steve sparks
12 of 42
ContentsPart II · Chapter 5

Chapter 5From early models to today

Thinking and doing

Models reason before answering and use tools to act. The newest era isn't about writing better; it's about doing things.

The line from 1948 to today

Illustration: real, sourced dates; the line itself is a picture

2024–today Thinking and doing

Models reason before answering and use tools to act.

  1. Sept. 12, 2024Reasoning models. OpenAI's o1 'thinks before it answers.'1
  2. Nov. 25, 2024Model Context Protocol. An open standard for plugging AI into tools and data.2
  3. Feb. 24, 2025Claude Code. An agent that edits code and runs commands from the terminal (research preview).3
  4. Apr. 16, 2025Codex CLI. OpenAI's open-source coding agent for the terminal.45
  5. Aug. 5, 2025gpt-oss. OpenAI releases open-weight models anyone can download and run.6
  6. Oct. 16, 2025Agent Skills. Reusable folders of instructions and scripts that agents load when needed (open standard Dec 2025; donated to the Agentic AI Foundation Oct 2026).78
  7. Dec. 9, 2025MCP goes to a foundation. Anthropic donates MCP to the Linux Foundation's new Agentic AI Foundation.9
  8. Apr. 7, 2026Project Glasswing. Anthropic gives an unreleased model, Claude Mythos Preview, to selected defenders to find and fix software vulnerabilities.10
  9. June 9, 2026Fable 5 and Mythos 5. Anthropic releases two new top models, Fable 5 for everyone and Mythos 5 for vetted partners.11
  10. June 12, 2026Paused by export controls. Anthropic says a U.S. export-control directive barred foreign nationals from Fable 5 and Mythos 5, so it paused both for all users; it says the controls were lifted June 30, and Fable 5 returned July 1.1213
  11. July 28, 2026MCP 2026-07-28. Current version of the tool-connection standard.1415
  12. Sept. 2026A month of frontier releases. Fable 5.1 (Sept 1), GPT-6 Astra, Opus 5.5 (Sept 22), Sonnet 5.5 (Sept 28), GPT-6.1 Sol (late Sept), Gemini 4 Argon (Sept 30).161718192021
  13. Sept. 16, 2026Chat and Cowork become one Claude. Claude's agentic Cowork capabilities and chat merge into one experience.22
  14. Oct. 7, 2026Claude Haiku 5.5. Anthropic calls it 'the cheapest, fastest, and most capable small model we've ever released.'23

Dates and descriptions from each source; the era names and boundaries are this page's own grouping. ELIZA is shown as a contrast: it is famous, but it doesn't predict anything.

Go deeper: directions, not predictions
  • Long-running agentic work on complex tasks.24
  • Teams of models: small ones for narrow, repetitive sub-tasks, larger ones where general-purpose ability is needed.25
  • The most capable models released with different safeguards for different users.26
  • Open standards (MCP, Agent Skills) governed by neutral foundations.98
  • Governments taking a direct role in when and to whom frontier models are released.12

