steve sparks
11 of 42
ContentsPart II · Chapter 5

Chapter 5From early models to today

Assistants for everyone

Trained to follow instructions and chat.

The line from 1948 to today

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

2022–2024 Assistants for everyone

Trained to follow instructions and chat.

  1. Jan. 2022Chain-of-thought. Asking a model to write out its steps improves reasoning.1
  2. Mar. 4, 2022InstructGPT. Human feedback teaches a model to follow instructions; a 1.3B tuned model is preferred over 175B GPT-3.2
  3. Oct. 6, 2022ReAct. Reason, act with a tool, observe, repeat: the pattern behind today's agents.3
  4. Nov. 30, 2022ChatGPT. A chat interface brings language models to the public.4
  5. Dec. 2022Constitutional AI. Anthropic trains models to critique their own answers against written principles.5
  6. Mar. 14, 2023GPT-4 and Claude. GPT-4's technical report (Mar 15), and Anthropic opens Claude more broadly after a closed alpha (Mar 14).67
  7. Dec. 6, 2023Gemini. Google says Gemini was 'built from the ground up to be multimodal.'8
  8. Mar. 4, 2024Claude 3 family. Haiku, Sonnet and Opus: pick your balance of intelligence, speed and cost.9
  9. May 13, 2024GPT-4o. Real-time voice and vision in one model.10

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.

Notes (10)
  1. 1Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, Brian Ichter, Fei Xia, Ed Chi, Quoc Le, and Denny Zhou, “Chain-of-Thought Prompting Elicits Reasoning in Large Language Models,” in Advances in Neural Information Processing Systems 35 (2022), 24824–24837, https://doi.org/10.52202/068431-1800.
  2. 2Long Ouyang, Jeff Wu, Xu Jiang, Diogo Almeida, Carroll L. Wainwright, Pamela Mishkin, Chong Zhang, et al., “Training language models to follow instructions with human feedback,” in Advances in Neural Information Processing Systems 35 (2022), 27730–27744, https://doi.org/10.52202/068431-2011.
  3. 3Shunyu Yao, Jeffrey Zhao, Dian Yu, Nan Du, Izhak Shafran, Karthik Narasimhan, and Yuan Cao, “ReAct: Synergizing Reasoning and Acting in Language Models” (preprint, arXiv, October 6, 2022), https://doi.org/10.48550/arXiv.2210.03629.
  4. 4OpenAI, “ChatGPT: Optimizing language models for dialogue (now 'Introducing ChatGPT'),” November 30, 2022, https://openai.com/index/chatgpt/ (accessed October 26, 2026).
  5. 5Yuntao Bai, Saurav Kadavath, Sandipan Kundu, Amanda Askell, Jackson Kernion, Andy Jones, Anna Chen, et al., “Constitutional AI: Harmlessness from AI Feedback” (preprint, arXiv, December 15, 2022), https://doi.org/10.48550/arXiv.2212.08073.
  6. 6OpenAI, Josh Achiam, Steven Adler, Sandhini Agarwal, Lama Ahmad, Ilge Akkaya, Florencia Leoni Aleman, et al., “GPT-4 Technical Report” (preprint, arXiv, March 15, 2023), https://doi.org/10.48550/arXiv.2303.08774.
  7. 7Anthropic, “Introducing Claude,” March 14, 2023, https://www.anthropic.com/news/introducing-claude (accessed October 26, 2026).
  8. 8S. Pichai and D. Hassabis, “Introducing Gemini: our largest and most capable AI model,” Google, December 6, 2023, https://blog.google/technology/ai/google-gemini-ai/ (accessed October 26, 2026).
  9. 9Anthropic, “Introducing the next generation of Claude,” March 4, 2024, https://anthropic.com/news/claude-3-family (accessed October 26, 2026).
  10. 10OpenAI, “Hello GPT-4o,” May 13, 2024, https://openai.com/index/hello-gpt-4o/ (accessed October 26, 2026).

Chapter 5 · page 2 of 3

Turn pages with the ← and → keys.

Cite this page

https://aifromthepartsup.com/from-early-models-to-today/2/

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'.