Why did Google hold back its AI chatbot?

Google CEO Sundar Pichai stated that an early internal version of their AI chatbot was deemed too toxic for public release, indicating a deliberate choice to uphold higher product quality standards.

Was Google caught unprepared for the AI chatbot wave?

Sundar Pichai claims Google was not unprepared but deliberately timed its AI chatbot releases, holding back advanced versions like LaMDA due to concerns about their toxicity and refinement.

How did ChatGPT's success affect Google's AI strategy?

Following ChatGPT's viral success in late 2022, Google experienced internal pressure, prompting a 'code red' and acceleration of its AI development efforts, leading to the eventual release of products like Bard and Gemini.

Home / Technology / Google CEO: AI Chatbot 'Too Toxic' for Release

Google CEO: AI Chatbot 'Too Toxic' for Release

10 Apr

•

Summary

Google CEO revealed an early AI chatbot version was too toxic.
The company deliberately held back advanced AI models for quality.
Google rushed its AI response after ChatGPT's viral success.

Google CEO: AI Chatbot 'Too Toxic' for Release

Credits: Nguyen Hung Vu from Hanoi, Vietnam, CC BY 2.0, via Wikimedia Commons

Google CEO Sundar Pichai revealed that the company deliberately delayed releasing its advanced AI chatbot technology, an early iteration of LaMDA, because an internal version was deemed "too toxic" for public release. Pichai clarified on a podcast that while an engineer interacted with a version similar to early ChatGPT, Google chose not to ship it publicly due to its unrefined state and toxic outputs.

This strategic decision, Pichai emphasized, was a conscious choice driven by Google's commitment to a higher product quality standard, contrasting with the rapid rollout of competitors' models. The company faced internal pressure and a "code red" following ChatGPT's viral success in late 2022, prompting an accelerated AI development strategy that led to the eventual releases of Bard and Gemini. Pichai drew parallels to Google's past navigation of internet innovations, suggesting that their approach was a deliberate timing call rather than a sign of being unprepared for the generative AI wave.

Disclaimer: This story has been auto-aggregated and auto-summarised by a computer program. This story has not been edited or created by the Feedzop team.