top of page

The Dawn of Dedicated AI: OpenAI and Google's Push for Physical Devices and Desktop Interfaces

Updated: 7 days ago

Tech giants OpenAI and Google are at the forefront of this transformation, each pursuing distinct yet complementary visions for how humans will interact with advanced AI agents. This article delves into their respective initiatives, exploring the motivations, current developments, and potential future impacts of these dedicated AI hardware and software solutions.

Ā 

OpenAI's Vision: The Jony Ive Collaboration and 'Operator' Agent

OpenAI, renowned for its groundbreaking large language models like GPT, is venturing into the realm of physical hardware through a significant collaboration with former Apple design chief Jony Ive and his design firm, LoveFrom. This partnership aims to create a dedicated AI device, signaling a strategic shift towards tangible, user-centric AI experiences.

Ā 

The 'io' Project: A New Paradigm for AI Interaction

The collaboration, initially dubbed the 'io' project, saw the team from io Products, Inc. merge with OpenAI in July 2025 [1]. Jony Ive and LoveFrom have taken on deep design and creative responsibilities across OpenAI, emphasizing the importance of design in shaping the future of AI interaction. Sam Altman, OpenAI's CEO, highlighted that great tools require work at the intersection of technology, design, and understanding people and the world, and that Jony Ive and his team are uniquely positioned to deliver this [1]. Ive himself expressed a growing sense that his 30 years of experience have led him to this moment, grateful for the opportunity to be part of such an important collaboration.

Ā 

While specific details about the physical device remain under wraps, reports suggest that OpenAI is targeting a launch in the second half of 2026 [2, 3, 4]. Early prototypes are described as small, screenless, and potentially wearable, hinting at a focus on seamless, intuitive interaction rather than traditional screen-based interfaces [2]. The ambition is to create a device that fundamentally changes how we interact with AI, moving towards a more natural and integrated experience. However, the project has faced challenges, including compute shortages and design struggles, leading to some delays [5, 6].

Ā 

A potential wearable form factor for the OpenAI/Jony Ive device
Figure 1: A potential wearable form factor for the OpenAI/Jony Ive device.

OpenAI Operator: A Desktop Interface for AI Agents

Beyond physical devices, OpenAI is also developing advanced software interfaces to enable AI agents to interact with digital environments. The introduction of OpenAI Operator marks a significant step towards AI agents that can control entire computers and perform tasks on behalf of users [7, 8]. Operator is trained to interact with graphical user interfaces (GUIs), understanding buttons, menus, and text fields on a screen. This capability allows AI agents to navigate complex software environments, automate workflows, and potentially revolutionize productivity by offloading mundane digital tasks to AI.

Ā 

OpenAI Operator demonstrating computer control features
Figure 2: OpenAI Operator demonstrating computer control features.

Google's Approach: Project Astra and the Universal AI Assistant

Google, through its DeepMind division, is also heavily invested in shaping the future of AI interaction, with a strong focus on multimodal AI assistants and new form factors. Project Astra stands as a cornerstone of this vision, aiming to create a universal AI assistant that is intelligent, context-aware, and capable of planning and taking action across various devices.

Ā 

Project Astra: Multimodal AI Across Devices

Project Astra is a research prototype that explores breakthrough capabilities for Google products, with the ultimate goal of transforming the Gemini app into a universal AI assistant [10, 11]. Key capabilities of Project Astra include natural interaction with improved audio input and output, proactive responses, and context-aware dialogue that ignores distractions. It also boasts action intelligence, helping users by reacting to their needs and taking actions on their behalf, including tool use with applications like Search, Gmail, Calendar, and Maps [10].

Ā 

A significant aspect of Project Astra is its intelligent personalization, learning and retaining user preferences to provide tailored answers and explanations. It integrates multimodal memory, allowing it to remember key details from past interactions and retrieve relevant content like PDFs or recipes to provide assistance [10].

Ā 

Project Astra is designed to work across various devices, including Android phones and prototype glasses [10]. This cross-device memory ensures that users can switch between devices and continue conversations seamlessly. The prototype glasses are particularly noteworthy, as they represent Google's long-term plan for hardware with augmented reality and multimodal AI capabilities [12, 13]. Reports suggest that Google's AI glasses could see an earliest release in Q4 2026, with some models shipping in Fall 2026 [14, 15].

