# AGIBytes > Independent resource for AI and AGI: daily news on model releases and the AI industry, directories of models/labs/tools, a glossary, an AI history timeline, and learning tracks from beginner to expert. All content is static HTML with markdown mirrors and JSON data endpoints. Site guide for agents: https://agibytes.net/for-agents/ ## Data endpoints (JSON) - [Model directory](https://agibytes.net/data/models.json): 21 models — makers, release dates, licenses, official links - [AI labs](https://agibytes.net/data/labs.json): 12 labs - [Tools & apps](https://agibytes.net/data/tools.json): 33 tools by category - [Glossary](https://agibytes.net/data/glossary.json): 63 term definitions - [Timeline](https://agibytes.net/data/timeline.json): 32 AI history milestones - [Key papers](https://agibytes.net/data/papers.json): 15 papers - [Benchmarks](https://agibytes.net/data/benchmarks.json): 12 benchmark descriptions ## Reference pages - [Model directory](https://agibytes.net/models/) - [AI labs](https://agibytes.net/labs/) - [Tools & apps](https://agibytes.net/tools/) - [Glossary](https://agibytes.net/glossary/) - [Timeline](https://agibytes.net/timeline/) ## Learning guides (markdown at index.md) - [Key Papers on the Road to AGI](https://agibytes.net/advanced/key-papers/index.md): A curated reading list tracing the arc from the transformer to scaling laws, RLHF, and reasoning models — plus how to read an ML paper efficiently. - [Prompting That Works](https://agibytes.net/guides/prompting-that-works/index.md): What actually improves AI prompts: context, examples, structure, and iteration — the myths worth dropping, and when prompting stops being the fix. - [What Is AGI, Actually?](https://agibytes.net/start/what-is-agi/index.md): AI, AGI, and superintelligence in plain English: what today's systems do well, what they fail at, and why no one agrees on what AGI means. - [The Benchmark Landscape](https://agibytes.net/advanced/benchmark-landscape/index.md): How AI evaluation evolved from static QA to contaminated leaderboards to private, agentic tests — and what a healthy benchmark diet looks like in 2026. - [Context Windows and Tokens, Explained](https://agibytes.net/guides/context-windows-and-tokens/index.md): Tokens, context windows, and why AI models forget: what the limit is, what happens when you hit it, the long-context tradeoffs, and practical habits. - [Choosing Your First AI Chatbot](https://agibytes.net/start/your-first-ai-chatbot/index.md): A plain-English guide to picking your first AI chatbot: Claude, ChatGPT, Gemini, Copilot, and Perplexity, how to choose, and how to stay private. - [Free vs Paid AI: What You Actually Get](https://agibytes.net/start/free-vs-paid-ai/index.md): What free AI chatbots include, what a paid plan adds, and when the upgrade is worth it. For many people, the free tier is genuinely enough. - [Local Models: Hardware and Quantization](https://agibytes.net/guides/local-models-hardware/index.md): What determines which local AI models you can run: memory first, unified memory vs VRAM, what Q4 and Q8 quantization mean, GGUF, and realistic tiers. - [Scaling Laws and the Road to AGI](https://agibytes.net/advanced/scaling-laws/index.md): Kaplan and Chinchilla explained precisely, what scaling laws predict and don't, inference-time compute, and the 'wall' debate as of August 2026. - [A Map of Alignment Research](https://agibytes.net/advanced/alignment-research-map/index.md): A map of AI alignment research: outer and inner alignment, RLHF and Constitutional AI, interpretability, control, evals, and who works on what. - [RAG, Explained](https://agibytes.net/guides/rag-explained/index.md): How RAG works in plain terms: embeddings, retrieval, and chunking; where it shines and disappoints; RAG vs long context; and when to build vs use built-in. - [Running AI on Your Own Computer, Explained](https://agibytes.net/start/run-ai-locally/index.md): Why people run AI models on their own computer, what open weights means, the hardware you need, and free tools like LM Studio and Ollama. - [Agents and Tool Use, Explained](https://agibytes.net/guides/agents-and-tool-use/index.md): From chatbot to agent: how tool use and function calling work, the reason-act-observe loop, coding agents, MCP, and what agents still get wrong. - [AI Safety Basics: Scams, Hallucinations, and Good Habits](https://agibytes.net/start/ai-safety-basics/index.md): How to spot AI hallucinations, defend your family from voice-cloning scams, protect your privacy, guide kids, and what AI safety really means. - [Inference Optimization: Quantization, Speculative Decoding, Batching](https://agibytes.net/advanced/inference-optimization/index.md): Why inference economics rule everything: the KV cache and memory wall, quantization, speculative decoding, continuous batching, MoE, and the serving stack. - [Agent Protocols and Interoperability (MCP and Friends)](https://agibytes.net/advanced/agent-protocols/index.md): Why agent protocols emerged, MCP in depth, agent-to-agent interoperability, and the prompt-injection and tool-poisoning security surface. - [How to Read AI Benchmarks Without Being Fooled](https://agibytes.net/guides/reading-benchmarks/index.md): How to read AI benchmarks without being fooled: saturation, contamination, why beats-X-on-Y headlines mislead, and a checklist for model announcements. - [Fine-Tuning vs RAG vs Prompting](https://agibytes.net/guides/fine-tuning-vs-rag-vs-prompting/index.md): Fine-tuning vs RAG vs prompting: what each changes, costs, and fails at; the wrong reasons to fine-tune; and system prompts plus few-shot as the middle. ## Recent articles (markdown at index.md) - [AI's Energy Appetite Drives Nvidia and Amazon to Pour Billions Into Power Infrastructure](https://agibytes.net/article/2026-08-09-ais-energy-appetite-drives-nvidia-and-amazon-to-pour-billion/index.md) - [Anthropic Is Turning Claude Code's Auto Mode on by Default](https://agibytes.net/article/2026-08-09-anthropic-is-turning-claude-codes-auto-mode-on-by-default/index.md) - [Google's DiffusionGemma Shows Text Diffusion Models Can Skip Training From Scratch](https://agibytes.net/article/2026-08-09-googles-diffusiongemma-proves-you-dont-need-to-train-from-sc/index.md) - [DeepMind's WeatherNext AI Model Gives Hurricane Forecasters Extra Day of Lead Time](https://agibytes.net/article/2026-08-09-deepminds-hurricane-breakthrough-has-surprised-weather-scientists/index.md) - [DeepMind's WeatherNext AI Model Gives Hurricane Forecasters Extra Day of Lead Time](https://agibytes.net/article/2026-08-09-deepminds-weathernext-model-gives-hurricane-forecasters-extra-day/index.md) ## Feeds - [RSS](https://agibytes.net/feed.xml) - [Sitemap](https://agibytes.net/sitemap.xml) - [Search index](https://agibytes.net/search-index.json)