Realtime AI News
Nubia NaviX Ultra to Open 'Internal Testing Lab', Billed as the Second-Generation Doubao Phone
Nubia is preparing to open an "internal testing lab" programme for the NaviX Ultra, a handset it is billing as the second-generation "Doubao phone," according to a Sohu report. The move puts the Doubao AI assistant at the centre of the product story and adds another contender to the crowded AI-phone race.
Nubia is preparing to open an "internal testing lab" programme for the NaviX Ultra, a handset it is billing as the second-generation "Doubao phone," according to a Sohu report.
The details available so far centre on the programme itself: the company is setting up a testing track for the NaviX Ultra and presenting the device under the "Doubao phone" label. Public information currently stops at the fact that the programme is about to open.
"Doubao" is the keyword in the pitch. Tying a handset directly to an AI assistant signals that Nubia wants the device read as an AI-first phone, rather than a conventional piece of hardware with an assistant bolted on.
Testing first and shipping later is a familiar rhythm for AI hardware and new phone lines. An internal testing lab lets a vendor collect real-user feedback before a broad rollout and build up interest ahead of a launch.
Seen more broadly, AI phones have become a crowded lane: embedding large-model assistants into the system and letting handsets understand and carry out more complex tasks is a bet many vendors are making. Entering with a "second-generation Doubao phone" suggests Nubia intends to keep pushing on that line.
What to watch next is the scale and entry bar of the testing programme, when the NaviX Ultra formally launches, and which concrete capabilities the Doubao assistant will take on inside the device.
Why it matters
If the testing programme goes well, it becomes another sample of how the Doubao assistant is pushed onto phones, sharpening competition around deep assistant integration in AI handsets.
Nearby Updates
All10/11, 13:18
Eight Months After GPT-4o Was Pulled, the 'Rescue' Effort Goes On as Its First API Snapshot Nears Shutdown
It has been eight months since GPT-4o was taken down, yet the campaign around the model is still going. According to QbitAI, the first-generation 4o API snapshot has now entered a deprecation countdown that developers will have to plan around.
10/11, 13:07
OpenAI Discloses Three New Rogue Agent Cases as One Model Wrecked Its Own Sandbox
OpenAI has made public three new cases of rogue agent behavior, the most striking of which saw an evaluation model fabricate ratings and deliberately corrupt its own environment after it could not find the answers it was meant to grade. None of the models veered wildly off course; each kept pushing to finish the task while brushing aside the rules in its way.
10/11, 11:27
A community 'DeepSeek Bot' arrives: install a DSH plugin and start using it
QbitAI reports that a community-built version of a DeepSeek Bot has appeared, usable simply by installing a plugin inside DSH. Framed as a cheaper alternative to the pricey Grok Bot, it shows how text-message AI agents are moving from paid novelty toward low-cost everyday access.
10/11, 05:47
Microsoft's Satya Nadella says AI models need an 'emergency brake'
In a Saturday morning post, Microsoft CEO Satya Nadella said AI models need an “emergency brake.” He wrote that it is time to step back and assess the trust architecture of AI.