Timeline
A chronological record of programming's AI era — the events, tools, and moments shaping the shift
27 events
The Last Programmers Goes Live
Two days after Block's historic layoff, The Last Programmers launches — a documentary project chronicling the last generation to write code by hand, across website, YouTube, X, and Xiaohongshu.
Source →Block Cuts 40% of Workforce — Largest AI-Attributed Layoff
Jack Dorsey cuts Block from 10,000+ to under 6,000 in a single day, explicitly citing AI as the reason. 'Intelligence tools paired with smaller teams fundamentally change what it means to build a company.' The event that triggered The Last Programmers project.
Source →OpenAI Codex Agent
OpenAI releases a cloud-based coding agent that can work on tasks in parallel — spinning up sandboxed environments to write, test, and iterate on code without human intervention.
Source →Claude Code (Agentic CLI) Launched
Anthropic releases Claude Code, an agentic command-line tool that can navigate codebases, make edits, run tests, and commit changes autonomously. Terminal-native AI coding arrives.
Source →"Vibe Coding" Enters the Lexicon
Andrej Karpathy coins the term "vibe coding" — describing code as something you feel out rather than write. The term instantly resonates and goes viral among developers.
Source →DeepSeek R1 Disrupts the Market
Chinese AI lab DeepSeek releases R1, a reasoning model rivaling top Western models at a fraction of the cost. Open-source AI coding gets a major boost.
Source →Y Combinator: 25% Solo AI Founders
Y Combinator reveals that 25% of its latest batch are solo founders using AI to build their entire product. The one-person startup becomes viable.
Source →Google: AI Writes 25% of New Code
Google CEO Sundar Pichai reveals that AI now generates more than 25% of all new code at Google, reviewed and accepted by engineers. A watershed moment for the industry.
Source →GitHub: 100M Developers, AI-Powered
GitHub surpasses 100 million developers. CEO Thomas Dohmke declares that AI-assisted development is now the default, not the exception.
Source →Claude Computer Use
Anthropic introduces computer use capability for Claude — AI can now interact with desktop applications, browsers, and terminals directly. AI moves from text-based coding to full computer operation.
Source →Bolt, Lovable, v0 — No-Code AI Building
A wave of AI app builders (Bolt, Lovable, Vercel v0) lets anyone describe a web app and get a working prototype in minutes. The line between programmer and non-programmer blurs further.
Source →Cursor Valued at $400M
Cursor's parent company Anysphere raises funding at a $400M valuation. AI-native code editors become one of the hottest categories in tech investing.
Source →"We Stopped Hiring Juniors" Goes Viral
Multiple tech leads share that their teams stopped hiring junior developers because AI handles the tasks juniors used to do. The post sparks massive debate about the future of entry-level programming.
Source →Claude 3.5 Sonnet Released
Anthropic releases Claude 3.5 Sonnet, which tops coding benchmarks while being faster and cheaper than Opus. AI coding quality takes another leap.
Source →GitHub Copilot Workspace
GitHub unveils Copilot Workspace — an AI-native development environment where developers describe tasks in natural language and AI handles implementation.
Source →AI Agents Pass SWE-bench
AI coding agents begin solving real GitHub issues from open-source projects. SWE-bench becomes the benchmark for measuring autonomous coding ability.
Source →EU AI Act Passes
The European Parliament passes the AI Act, the world's first comprehensive AI regulation. Sets rules for AI systems including those used in employment and code generation.
Source →Devin: First AI Software Engineer
Cognition Labs introduces Devin, marketed as the first AI software engineer. It can plan, code, debug, and deploy autonomously. The narrative of AI replacing programmers goes mainstream.
Source →Claude 3 Family Released
Anthropic releases Claude 3 Opus, Sonnet, and Haiku. Opus matches GPT-4 on coding benchmarks while excelling at nuanced reasoning about code architecture.
Source →Google Launches Gemini
Google unveils Gemini, its most capable AI model, designed for multimodal reasoning. AI coding competition intensifies with a third major player.
Source →GPT-4 Turbo + Assistants API
OpenAI DevDay introduces GPT-4 Turbo with 128K context and the Assistants API. Developers can now build AI agents that write, test, and debug code autonomously.
Source →Cursor IDE Gains Traction
Cursor, an AI-native code editor built on VS Code, gains rapid adoption. The first IDE designed from the ground up around AI pair programming.
Source →Meta Releases Code Llama
Meta open-sources Code Llama, a family of code-specialized LLMs. Open-source AI coding tools begin competing with proprietary solutions.
Source →Claude 2 Released
Anthropic releases Claude 2 with 100K context window, capable of processing entire codebases at once. Long-context AI changes how developers interact with code.
Source →Stack Overflow Traffic Drops 35%
Stack Overflow sees a dramatic decline in traffic as developers increasingly turn to AI chatbots for coding answers. A symbolic shift in how programmers learn.
Source →GitHub Copilot X Announced
GitHub unveils Copilot X with chat interface, pull request summaries, and docs integration. AI moves from autocomplete to full development partner.
Source →GPT-4 Released
OpenAI releases GPT-4, a multimodal large language model capable of understanding images and generating sophisticated code. Marks a step-change in AI coding ability.
Source →