3 Years with ChatGPT: How OpenAI “Turned On” AI for Everyone

3 years ago today, on November 30, 2022, OpenAI posted on their official account announcing ChatGPT, their “new AI system which is optimized for dialogue.” This release was major not only because ChatGPT immediately became the best way to talk to a machine, but also because it made talking to a machine ridiculously straightforward.

Instead of interacting with AI through autocomplete, recommendation systems, voice assistants, or developer APIs, with ChatGPT you could just go to chat.openai.com and start talking right away. This simplicity enabled even non-tech people to make AI a part of their everyday lives like never before.

In the months that followed, several big players rushed to respond, and new players appeared. In February 2023, Google announced the beta testing phase for Bard, an “experimental conversational AI service powered by LaMDA,” which later rebranded to Gemini.

The day after the announcement of Bard, Microsoft introduced updated versions of Bing and Edge on February 7, with a post titled “Reinventing search with a new AI-powered Microsoft Bing and Edge, your copilot for the web.” Microsoft highlighted that the new AI-powered Bing used an OpenAI model that was “more powerful than ChatGPT.”

In the last days of February 2023, Meta introduced LLaMA, branded as “a foundational, 65-billion-parameter large language model.” Meta’s strategy was to become the open-source base layer of the LLM ecosystem. A later Wired piece described “Meta AI” as “Zuckerberg’s answer to ChatGPT,” built on LLaMA 2 and integrated into Facebook, Instagram, and WhatsApp with built-in search and image generation.

Later, on Pi Day, March 14, OpenAI introduced GPT-4, their next flagship model. GPT-4 arrived on ChatGPT for Plus subscribers and made ChatGPT even smarter, allowed a larger context window, and enabled image inputs.

On the same day, Anthropic publicly announced Claude. It was framed as a ChatGPT-like assistant trained with “constitutional AI,” emphasizing safety and rule-following as a differentiator rather than raw power.

In late April 2023, OpenAI released “GPT-3.5 with browsing (alpha)” and “GPT-4 with browsing (alpha)” for Plus users, giving ChatGPT independence from its knowledge cutoff. With this update, ChatGPT could search the web and deliver up-to-date information.

One month later, ChatGPT received its iOS app. Along with nearly all web features, the app included speech-to-text powered by Whisper for fast and accurate prompting. The Android app launched in July.

When summer ended, OpenAI published a post titled “ChatGPT can now see, hear, and speak.” This update introduced Voice Mode, enabling natural back-and-forth conversation, and the ability to send images directly to ChatGPT, unlocking even more ways it could help.

In October 2023, OpenAI announced two major updates: built-in web search and DALL-E image generation. Web search no longer needed to be manually enabled. ChatGPT could now detect when a query required current information and automatically search the web. The DALL-E integration allowed users to generate images directly inside chat.

One of the biggest updates for many arrived in February 2024: Memory. ChatGPT could now learn about you and use that information when needed. This made ChatGPT more personal than ever before. From that moment on, no two ChatGPTs were the same.

On May 13, 2024, OpenAI held a livestream introducing GPT-4o, with the “o” standing for omni. The livestream showcased a world of multimodal capabilities inside ChatGPT. The new voice mode enabled far more natural conversations with real-time interruption, camera access, and screen sharing. The macOS app was also introduced.

Since then, OpenAI has continued releasing updates regularly, such as GPT-5, Codex, Deep Research, Agent, Connectors, Apps, Pulse, Shopping Research and more. Each release unlocks entirely new ways to use ChatGPT and makes it even more personal, capable and useful as an everyday assistant.

And that’s the entire timeline of how AI transformed from a research topic behind the scenes into your personal assistant. I expect to see completely new use cases we can’t even imagine by the next birthday of ChatGPT.

As always, I’ll be covering everything important happening in the AI world with all of you.

Stay tuned, and keep reading!