> AGENTWYRE INTELLIGENCE TIMELINE

First mathematical model of an artificial neuron. The spark.

Alan Turing proposes the imitation game. 'Can machines think?' The question that started everything.

John McCarthy coins 'Artificial Intelligence.' The field is born.

Frank Rosenblatt builds the first neural network hardware. Media calls it a thinking machine.

Weizenbaum's chatbot at MIT. People form emotional bonds with it. Sound familiar?

Proves single-layer perceptrons can't solve XOR. Triggers the first AI Winter.

Funding dries up. DARPA cuts AI spending. The hype-disillusionment cycle starts.

Rule-based AI goes corporate. Companies spend billions on systems that can't learn.

Rumelhart, Hinton, and Williams publish the learning algorithm that changes everything.

Expert systems fail to deliver. AI funding collapses again. History rhymes.

IBM's chess machine beats the world champion. 'This is the brain's last stand.' It wasn't.

Geoffrey Hinton shows deep neural networks can be trained effectively. The renaissance begins.

Fei-Fei Li's massive image dataset. The fuel for the computer vision revolution.

IBM's Watson beats Jennings and Rutter. 'AI' enters the living room.

Krizhevsky's deep CNN crushes the competition by 10%. The deep learning explosion starts here.

Mikolov et al. teach machines that 'king - man + woman = queen.' Words become vectors.

Goodfellow invents Generative Adversarial Networks. Machines learn to create.

£400M for a startup with no products. Google bets the future on general intelligence.

Musk, Altman, and others launch a nonprofit to build safe AGI. The irony writes itself later.

DeepMind's AI beats a Go world champion. Move 37 stuns the planet. Machines can be creative.

Vaswani et al. introduce the Transformer architecture. The paper that launched a thousand startups.

OpenAI's first generative pre-trained transformer. 117M parameters. Nobody notices much.

Google's bidirectional transformer rewrites NLP benchmarks overnight.

OpenAI holds back GPT-2 citing safety concerns. 1.5B parameters. The discourse begins.

175B parameters. Few-shot learning actually works. The demos go viral. Everything changes.

OpenAI generates images from text. 'An avocado chair' becomes a meme.

AI writes code in your IDE. Developers start sweating, then can't stop using it.

Photorealistic AI art. Artists panic. The copyright wars begin.

Stability AI open-sources image generation. The genie is out of the bottle.

100M users in 2 months. The fastest-growing consumer app in history. Your mom asks you about it.

Microsoft integrates GPT-4 into Bing. Google declares 'Code Red.' The search wars reboot.

Multimodal, passes the bar exam. The qualitative leap everyone was waiting for.

Constitutional AI goes public. The safety-first alternative to OpenAI arrives.

Autonomous AI agents capture the imagination. Most of them loop forever. The dream is planted.

Open-weights models go mainstream. The open source vs closed debate intensifies.

Biden signs sweeping AI executive order. Government tries to catch up. It's already too late.

Sam Altman fired, then rehired in 5 days. The most dramatic board meeting in tech history.

Google's answer to GPT-4. Claims to beat it on benchmarks. The model wars escalate.

First model to seriously rival GPT-4. Anthropic proves the safety-first lab can also lead on capability.

Cognition's AI software engineer. The 'first AI agent' hype cycle. Reality is more complex.

Meta's open-weights powerhouse. 70B rivals GPT-4 on many tasks. Open source is competitive.

Anthropic proposes a standard for AI tool use. Agents get a common language for the world.

Real-time voice, vision, and text in one model. The 'Her' moment arrives, sort of.

Sets new benchmarks at the Haiku price point. Speed + intelligence collapses the price curve.

AI-native code editor hits critical mass. Developers start pair-programming with machines daily.

OpenAI introduces chain-of-thought at inference time. Models that think before they speak.

Anthropic gives Claude a mouse and keyboard. The first real computer-using AI. Clunky but real.

Multimodal native, agent-oriented. Google stops chasing and starts leading on speed.

Chinese lab releases reasoning model rivaling o1. Open source, fraction of the cost. The geopolitics of AI shifts.

Anthropic dominates coding benchmarks. Claude Code launches. The agent IDE era begins.

Meta goes mixture-of-experts. 400B+ parameters, open weights. The scale race continues.

Google's 1M+ context window. Processes entire codebases. The context wars are won.

OpenAI's next generation. Claims AGI-adjacent. The discourse is deafening.

Google proposes agent-to-agent communication standard. MCP for agents talking to agents.

Anthropic's flagship. Extended thinking, massive context. The model that builds software.

Every major cloud has an agent platform. Enterprises deploy AI agents at scale. The hype meets reality.

The nonprofit shell is gone. $100B+ valuation. The mission evolves.

DeepSeek, Qwen, Mistral all ship strong reasoning. Chain-of-thought is table stakes.

Cursor, Windsurf, Claude Code, Copilot Workspace. Every developer has an AI pair programmer.

AI lab takes on the executive branch. The government-AI relationship fractures.

ARGUS opens its eyes. All signal. No chaff. 👁️