Tech & Philosophy

The Evolution of AI

An Interactive Visual History

Israel Canyon Saratoga Springs Stake

Journey through seven eras of artificial intelligence — from the earliest symbolic dreams to modern reasoning models — visualized entirely through CSS and SVG art. Hover or tap each card to reveal key milestones.

0 1 1 0 1 0 0 1 01001000 01100101 01101100 01101100 01101111 00100000 01010111 01101111 01110010
1950s – 1960s
The Dream

The birth of artificial intelligence as a field — bold visions, symbolic reasoning, and the first programs that could think.

Key Milestones

  • 1950 — Turing publishes "Computing Machinery and Intelligence"
  • 1956 — Dartmouth Conference coins "Artificial Intelligence"
  • 1958 — Frank Rosenblatt builds the Perceptron
  • 1966 — ELIZA chatbot demonstrates natural language processing
ROOT IF temp>100 IF age>50 THEN A THEN B THEN C THEN D IF symptom = fever AND cough = true THEN diagnosis = "influenza" IF pressure > 150 AND age > 60 THEN risk = "high"
1970s – 1980s
Expert Systems

Hand-crafted knowledge encoded as IF-THEN rules — the era when AI first entered industry through diagnosis, configuration, and decision support.

Key Milestones

  • 1972 — MYCIN diagnoses blood infections using 600 rules
  • 1976 — DENDRAL analyzes chemical compounds
  • 1980 — XCON configures VAX computers, saving DEC $40M/year
  • 1984 — Cyc project begins encoding common-sense knowledge
× × FUNDING WITHDRAWN
1980s – 1990s
The AI Winter

Hype exceeded reality. Funding dried up, projects collapsed, and the field retreated into quiet, persistent research.

Key Milestones

  • 1987 — Lisp machine market collapses
  • 1988 — Lighthill Report's pessimism proven right for the second time
  • 1993 — Expert systems widely abandoned by industry
  • 1997 — Deep Blue defeats Kasparov (a bright spot in the cold)
2000s
Machine Learning Rise

Data-driven approaches overtook hand-coded rules. SVMs, random forests, and statistical models proved that learning from data beats programming by hand.

Key Milestones

  • 2001 — Random Forests paper by Leo Breiman
  • 2006 — Hinton's deep belief networks reignite neural research
  • 2006 — Netflix Prize competition drives ML innovation
  • 2009 — ImageNet dataset created with 14 million labeled images
GPU COMPUTE
2012 – 2017
Deep Learning Revolution

Neural networks went deep. GPU computing unlocked layers upon layers of representation, shattering benchmarks in vision, speech, and games.

Key Milestones

  • 2012 — AlexNet wins ImageNet, halving error rate overnight
  • 2014 — GANs introduced by Ian Goodfellow
  • 2015 — ResNet achieves superhuman image recognition
  • 2016 — AlphaGo defeats world champion Lee Sedol
Q K V Attention Is All You Need ENCODE DECODE
2017 – 2022
The Transformer Era

Self-attention replaced recurrence. A single architecture conquered language, vision, and beyond — spawning GPT, BERT, and the large language model revolution.

Key Milestones

  • 2017 — "Attention Is All You Need" paper introduces the Transformer
  • 2018 — BERT achieves breakthrough in language understanding
  • 2020 — GPT-3 demonstrates few-shot learning with 175B parameters
  • 2021 — DALL-E generates images from text descriptions
👁 💬 🎨 🧠 ✨
2023 – Present
The Age of Intelligence

Reasoning, agency, and multimodal understanding. AI now writes code, creates art, plans actions, and converses with nuance — reshaping every field it touches.

Key Milestones

  • 2022 — ChatGPT reaches 100M users in two months
  • 2023 — GPT-4 demonstrates multimodal reasoning
  • 2024 — AI agents begin writing and debugging software autonomously
  • 2025 — Claude, Gemini, and frontier models achieve expert-level performance