AI Evolution

2017 to present — the architectures that reshaped machine intelligence

Custom CSS — spine + dots + cards · /visualcave AI evolution timeline, rich CSS, color-coded by era
2017

Transformer — Attention Is All You Need

Google Brain · Vaswani et al.

Self-attention replaces recurrence entirely. Enables massively parallel training on long sequences. Every production LLM today — GPT, Claude, Gemini — descends from this architecture.

foundational
2018

BERT — Bidirectional Encoder

Google · Devlin et al.

Masked language modeling and next-sentence prediction. Bidirectional context captures richer semantics than left-to-right GPT. Transfer learning for NLP becomes standard practice.

encoder
2020

GPT-3 — 175B Parameters

OpenAI · Brown et al.

Scale reveals in-context learning as an emergent property. Few-shot prompting replaces fine-tuning for many tasks. API-first deployment establishes the LLM-as-service model.

scale
2022

ChatGPT — RLHF at Scale

OpenAI · InstructGPT lineage

Reinforcement Learning from Human Feedback teaches models to follow instructions and refuse harmful requests. Consumer adoption goes mainstream: 100M users in 60 days.

alignment
2023

GPT-4 + Claude 2 — Multimodal Reasoning

OpenAI · Anthropic

Vision inputs, 100K+ context windows, and bar-exam-level reasoning. Models begin passing professional licensing exams. Constitutional AI and process supervision improve safety.

multimodal
2024

MCP — Model Context Protocol

Anthropic · Open standard

Standardises tool and resource interfaces so agents can compose capabilities across vendors without custom glue code. Marks the shift from chatbots to autonomous agentic systems.

agentic