← All artifacts

// Artifact · AI/ML Research

The Climate of Machine Intelligence

An interactive timeline of AI history (1943–2025), framed as booms, winters, and thaws

AI/ML ResearchData VisualizationTypeScriptGitHub Pages
Open full experience

// 01 · Introduction

Introduction

An interactive, publicly hosted web timeline that tells the story of artificial intelligence from 1943 to 2025 through a weather metaphor — booms, winters, and thaws — arguing that AI winters were driven by hardware and funding constraints rather than failures of the underlying theory.

// 02 · Description

Description

The timeline presents 40 curated milestones across eight eras — from Foundations (1943–1955) and the Golden Years (1956–1973), through the First AI Winter, the Expert-Systems boom, and the Second AI Winter (1974–1993), to Quiet Progress (1993–2011), the Deep Learning Boom (2012–2019), and the Generative Era (2020–2025). Milestones are categorized (Theory, Hardware, Industry, Model Release, AI Winter) and filterable by theme; a clickable "barometer" lets visitors jump between eras; and a log-scale training-compute visualization shows how compute flatlined during the winters and exploded afterward, reinforcing the central thesis.

// 03 · Objective

Objective

Created for the AI & ML Timelines activity in the IWU AI/ML program, with two goals: (1) demonstrate the ability to research, synthesize, and communicate seven decades of AI history for a technical-professional audience, and (2) publish it as a genuinely useful, interactive reference rather than a static document.

// 04 · Process

Process

01Researched primary sources and milestone candidates across 1943–2025.
02Developed the "climate" narrative frame and grouped milestones into six eras.
03Designed the interaction model (era barometer, category filters, compute chart).
04Implemented as a TypeScript web app.
05Collected and log-scaled training-compute data for landmark models.
06Deployed to GitHub Pages and iterated on feedback.

// 05 · Tools & Technologies

Tools & Technologies

TypeScriptHTML & CSSJavaScriptGitHub PagesGit/GitHubHistorical sources: peer-reviewed papers, Stanford AI Index, Epoch AI compute data

// 06 · Value Proposition

Value Proposition

Unique Value

Goes beyond a list of dates — it advances an argument (winters were resource constraints, not theoretical dead ends) and lets the audience test that argument interactively against the compute data. It demonstrates research synthesis, front-end engineering, and data-storytelling in a single artifact.

Relevance to My Audience

For hiring managers and collaborators in AI/ML it shows historical literacy — understanding why the field moves in cycles — which directly informs sober judgment about today's generative-AI boom; for peers and learners it serves as a free, reusable teaching resource.

// 07 · References

References

  • McCulloch & Pitts (1943), "A Logical Calculus of the Ideas Immanent in Nervous Activity"
  • Turing (1950), "Computing Machinery and Intelligence"
  • Dartmouth Summer Research Project on AI (1956)
  • Krizhevsky et al. (2012), "ImageNet Classification with Deep CNNs"
  • Vaswani et al. (2017), "Attention Is All You Need"
  • Stanford HAI AI Index Report
  • Epoch AI training-compute dataset