Alibaba launches Qwen Glasses N1 at $450, taking on Meta Ray-Ban
A quiet Tuesday morning turns on two poles of the same argument — who gets to put AI on your face, and whether the boom behind it is lifting the world economy or priming a correction.
Alibaba put its Qwen Glasses N1 on sale today across China's major marketplaces, starting at 2,999 yuan (about $450), with a 3,499-yuan Pro model that adds eye tracking. The base N1 carries a 50-megapixel first-person camera and handles translation across 89 languages; the Pro layers on eye tracking for object recognition and iris-based payments. Alongside the glasses Alibaba launched the Qwen Clip earbuds, co-engineered with Bose, at 1,099 yuan in a first-sale promotion against a 1,299-yuan list price — Bose's chief executive joined the launch event onstage. Pre-orders opened in late September, but same-day availability in China's biggest online stores makes this the highest-profile Chinese challenger to Ray-Ban Meta now shipping as a finished consumer product, with Alibaba's Qwen family doing the model work. Why it matters: smart glasses are the contested interface after the phone, and Meta has had the category nearly to itself — Alibaba just showed up with vertical integration, local pricing, and a distribution machine that can move units on day one. If the N1 sells through, the fight for what sits on your face becomes a US-China contest fought on hardware margins, not just benchmark charts.
The IMF raised its 2026 global growth forecast to 3.1%, up from 3% in July, and says an AI investment boom is what lets the world economy absorb a Middle East war and expensive energy. The fund's World Economic Outlook, released today at its Annual Meetings in Bangkok, holds 2027 growth at 3.4% and credits AI spending with offsetting shocks that would otherwise have cut much deeper — without mitigation measures, the IMF says growth would have fallen to 2.5%. The same document cuts the other way: an AI productivity disappointment would knock US equities 20% and leave global growth 1.1% below baseline by 2028, while the upside scenario adds 0.6% to global GDP by 2029. That duality is precisely the tension we wrote about this morning — IMF warns an AI share crash threatens global financial stability.
MIT Technology Review's latest explainer argues the open-models debate has outgrown its binary framing — and that the commercial incentives on both sides are hiding a genuine trade-off. Grace Huckins' piece lays out why open weights matter (research access, cheaper deployment, local processing of sensitive data, a check on labs' concentration of power) against the harder safety reality: downloaded models come with no usage filters and no way to revoke access, and safety training can be retrained away by anyone with the compute. The counter-case gets a real hearing, too — Hugging Face leaned on an open model to analyze the OpenAI-agent hack on it precisely because the closed models' security filters blocked the queries. The timing matters: the piece lands as Anthropic's report framing Z.ai's freely downloadable GLM-5.3 as a cyberattack risk keeps rippling through policy circles, and both arguments are about to collide in front of regulators who need an answer that isn't a slogan.
What to watch: whether Alibaba dates an international launch for the N1, and how Meta prices Ray-Ban Meta now that a well-resourced rival is on the shelf.
Would you buy AI glasses at $450 if the assistant actually worked — or is the interface still a solution looking for a problem? Tell us in the comments.
