Google's most advanced AI tools aren't meant for us

Google's AI showcases at I/O 2025 felt like a retread of past advancements without new transformative features.

: During the I/O 2025 event, Google's offerings centered around AI capabilities were mainly a reiteration rather than groundbreaking developments. Notably, tools like AI Mode, Jules, Flow, and the new $250 AI Ultra plan were highlighted. Google's past innovations, such as Night Sight and Call Screening, had a more significant impact than this year's AI introductions. The event appeared more focused on promoting Google's AI potential to developers without introducing widely applicable user-oriented advancements.

Google's most advanced AI tools, like Veo 3 and Project Astra, are currently not designed for the general public. These tools are primarily being made available to enterprise users, developers, and select testers. While the demos at Google I/O 2025 showcased impressive capabilities—such as real-time video understanding and lifelike video generation—the tools remain locked behind closed access, raising questions about inclusivity and long-term vision.

Veo 3, Google's latest AI video generator, can produce highly realistic 1080p videos with smooth camera motions, dialogues, and sound effects. However, it is available only to U.S.-based subscribers of Google AI Studio Ultra, which costs $249/month. This pricing model and geographic restriction limit experimentation and adoption by independent creators and educators who could benefit from such a powerful tool.

Project Astra, a universal AI agent developed by Google DeepMind, was introduced with much hype. It can interpret live video input, answer contextually, and remember previous interactions. Yet, it is still in the prototype phase and unavailable to the public. Google’s strategy seems to prioritize internal testing and partner collaborations before a broader release, possibly to ensure safety and prevent misuse.

This gated access reflects a broader trend among tech giants: keeping their most powerful AI systems restricted either for monetization or to avoid public scrutiny and misuse. While this can ensure a more controlled rollout, it also exacerbates the divide between tech elites and general users, who remain dependent on less capable versions.

The situation underscores a philosophical and commercial shift in how AI is distributed—from open-source or consumer-first tools to proprietary, enterprise-level platforms. Critics argue that this hoarding of innovation by a few companies stifles democratization, creativity, and educational opportunity.

Sources: TechCrunch, The Verge, Wired, Google Blog, Ars Technica

AITechnologySoftware