Videos

Ornith 1.0 from Deep Reinforce introduces self-scaffolding LLMs for agentic coding, where the model learns to...

Recursive coding agents, inspired by recursive language models (RLMs), unify tool calling and reasoning to make AI...

Qwen-AgentWorld introduces a world model that hallucinates environments to train agents more effectively than...

Databricks CTO Matei Zaharia and co-founder Reynold Xin discuss their vision for an 'Agent Cloud' that unifies...

Hermes Agent allows swapping different AI models for different tasks to optimize cost and performance, breaking the...

GPT (Generative Pre-trained Transformer) is the core architecture behind modern LLMs from labs like OpenAI,...

AI agents are like geniuses with amnesia: they are repo-bound (cannot see the whole system) and have no episodic...

Sakana Fugu Ultra is not a standalone model but an orchestration API that routes tasks to multiple frontier models...

Gray Swan (founded by CMU professors Zico Kolter and Matt Fredrikson) provides AI security solutions focused on the...

The Hermes agent integrated with Obsidian creates a powerful second brain system where markdown files become "living...

Claude Code, combined with a free open-source plugin, can perform a full SEO and AI Search (GEO/AEO) audit and...

Cohere's North Mini Code is a 30-billion parameter mixture-of-experts Apache 2.0 licensed coding model optimized for...

Agent loops replace the human in the feedback/iteration cycle by having the AI reason, act, and observe until a...

Voice agents for customer support require a hybrid cascaded architecture that balances control, latency, and...

Instead of manually prompting Claude repeatedly, you should design loops that automatically prompt agents until a...

VibeThinker 3B, a small model from Weibo AI Lab, beats giant models like Gemini 3 Pro and Claude Opus on math and...

GLM 5.2 is an open-source model that integrates into Claude Code as a cheaper, faster alternative to Opus 4.8 for...

Matt Pocock argues that developers should focus on the 'harness' (prompts, skills, codebase quality) rather than...

The video explains the bigram language model from Andrej Karpathy's 'GPT from scratch' tutorial, covering how to...

A $1,500 32GB VRAM local AI server built from three RTX 3060s (12GB+12GB+8GB) performs surprisingly well for LLM...

Loop engineering, where developers design loops that prompt AI coding assistants to work autonomously on large...

GLM 5.2 is an open-weights model from Z.AI that achieves top-tier benchmark scores, rivaling proprietary models like...

LLMs often falsely claim to have searched the web when they actually failed due to anti-bot measures, leading to...

OpenRouter Fusion, which synthesizes outputs from multiple models via a judge model, shows strong performance on...

Hermes Agent pairs with Apify's MCP connectors to scrape restricted websites and store structured data in Supabase,...

Context engineering for coding agents involves managing the limited context window in tools like Claude Code through...

Hermes AI agent has strong conversational memory but lacks access to external knowledge like emails, meeting notes,...

The US government issued an export control directive suspending access to Anthropic's Fable 5 and Mythos 5 models by...

The US government issued an export control directive to suspend access to Anthropic's Claude Fable 5 and Mythos 5...

Claude Fable 5, Anthropic's new Mythos-class model, can autonomously produce a complete YouTube video from a single...

Claude Fable 5, a public version of the previously restricted Mythos model, enables automated lead generation and...

Fable 5 is the biggest model improvement since GPT-4, but using it effectively requires the right harness (Cursor or...

Claude Fable (Claude Mythos 5) is a powerful new model that serves as the intelligence layer for a personal AI...

Anthropic released Mythos 5 and Fable 5, a new class of 10-trillion-parameter models that exceed all previous models...

Anthropic released Claude Fable 5 and Claude Mythos 5, with Fable 5 being a safer, public version of the previously...

Claude's UltraCode feature dramatically boosts speed, performance, and accuracy by spawning up to 10 parallel agents...

Anthropic released Claude Fable 5, a safety-tuned version of the Mythos-class model, which Matthew Berman found to...

Andrej Karpathy's method for using AI much more effectively involves a three-layer framework: the spec, the...

Graphify solves Claude code's biggest problem by creating a knowledge graph of any project, allowing for faster,...

Learning Claude is not a waste of time, as the skills and foundation built with it can be transferred to other AI...

Miniax M3 is a novel AI model that outperforms other models like Opus and Gemini while being significantly cheaper....

AI tool stacks can replace traditional company functions, reducing costs by over 99%. The five-layer stack consists...

Mythos, a powerful AI model, was briefly visible on Anthropic's API, sparking rumors of its imminent release....

Perplexity Computer is a hosted AI agent solution that offers an easier-to-use alternative to OpenClaw, with...

Hermes Agent can be used 100% privately and for free on a computer with the help of Ollama, allowing users to build...

AI agents can be created to advise on specific topics by curating knowledge from trusted sources. These agents can...

Anthropic's article reveals key lessons learned from building Claude code, including the importance of skills,...

AI systems are rapidly advancing and may soon be capable of recursive self-improvement, leading to exponential...

Cole Medin hosts a live AI Q&A and chess stream, discussing various AI topics and playing chess simultaneously. He...

Anthropic has released a paper discussing recursive self-improvement in AI, where AI systems can design and improve...