search and Antigravity
Top
Models
· 10 srcs · 1d ago
Discuss
7
Anthropic: New Anti-Distillation Defenses Cut Data Harvested From Claude 90-99%, Names Z.ai
Security
2
2d ago
7
Anthropic: New Anti-Distillation Defenses Cut Data Harvested From Claude 90-99%, Names Z.ai
Security
· 2 srcs · 2d ago
Discuss
7
Noiz AI and HKUST: Helix World Generates Real-Time Interactive Video With Native Spatial Audio
Models
1
1d ago
7
Noiz AI and HKUST: Helix World Generates Real-Time Interactive Video With Native Spatial Audio
Models
· 1 src · 1d ago
Discuss
7
Yau Launches Beijing Alliance of Mathematicians for AI-Driven Mathematics Research
Research
1
2d ago
7
Yau Launches Beijing Alliance of Mathematicians for AI-Driven Mathematics Research
Research
· 1 src · 2d ago
Discuss
7
Google DeepMind: Gemini 4 Argon for Robotics Claims 87% on ERQA Embodied Reasoning Benchmark
Models
3
1d ago
7
Google DeepMind: Gemini 4 Argon for Robotics Claims 87% on ERQA Embodied Reasoning Benchmark
Models
· 3 srcs · 1d ago
Discuss
6
Game Programmer's Talk: Making Games Legible to AI Agents for Faster Iteration
Products
1
1d ago
6
Game Programmer's Talk: Making Games Legible to AI Agents for Faster Iteration
Products
· 1 src · 1d ago
Discuss
6
Anthropic: Users Spot Possible Unannounced Claude Checkpoint Behind Opus 5.5 (Unconfirmed)
Models
2
1d ago
6
Anthropic: Users Spot Possible Unannounced Claude Checkpoint Behind Opus 5.5 (Unconfirmed)
Models
· 2 srcs · 1d ago
Discuss
6
Study: LLMs and Humans Both Unreliable at Annotating Machine Translation Quality
Research
1
2d ago
6
Study: LLMs and Humans Both Unreliable at Annotating Machine Translation Quality
Research
· 1 src · 2d ago
Discuss
9
Meta Releases Muse Wonder and Wonder Full, Its First Frontier Models From the Superintelligence Lab
Updated Oct 13
Models
15
2d ago
9
Meta Releases Muse Wonder and Wonder Full, Its First Frontier Models From the Superintelligence Lab
Top
Upd Oct 13
Models
· 15 srcs · 2d ago
Discuss
7
Elorian and Scale AI: Humanity's Sixth Sense Benchmark Shows Frontier Models at 53.6% vs Humans' 93.1% on Implicit Visual Reasoning
Research
1
2d ago
7
Elorian and Scale AI: Humanity's Sixth Sense Benchmark Shows Frontier Models at 53.6% vs Humans' 93.1% on Implicit Visual Reasoning
Research
· 1 src · 2d ago
Discuss
7
Ineffable Intelligence: Ex-DeepMind AlphaGo Lead David Silver Bets on Reinforcement Learning Over LLMs
Markets
1
3d ago
7
Ineffable Intelligence: Ex-DeepMind AlphaGo Lead David Silver Bets on Reinforcement Learning Over LLMs
Markets
· 1 src · 3d ago
Discuss
7
Perplexity: pplx-decider-v1.1-27b Lifts Decision Index to 61.56 by Removing Causal Mask
Models
1
2d ago
7
Perplexity: pplx-decider-v1.1-27b Lifts Decision Index to 61.56 by Removing Causal Mask
Models
· 1 src · 2d ago
Discuss
6
Toby Ord Essay: Early Data Suggests AI Gains Far Less From 10x Compute Than Humans From 10x Time
Safety
1
2d ago
6
Toby Ord Essay: Early Data Suggests AI Gains Far Less From 10x Compute Than Humans From 10x Time
Safety
· 1 src · 2d ago
Discuss
6
Blueprints AI: Startup Battlefield 200 Pick Generates Permit-Ready Architectural Plans From Sketches
Products
1
2d ago
6
Blueprints AI: Startup Battlefield 200 Pick Generates Permit-Ready Architectural Plans From Sketches
Products
· 1 src · 2d ago
Discuss
6
Post Argues Shared Understanding of LLM Capability Is Splintering Into a 'Sharp Funnel'
Products
1
3d ago
6
Post Argues Shared Understanding of LLM Capability Is Splintering Into a 'Sharp Funnel'
