Google Gets Its Act Together

Google seems to have finally learned its lesson about announcing products that aren’t yet ready for release to the public. At its annual I/O developer conference on Tuesday, the company launched a flurry of products, many of which—including its chatbot-style search tab AI Mode, its new Veo video model and its coding assistant Jules—were immediately available for users.

Reactions were largely positive. Developers told us that they were excited to try out Google’s new AI models that can understand and produce audio, as well as generate video with sound effects and dialogue. That video model—Veo 3—places Google markedly ahead of rivals also making video AI, including OpenAI and Runway.

The upbeat mood fueled excitement among developers about Google’s progress that had already been building for months. Developers we spoke with said that Google’s release of its Gemini 2.5 Pro model in March marked the start of a noticeable shift in how developers feel about its conversational AI models. One said that Google’s push to quickly make the model available in popular coding apps like Cursor was one big driver of this vibe shift.