Skip to content
forward pass

Claude

Anthropic releases its first assistant, trained with 'constitutional AI' to critique its own answers against written principles; a second frontier chatbot with a different alignment recipe.

category
model
significance
3 of 5
people
Dario Amodei, Jared Kaplan, Yuntao Bai
organisations
Anthropic

what had to happen · 51 events back to 1943

Every event this one built on, transitively, in order. The organism has the same path lit. Direct influences are marked.

00 · One neuron · 1

  1. 1943A logical calculus of nervous activity

I · Foundations · 9

  1. 1948A mathematical theory of communication
  2. 1949Cells that fire together wire together
  3. 1950Programming a computer for playing chess
  4. 1958The perceptron learns
  5. 1959Samuel's checkers program coins 'machine learning'
  6. 1960ADALINE and the least-mean-squares rule
  7. 1965Moore's law
  8. 1969Perceptrons
  9. 1970Reverse-mode automatic differentiation

W1 · The first winter · 2

  1. 1974Werbos applies backpropagation to neural networks
  2. 1980The Neocognitron

II · Connection · 3

  1. 1982The Hopfield network
  2. 1985The Boltzmann machine
  3. 1986Backpropagation

W2 · The second winter · 6

  1. 1988Temporal-difference learning
  2. 1989Q-learning
  3. 1989LeNet reads handwritten postcodes
  4. 1990Finding structure in time
  5. 1991The vanishing gradient problem
  6. 1992TD-Gammon reaches world-class backgammon

III · Statistics and data · 10

  1. 1997Long short-term memory
  2. 1998MNIST and LeNet-5
  3. 1999The first GPU
  4. 2003A neural probabilistic language model
  5. 2006Deep belief networks and the word 'deep'
  6. 2007CUDA
  7. 2009Deep learning moves to GPUs
  8. 2009ImageNet
  9. 2010Rectified linear units
  10. 2010DeepMind is founded

IV · Deep learning · 11

  1. 2012Google Brain's network discovers cats
  2. 2012Dropout
  3. 2012AlexNet wins ImageNet
  4. 2013Deep Q-networks play Atari
  5. 2014Google buys DeepMind
  6. 2014Superintelligence
  7. 2014Attention
  8. 2014Sequence to sequence learning
  9. 2015Batch normalisation
  10. 2015Residual networks
  11. 2015OpenAI is founded

V · Transformers · 8

  1. 2017Attention is all you need
  2. 2017Deep reinforcement learning from human preferences
  3. 2018GPT: generative pre-training
  4. 2019GPT-2 and the model too dangerous to release
  5. 2020Scaling laws for neural language models
  6. 2020GPT-3
  7. 2020Learning to summarise from human feedback
  8. 2021Anthropic is foundeddirect

VI · Everyone · 1

  1. 2022InstructGPTdirect

Anthropic announced Claude on the same day as GPT-4. It had been in private testing for months with companies including Notion and Quora, and it was, by the benchmarks of the time, roughly a GPT-3.5-class model with a longer memory and a manner that users described as careful. What distinguished it was how it had been aligned. Rather than rely only on human labellers ranking outputs, the model was given a written constitution, a list of principles drawn from sources including the UN Declaration of Human Rights, and trained to critique and revise its own responses against it. Humans supplied the principles; the model supplied the feedback.

Constitutional AI, described in a paper the previous December, was the first alternative to RLHF that a frontier laboratory shipped, and it made the values a model was trained on into a document that could be read and argued with. Later versions published the constitution and, in 2025, a longer statement of the model's intended character.

Claude 2 in July 2023 raised the context window to 100,000 tokens, Claude 3 in March 2024 caught GPT-4, and the models became, by 2025, the ones most used for writing code. The March 2023 release is on this timeline as the moment the frontier acquired a second laboratory with a stated method for making models safe.

what it led to · 7 events downstream, through 2026

Built on it directly:

  1. 2024Claude 3 catches GPT-4VI

And, through them, by era:

VI · Everyone · 2
  1. 2024Claude learns to use a computer
  2. 2024The Model Context Protocol
VII · Agents · 4
  1. 2025Claude 4 and Claude Code
  2. 2025MCP is donated to the Agentic AI Foundation
  3. 2026Claude Fable 5 and the Mythos class
  4. 2026Claude Fable 5.1

sources · 2

Trace the lineage of this event on the timeline →Back to the ledger