From Socks to Smart Agents: DeepMind Gemini Robots Master Household Tasks
What Gemini Robotics 1.5 can do
Google DeepMind has introduced Gemini Robotics 1.5, a system that brings advanced reasoning and physical adaptability to robots. The demonstrations show machines performing familiar household chores, from sorting laundry by color to choosing appropriate items to pack for a rainy day.
Reasoning combined with real-world adaptability
The key upgrade is the integration of reasoning capabilities with sensors and manipulation skills. That means robots can evaluate ambiguous situations, weigh options and act in environments that are not tightly controlled. Tasks that once seemed to require human judgment are now within reach for these systems.
Web-informed actions and instant knowledge sharing
One striking capability is the ability to pull information from the web and apply it to local tasks. A robot can look up recycling rules for London and apply them at home, while another robot in Tokyo could do the same for local regulations. Those agents can then share what they learn, creating a network effect where localized knowledge scales quickly across devices.
Agentic AI moves beyond the home
This blend of agency and connectivity is not limited to domestic settings. Corporations are testing similar agentic AI to manage multi-step workflows. For example, Citi recently began trials of AI agents that can navigate internal systems and complete multi-stage tasks, reducing the need for employees to switch between multiple apps and interfaces.
Economic and geopolitical stakes
Policy discussions are already catching up. At the Global AI Summit, experts warned that AI advances could affect as much as 60% of jobs in advanced economies. A robot that folds socks appears harmless until it is seen as one component of a broader automation wave that could reshape labor markets, required skills and international trade balances.
A personal reaction
There is a tension between excitement and unease. On one hand, labor-saving domestic robots promise convenience and freed time. On the other hand, the rapid pace at which agentic AI gains new skills highlights gaps in regulation and raises questions about how societies will manage economic displacement. The laundry basket is mundane, but in 2025 it has become a symbol of how fast machines are closing the gap with human tasks.