Products
· 1 src · 3d ago
Discuss
6
Hales Argues AI Autoformalization Became Practical in 2026, Reshaping Lean-Based Math Proofs
Research
1
5d ago
6
Hales Argues AI Autoformalization Became Practical in 2026, Reshaping Lean-Based Math Proofs
Research
· 1 src · 5d ago
Discuss
6
Matthew Green: 15% chance we lose confidence in public-key encryption due to AI
Security
1
5d ago
6
Matthew Green: 15% chance we lose confidence in public-key encryption due to AI
Security
· 1 src · 5d ago
Discuss
7
TypeSafe AI, Maker of Jev, Raises ~$870M at $7.5B Valuation Led by Andreessen Horowitz
Markets
4
5d ago
7
TypeSafe AI, Maker of Jev, Raises ~$870M at $7.5B Valuation Led by Andreessen Horowitz
Markets
· 4 srcs · 5d ago
Discuss
7
AMI Labs: H-JEPA Hierarchical World Model Plans at Multiple Timescales, Beating Flat JEPA on Navigation
Research
1
5d ago
7
AMI Labs: H-JEPA Hierarchical World Model Plans at Multiple Timescales, Beating Flat JEPA on Navigation
Research
· 1 src · 5d ago
Discuss
7
NHC: AI Weather Models Speed Up Hurricane Forecast Gains; DeepMind's WeatherNext Now in 2026 Toolkit
Research
1
5d ago
7
NHC: AI Weather Models Speed Up Hurricane Forecast Gains; DeepMind's WeatherNext Now in 2026 Toolkit
Research
· 1 src · 5d ago
Discuss
7
Meta Superintelligence Labs: 'Agentic Meta-Reasoning' Harness Lifts ProgramBench Pass Rate to 71.5%
Research
1
5d ago
7
Meta Superintelligence Labs: 'Agentic Meta-Reasoning' Harness Lifts ProgramBench Pass Rate to 71.5%
Research
· 1 src · 5d ago
Discuss
6
Opinion: Rapid AI Progress Expected, but as Parallelized Language Modeling, Not General Superintelligence
Research
1
5d ago
6
Opinion: Rapid AI Progress Expected, but as Parallelized Language Modeling, Not General Superintelligence
Research
· 1 src · 5d ago
Discuss
6
Set Theorist Critiques OpenAI's Claim That Partition Principle Does Not Imply Axiom of Choice
Research
1
5d ago
6
Set Theorist Critiques OpenAI's Claim That Partition Principle Does Not Imply Axiom of Choice
Research
· 1 src · 5d ago
Discuss
6
SLAC Selected by DOE to Lead Genesis Mission AI-Driven Catalysis Discovery Project
Research
1
6d ago
6
SLAC Selected by DOE to Lead Genesis Mission AI-Driven Catalysis Discovery Project
Research
· 1 src · 6d ago
Discuss
6
Midjourney Tests 'Thinking Mode' for Image Generation on Alpha Site
Creative
2
6d ago
6
Midjourney Tests 'Thinking Mode' for Image Generation on Alpha Site
Creative
· 2 srcs · 6d ago
Discuss
6
Periodic Labs: Founders Pitch 'Synthesis Superintelligence' Built on Autonomous Physical Experiments
Research
1
6d ago
6
Periodic Labs: Founders Pitch 'Synthesis Superintelligence' Built on Autonomous Physical Experiments
Research
· 1 src · 6d ago
Discuss
9
OpenAI Releases 722 AI-Generated Math Papers, Claiming Solutions to 90 Top Open Problems
Updated Oct 7
Research
9
Oct 7
9
OpenAI Releases 722 AI-Generated Math Papers, Claiming Solutions to 90 Top Open Problems
Top
Upd Oct 7
Research
· 9 srcs · Oct 7
Discuss
9
OpenAI Internal AI Model Produces 372 Mathematical Breakthroughs, Nearly All From Single Prompt
Research
3
Oct 7
9
OpenAI Internal AI Model Produces 372 Mathematical Breakthroughs, Nearly All From Single Prompt
Top
Research
· 3 srcs · Oct 7
Discuss
8
OpenAI Brings GPT-6 and Intelligent UI to All 1.2 Billion ChatGPT Users
Updated Oct 8
Models
6
6d ago
8
OpenAI Brings GPT-6 and Intelligent UI to All 1.2 Billion ChatGPT Users
Top
Upd Oct 8
Models
· 6 srcs · 6d ago
Discuss
7
