OpenAI Employee Reveals Google Had ChatGPT-Like AI Before OpenAI’s Launch

OpenAI’s ChatGPT changed the AI landscape when it launched nearly four years ago, forcing Google to accelerate its own AI efforts. However, according to Thibault ‘Tibo’ Sottiaux, a former Google AI team member now at OpenAI, Google had a similar chatbot almost a year before ChatGPT’s release but chose not to launch it.

Key Highlights

• OpenAI’s ChatGPT launched nearly four years ago and quickly changed the AI landscape.
• Google developed a similar chatbot called LMChat about a year before ChatGPT but did not release it.
• Google withheld its chatbot due to concerns about disrupting its search business and accuracy issues.
• ChatGPT reached one million users in five days and now has about one billion monthly active users.
• Google’s Gemini has nearly 950 million users, with both companies leading AI development alongside Anthropic.

Google’s Early AI Chatbot Development

Tibo responded to a post on X by AI engineer Cheng Lou, who referenced an interview with Jeff Dean, Google’s chief scientist. Dean had mentioned that Google had developed an internal chatbot before ChatGPT but did not release it, believing it was not better than traditional search. Tibo confirmed this, stating, “I was part of that team. Basically ChatGPT one year before it came out. Called LMChat and then another codename.”

Tibo claimed that Google blocked the release of its chatbot due to concerns it could disrupt its core search business. He added, “Google was too nervous to release it and DeepMind was blocked from shipping products that could disrupt Google.” Tibo left Google DeepMind in July 2024 to join OpenAI.

Google played a key role in developing the architecture behind large language models. In 2017, Google published the “Attention is All You Need” paper, introducing the Transformer architecture. This technology underpins models like ChatGPT, where GPT stands for Generative Pre-trained Transformer.

Concerns Over Accuracy and Business Impact

Despite its technical leadership, Google decided not to release its chatbot. After ChatGPT’s debut, Google responded with Bard and later released Gemini. In 2022, CEO Sundar Pichai and Jeff Dean acknowledged that Google had similar AI capabilities to OpenAI but cited higher risks. They said users expected reliable answers from Google, and the company was concerned about AI “hallucinations,” where chatbots generate inaccurate information. This remains a challenge for AI systems today.

Before ChatGPT, Google was widely regarded as the leader in AI research, with teams like Google Brain, DeepMind, and Google Research. The company developed advanced models such as LaMDA, PaLM, and Meena. However, these systems were often kept internal or not fully integrated into Google’s products due to concerns about accuracy, safety, and the risk of disrupting its search business.

Market Impact and User Growth

OpenAI’s launch of ChatGPT changed the pace of AI development. ChatGPT reached one million users in five days and 100 million in two months. Currently, ChatGPT has about one billion monthly active users, while Google’s Gemini has nearly 950 million.

Today, Google and OpenAI are leading AI model development, alongside Anthropic. The decision not to release Google’s early chatbot has sparked repeated debate online. Some suggest that if Google had launched its chatbot, OpenAI might not have achieved the same prominence in the AI field.