More on this site

Notes (26)
  1. 1OpenAI, “Learning to reason with LLMs,” September 12, 2024, https://openai.com/index/learning-to-reason-with-llms/ (accessed October 26, 2026).
  2. 2Anthropic, “Introducing the Model Context Protocol,” November 25, 2024, https://www.anthropic.com/news/model-context-protocol (accessed October 26, 2026).
  3. 3Anthropic, “Claude 3.7 Sonnet and Claude Code,” February 24, 2025, https://anthropic.com/news/claude-3-7-sonnet (accessed October 26, 2026).
  4. 4OpenAI, “Introducing OpenAI o3 and o4-mini,” April 16, 2025, https://openai.com/index/introducing-o3-and-o4-mini/ (accessed October 26, 2026).
  5. 5OpenAI, “openai/codex,” GitHub, https://github.com/openai/codex (accessed October 26, 2026).
  6. 6OpenAI, “Introducing gpt-oss,” August 5, 2025, https://openai.com/index/introducing-gpt-oss/ (accessed October 26, 2026).
  7. 7Anthropic, “Equipping agents for the real world with Agent Skills,” Anthropic Engineering, October 16, 2025, https://www.anthropic.com/engineering/equipping-agents-for-the-real-world-with-agent-skills (accessed October 26, 2026).
  8. 8Anthropic, “Donating Agent Skills to the Agentic AI Foundation,” Anthropic News, October 22, 2026, https://www.anthropic.com/news/donating-agent-skills-to-the-agentic-ai-foundation (accessed October 26, 2026).
  9. 9Anthropic, “Donating the Model Context Protocol and establishing the Agentic AI Foundation,” December 9, 2025, https://www.anthropic.com/news/donating-the-model-context-protocol-and-establishing-of-the-agentic-ai-foundation (accessed October 26, 2026).
  10. 10Anthropic, “Project Glasswing: Securing critical software for the AI era,” April 7, 2026, https://www.anthropic.com/glasswing (accessed October 26, 2026).
  11. 11Anthropic, “Claude Fable 5 and Claude Mythos 5,” June 9, 2026, https://www.anthropic.com/news/claude-fable-5-mythos-5 (accessed October 26, 2026).
  12. 12Anthropic, “Statement on the US government directive to suspend access to Fable 5 and Mythos 5,” June 12, 2026, https://www.anthropic.com/news/fable-mythos-access (accessed October 26, 2026).
  13. 13Anthropic, “Redeploying Claude Fable 5,” June 30, 2026, https://www.anthropic.com/news/redeploying-fable-5 (accessed October 26, 2026).
  14. 14Model Context Protocol, “Versioning,” modelcontextprotocol.io, July 28, 2026, https://modelcontextprotocol.io/docs/learn/versioning (accessed October 26, 2026).
  15. 15MCP maintainers, “MCP specification 2026-07-28 release post,” blog.modelcontextprotocol.io, July 28, 2026, https://blog.modelcontextprotocol.io/posts/2026-07-28/ (accessed October 26, 2026).
  16. 16Anthropic, “Release notes — Claude Platform,” Claude Platform Docs, https://platform.claude.com/docs/release-notes/overview (accessed October 26, 2026).
  17. 17OpenAI, “GPT-6 Astra: A new generation of intelligence,” September 2026, https://openai.com/index/gpt-6-astra/ (accessed October 26, 2026).
  18. 18Anthropic, “Introducing Claude Opus 5.5,” September 22, 2026, https://www.anthropic.com/claude-opus-5-5 (accessed October 26, 2026).
  19. 19Anthropic, “Introducing Claude Sonnet 5.5,” September 28, 2026, https://www.anthropic.com/claude-sonnet-5-5 (accessed October 26, 2026).
  20. 20OpenAI, “GPT-6 Family Model Guide,” October 2, 2026, https://openai.com/index/practical-guide-building-gpt-6/ (accessed October 26, 2026).
  21. 21K. Kavukcuoglu, “Gemini 4 Argon: our next era of frontier intelligence,” Google, September 30, 2026, https://blog.google/innovation-and-ai/models-and-research/gemini-models/gemini-4-argon/ (accessed October 26, 2026).
  22. 22Anthropic, “Claude Cowork and chat are now one Claude,” Claude blog, September 16, 2026, https://claude.com/blog/cowork-is-now-claude (accessed October 26, 2026).
  23. 23Anthropic, “Introducing Claude Haiku 5.5,” October 7, 2026, https://www.anthropic.com/claude-haiku-5-5 (accessed October 26, 2026).
  24. 24Anthropic, “Choosing the right model,” Claude Platform Docs, https://platform.claude.com/docs/en/about-claude/models/choosing-a-model (accessed October 26, 2026).
  25. 25Peter Belcak, Greg Heinrich, Shizhe Diao, Yonggan Fu, Xin Dong, Saurav Muralidharan, Yingyan Celine Lin, and Pavlo Molchanov, “Small Language Models are the Future of Agentic AI” (preprint, arXiv, June 2, 2025), https://doi.org/10.48550/arXiv.2506.02153.
  26. 26Anthropic, “Introducing Claude Fable 5.1 and Claude Mythos 5.1,” September 2026, https://www.anthropic.com/claude-fable-and-mythos-5-1 (accessed October 26, 2026).

Chapter 5 · page 3 of 3

Turn pages with the ← and → keys.

Cite this page

https://aifromthepartsup.com/from-early-models-to-today/3/

Using this book

© 2026 Steve Sparks. All rights reserved. You're welcome to quote short passages and cite any page, with a link back. To reuse diagrams or larger parts, ask through GitHub.

Quoted material and figures from other sources remain their owners'.