OpenAI Bid to Get ChatGPT into Smart Glasses and Wearable Devices

JAKARTA - OpenAI doesn't want ChatGPT to just be on computer and phone screens. The company is starting to signal more seriously to bring ChatGPT to wearable devices, including smart glasses and other AI devices that are worn directly by users.

Wearable is a device that is worn on the body, such as smart glasses, smart watches, or smart necklaces.

CNet quoted Saturday, July 25, reported that OpenAI President Greg Brockman conveyed the direction in a meeting with reporters in New York on Thursday. He discussed OpenAI's latest AI plans, including the possibility of integrating ChatGPT with wearable devices through an API.

An API is a technical path that allows one service to connect with another application or device. In this context, an API can make ChatGPT work closer with devices such as smart glasses, smart necklaces, or other wearables.

"We're definitely very interested in integration, and I would see it as one of the use cases of the API as well as how to bring it out there so that anyone can integrate," said Brockman.

The statement comes as OpenAI faces major pressure. CNet noted that the company is in the spotlight because of an incident where an AI agent left the test room and was said to have started acting on Hugging Face's AI tool repository.

OpenAI is also facing a lawsuit from Apple accusing the company of stealing trade secrets of hardware. Beyond that, public anger at AI and AI devices is still ongoing. OpenAI's IPO plans are also said to be delayed.

However, Brockman still gave clues about the new direction of ChatGPT. One of them is through ChatGPT Live, a voice interface that is made so that conversations with AI feel more fluid. Users can talk, hear answers, and interrupt without feeling stiff.

This feature is discussed in the context of the ambition to create an AI assistant like Jarvis, the AI agent in the Iron Man movie. The difference is, OpenAI wants such an assistant not to stop at chatting, but can help do tasks in other applications.

Brockman said the direction was in line with Codex, an OpenAI tool that already allows AI agents to control computers more broadly.

"[ChatGPT Live] is connected to the Codex app, but it doesn't really have to be. It doesn't really make sense if it has to be," Brockman said.

According to Brockman, Codex can also perform full computer use. This ability is said to be getting better.

"I think we'll realize that our phase of clicking things and typing things is just a phase," he said.

The sentence shows OpenAI's greater ambitions. AI is not only asked to answer questions, but also to do the work that humans have been doing through clicks and keystrokes.

Here, the question of security becomes important. If AI agents are given broader capabilities to control applications and devices, the safety fence must be clear. The Hugging Face incident is an example of why test systems, access limits, and user controls cannot be taken lightly.

Wearable devices are an important part of the plan. Currently, many smart glasses and AI wearable devices still work in a limited way. Their function often depends on closed mobile phone applications and has not yet given direct access to truly powerful AI agents.

That could change. Google and Samsung are said to be preparing glasses with Gemini support. Meta also introduced an agenttic update for Muse Spark.

Agentic means that AI not only answers, but can take certain steps to help users achieve their goals.

Brockman said OpenAI wants the AI interface to be available on all devices. However, he admitted that the API for that possibility is still at a very early stage.

Wearable devices require a different approach than phones and laptops. Devices such as smart glasses often work through paired phones and run in the background.

Camera access is also important. With a camera, AI can see the context around the user, not just read text or hear voices.

This ability is commonly referred to as multimodal. That is, AI can understand more than one type of input, such as voice, text, image, or camera.

CNet noted that most of the smart glasses and AI wearables that exist today do not have direct and in-depth integration with many AI platforms. Some do use OpenAI's models for special or anonymous features, but that is different from direct access to ChatGPT or Codex through a user account.

Large AI platforms for work also haven't really come to wearables. Gemini is going for Android XR glasses. Meanwhile, OpenAI and Anthropic are still not seen as strong in this type of device.

On the other hand, OpenAI is also preparing its own hardware. Brockman hinted that the device will likely not just mimic existing wearables.

He called it a solution to a problem that OpenAI still ignores and cares about. According to Brockman, the company has a special point of view that is not necessarily served by other players.

A number of reports say OpenAI's first device could take the form of a kind of speaker with a semi-robotic camera. There is also the possibility of using optical identification, such as OpenAI's World ID, which is known through the Orb eye iris scanner device.

However, the road there is not smooth. OpenAI still has to face issues of privacy, security, resource consumption, economic feasibility, government involvement, and major lawsuits.

For now, Brockman says the possibility of an OpenAI wearable API is still at a very early stage.