Keep pulling the thread on Logan Kilpatrick.
Logan Kilpatrick predicts that within the next 12 months, the operational duration of autonomous AI agents without human intervention will increase from hours to days or weeks.
Google is laying the groundwork to allow developers to create their own agents using the Gemini API and plans to release new Google-hosted agents.
Google AI Studio's "Yap to App" feature uses Gemini to interpret unstructured voice input, formulate it into a coherent application idea, and then build the app.
A demonstration version of Google AI Studio Mobile has been shown internally to be capable of running local, on-device models.
Logan Kilpatrick predicts that within the next 12 months, AI-powered development tools will enable users to build any software they can imagine.
Logan Kilpatrick predicts that Google's various generative media models, such as Veo and Lyria, will be consolidated into the mainline version of the Gemini model in the future.
Google, Google AI Studio, and DeepMind are making a significant strategic bet on the Anti-Gravity team to advance their coding tools for developers.
Demand for AI compute across all of Google's AI products significantly exceeds the available supply.
Logan Kilpatrick predicts that over the next two to three years, the entire AI ecosystem will face a challenge where demand for AI outstrips supply, forcing businesses to strategically ration their use of AI tokens.
Logan Kilpatrick predicts that the field of robotics is approximately 12 to 18 months away from significant breakthroughs, similar to the state of coding models 18 months prior.
Google recently released two new versions of its research agent, named Deep Research and Deep Research Max, which offer more rigorous and in-depth research capabilities.
Logan Kilpatrick believes the "era of agents" has arrived, driven by platform progress at companies like Google.