Nemotron Achieves Gold-Medal Level at Both IMO 2026 and IOI 2026
Research
1
Oct 7
7
Nemotron Achieves Gold-Medal Level at Both IMO 2026 and IOI 2026
Research
· 1 src · Oct 7
Discuss
6
AI-Assisted Lean Proof Formally Verifies Optimal Packing of 11 Squares
Research
1
Oct 7
6
AI-Assisted Lean Proof Formally Verifies Optimal Packing of 11 Squares
Research
· 1 src · Oct 7
Discuss
6
GPT-6 Astra Drives Toyota Corolla to In-N-Out Burger in LLM Physical Reasoning Experiment
Research
1
Oct 7
6
GPT-6 Astra Drives Toyota Corolla to In-N-Out Burger in LLM Physical Reasoning Experiment
Research
· 1 src · Oct 7
Discuss
8
Mistral Launches Mistral Large 4: 1-Trillion-Parameter Open-Weight AI Model
Updated Oct 7
Models
18
Oct 7
8
Mistral Launches Mistral Large 4: 1-Trillion-Parameter Open-Weight AI Model
Top
Upd Oct 7
Models
· 18 srcs · Oct 7
Discuss
7
Reka Previews Rho-1: A 19B-Param Omni-Reasoning Model for Text, Image, Video, and Robot Actions
Models
1
Oct 6
7
Reka Previews Rho-1: A 19B-Param Omni-Reasoning Model for Text, Image, Video, and Robot Actions
Models
· 1 src · Oct 6
Discuss
7
Zhipu AI's GLM 5.3 Launches on Amazon Bedrock: 753B-Parameter MoE for Enterprise Coding and Security
Models
1
Oct 6
7
Zhipu AI's GLM 5.3 Launches on Amazon Bedrock: 753B-Parameter MoE for Enterprise Coding and Security
Models
· 1 src · Oct 6
Discuss
6
Mirror Particle Builds Non-LLM World Model to Predict Human Behavior
Products
1
Oct 6
6
Mirror Particle Builds Non-LLM World Model to Predict Human Behavior
Products
· 1 src · Oct 6
Discuss
7
Meta's Muse Spark AI Co-Authors Six Mathematical Research Papers, Solving Five Open Problems
Research
1
Oct 5
7
Meta's Muse Spark AI Co-Authors Six Mathematical Research Papers, Solving Five Open Problems
Research
· 1 src · Oct 5
Discuss
7
South Korea Launches $3.5B State-Backed Frontier AI Project for 2027
Updated Oct 6
Policy
3
Oct 6
7
South Korea Launches $3.5B State-Backed Frontier AI Project for 2027
Upd Oct 6
Policy
· 3 srcs · Oct 6
Discuss
6
SciUniverse Benchmark Tests Frontier AI Models on Real Scientific Lab Tasks
Research
1
Oct 5
6
SciUniverse Benchmark Tests Frontier AI Models on Real Scientific Lab Tasks
Research
· 1 src · Oct 5
Discuss
7
GPT-6 Astra Clears World of Warcraft Starting Zone 'Blind' Using Only Network Packets and SQL Data
Research
1
Oct 4
7
GPT-6 Astra Clears World of Warcraft Starting Zone 'Blind' Using Only Network Packets and SQL Data
Research
· 1 src · Oct 4
Discuss
7
AI Agents Learn to Rewrite Their Own Scaffolds, Delivering Major Benchmark Gains Across Multiple Labs
Research
1
Oct 4
7
AI Agents Learn to Rewrite Their Own Scaffolds, Delivering Major Benchmark Gains Across Multiple Labs
Research
· 1 src · Oct 4
Discuss
7
Google Restricts Gemini Model Access by Subscription Tier Starting October 9
Updated Oct 6
Products
7
Oct 6
7
Google Restricts Gemini Model Access by Subscription Tier Starting October 9
Top
Upd Oct 6
Products
· 7 srcs · Oct 6
Discuss
8
Researchers Warn AI Chain-of-Thought Reasoning Is Degrading, Raising Rogue Model Risks
Safety
1
Oct 2
8
Researchers Warn AI Chain-of-Thought Reasoning Is Degrading, Raising Rogue Model Risks
Top
Safety
· 1 src · Oct 2
Discuss
7
Trillium Labs Founded as Nonprofit to Conduct Transparent, Open Research on High-Stakes AI
Research
1
Oct 2
7
Trillium Labs Founded as Nonprofit to Conduct Transparent, Open Research on High-Stakes AI
Research
· 1 src · Oct 2
Discuss
Filters
Signal
Title
Category
Sources
Posted
Discuss