Ā 

A user wearing Google Project Astra glasses
Figure 3: A user wearing Google Project Astra glasses.

Project Mariner: Multitasking AI Agents for Digital Environments

Complementing Project Astra, Google has also been exploring how agentic capabilities can help people multitask with Project Mariner. This research prototype focuses on human-agent interaction within browsers, allowing a system of agents to complete multiple tasks simultaneously, such as looking up information, making bookings, and conducting research [11]. Project Mariner's computer use capabilities are being integrated into the Gemini API, with plans to bring more of its features to Google products throughout the year.

Ā 

Comparison and Future Outlook

Both OpenAI and Google are pushing the boundaries of AI interaction, albeit with slightly different immediate focuses. OpenAI, with its collaboration with Jony Ive, appears to be emphasizing a novel, perhaps minimalist, physical device that redefines the human-AI interface. Their Operator agent, meanwhile, focuses on deep integration with existing desktop environments.

Ā 

Google, through Project Astra, is building a more pervasive, multimodal AI assistant that seamlessly integrates across various form factors, from phones to smart glasses, and leverages its extensive ecosystem of applications. Project Mariner further enhances this by enabling advanced multitasking within digital interfaces.

Ā 

The race to create dedicated physical devices and advanced desktop interfaces for AI agents is well underway. OpenAI and Google are investing heavily in these areas, recognizing that the future of AI lies not just in powerful models, but also in intuitive and integrated ways for humans to interact with them. As these initiatives mature, we can expect a profound transformation in how we work, communicate, and experience the digital and physical worlds, with AI becoming an even more integral part of our daily lives.

Ā 

References

[1] OpenAI. (2025, July 9). A letter from Sam & Jony. https://openai.com/sam-and-jony/Ā 

[2] MediaCopilot.ai. (2026, January 20). OpenAI targets second half of 2026 for its first AI device. https://mediacopilot.ai/openai-ai-device-2026-jony-ive-davos/

[4] Axios. (2026, January 19). Exclusive: OpenAI aims to debut first device in 2026. https://www.axios.com/2026/01/19/openai-device-2026-lehane-jony-ive

[6] Reddit. (2025, October 20). OpenAI and Jony Ive are struggling to bring their AI device to market. https://www.reddit.com/r/technology/comments/1nzhrlh/openai_and_jony_ive_are_struggling_to_bring_their/

[7] Reddit. (Unknown). OpenAI launches Operatorā€”an agent that can use a computer for you. https://www.reddit.com/r/OpenAI/comments/1i89lt0/openai_launches_operatoran_agent_that_can_use_a/

[8] Reddit. (Unknown). OpenAI's new ChatGPT Agent can control an entire computer and do tasks for you. https://www.reddit.com/r/technews/comments/1m303pr/openais_new_chatgpt_agent_can_control_an_entire/

[9] DeepMind. (Unknown). Project Astra. https://deepmind.google/models/project-astra/

[10] Google Blog. (2025, May 20). Our vision for building a universal AI assistant. https://blog.google/innovation-and-ai/models-and-research/google-deepmind/gemini-universal-ai-assistant/

[11] TechCrunch. (Unknown). See what Google's Project Astra AR glasses can do (for a select few beta testers). https://techcrunch.com/video/see-what-googles-project-astra-ar-glasses-can-do-for-a-select-few-beta-testers/

[12] Reddit. (2025, July 20). Project Astra Prototype Glasses ā€” Will the display always be.... https://www.reddit.com/r/augmentedreality/comments/1heslga/project_astra_prototype_glasses_will_the_display/

[13] Digital Applied. (2026, May 20). Google Intelligent Eyewear: Gemini AI Glasses Fall 2026. https://www.digitalapplied.com/blog/google-intelligent-eyewear-ai-glasses-fall-2026

[14] 36kr. (2025, November 27). Exclusive Report: Google Restarts AI Glass Project. https://eu.36kr.com/en/p/3571198300289921


Video References

Video 1: OpenAI's First Hardware Device Revealed: Sam Altman...

Video 2: Project Astra: Our vision for the future of AI assistants

Video 3: Google I/O 2026: The Smart Glasses Reveals Everyone's...

Comments

Rated 0 out of 5 stars.
No ratings yet
Add a rating
bottom of page