Alibaba is moving into the smart eyewear market with devices powered by their own AI models that promote a wide range of AI models and cloud computing. Quark AI Glasses marks the company’s first step into the wearable category and is expected to be launched in China by the end of 2025.
The glasses run on Alibaba’s Qwen large language model and its AI assistant, Quark. Quark is already available as a Chinese app, but this is the first time the company has paired with hardware to contact more users.
The Hangzhou-based company is one of China’s more active AI developers, deploying models designed to compete with the systems of companies such as Openai. By moving to smart glasses, you will join a growth group of high-tech players who are betting on wearables as the next major computing platform along with smartphones.
Push it into the hardware
Quark AI glasses will enter the market, including Meta smart glasses made with Ray-Ban, and models launched by Xiaomi this year. The Alibaba version offers hands-free call, music streaming, real-time translation, transcription encounters, and built-in cameras.
Alibaba operates a wide range of services in China, and glasses connect to their ecosystem. Users can access navigation, make payments through Alipay, compare Taobao prices, and take advantage of other Alibaba-owned platforms, such as mapping and travel booking.
The company outlines several features but does not reveal prices or detailed specifications.
Data Behind the Device
Smart glasses like Alibaba rely on AI systems that can recognize images, interpret contexts, and respond in natural language. Capacity relies on vast amounts of labeled data. This is information reviewed and tagged by humans so that AI can then learn from it.
The process often involves the “loop human” (HITL) system. There, people provide input at the key stages of training and testing. To understand how this actually works, AI News We spoke with Henry Chen, co-founder of Sapien. Sapien is a company that manages a large, distributed workforce for data labeling. Chen discussed the general misconceptions, the demand for skilled contributors, and how the growth of AI in China is affecting the industry.
Misconceptions about HITL
One common belief is that HITL is simply data labeling. Chen said it would be more complicated and includes decisions on edge cases, calls for judgment and ongoing evaluations. “Continuous feedback is what makes HITL work instead of a one-time dataset,” he said.
Another misconception is that the work is low-skilled. Chen said the rise of industry-specific AI has created a demand for domain experts such as doctors, lawyers and scientists to provide knowledge.
Sapien works with 1.8 million contributors in 110 countries. For complex tasks such as contextual understanding and visual recognition, maintaining quality is important. Chen said the company adjusted its incentives to ensure peer verification, contributor reputation tracking, and consistent results.
China’s AI growth and labeling demand
China’s AI sector is expanding rapidly, with demand for data labeling keeping up to the US level. China has its own rules and regulations, but Chen said the types of projects are increasingly similar to other major market types.
With such a large, decentralized workforce, Sapien uses on-chain technology to make payments transparent and tells the community that the project is worth pursuing. By operating without a traditional office, Chen said he will focus on avoiding workplace issues and rewarding contributors for the value they offer.
Automation is changing the labeling of data, but Chen believes humans will remain central to certain types of work. Tasks that involve cultural nuances, irony, rare diseases, niche language, or complex emotions require human review. “Humans shift their focus to long tail data and new vertical domains,” he said, predicting an increase in AI-assisted labels while people handle the most challenging cases.
Strict control is required for sensitive projects, such as IPs of large companies and international organizations. According to Chen, Sapien trains enterprise contributors with veterinarians, uses data minimization and access controls, and follows compliance rules set by clients. The company works under frameworks such as Soc 2 Type 2, GDPR and HIPAA.
Looking ahead
Some expect the need for human labeling to be reduced as AI models improves learning from unlabeled data known as self-teacher learning. Rather than disappearing, Chen sees the role of human contributors change.
“We’re evolving into a more specialized industry,” he said. Sapien said he already does more work on assessing synthetic data and model output. He hopes future projects will focus on curating unique “ground truth” datasets, assessing AI performance, and providing domain-specific expertise.
From glasses to wider AI lace
Alibaba’s smart glasses highlight that AI has moved to daily products. It could be one of many wearable devices in the market by 2025, but it is a combination of Alibaba’s internal language model, existing services and hardware integration that stands out for Chinese users.
At the same time, such products rely on a complex supply chain of complex human expertise, from engineers building models to contributors improving the data used. Companies like Sapien are operating behind the scenes, ensuring that AI systems have the information they need to function more accurately and responsibly.
Whether in the form of smart glasses, virtual assistants, or other unreleased devices, AI-driven hardware is becoming a new way for businesses to directly serve consumers. For Alibaba, Quark AI Glasses is a product launch and a statement about where they are seeing the growth of technology that combines software, hardware and human input.
(Photo: Panos Sakalakis)
See: Alibaba’s AI coding tool raises security concerns in the West
Want to learn more about AI and big data from industry leaders? Check out the AI & Big Data Expo in Amsterdam, California and London. The comprehensive event will be held in collaboration with other major events, including the Intelligent Automation Conference, Blockx, Digital Transformation Week, and Cyber Security & Cloud Expo.
Check out other upcoming Enterprise Technology events and webinars with TechForge here.
