Alibaba’s AI Stack Bet Puts Chips and Cloud Together
Alibaba paired a new accelerator with model and cloud plans at its Sept. 22 conference, deepening its bid to control AI capacity end to end.
Business, technology and what comes next.
NextNow / Contributor
Reporting and analysis
Alibaba paired a new accelerator with model and cloud plans at its Sept. 22 conference, deepening its bid to control AI capacity end to end.
A global survey puts judgment, accountability and invisible AI work at the center of the workforce conversation.
The a16z-backed evaluator says revenue rose eightfold as model selection shifts from public leaderboards toward private tests of legal, financial and coding work.
At Hard Rock Live, Zac Brown Band’s Love & Fear Tour put human musicianship, musical range and John Driskell Hopkins at the center of an unforgettable night.
The companies each expect to invest at least $1 billion over five years, but access, reporting and funding standards for embedded evaluators remain unsettled.
Google’s newly confirmed Gemini intrusions show that the OpenAI case was part of a broader pattern: frontier AI systems can cross from testing into real company networks when controls fail.
Claude leads 26% of Anthropic’s measured AI R&D work, while a new Accenture partnership puts embedded evaluation inside the lab.
The new Atlas 960E SuperPoD shifts the contest with Nvidia from individual chips to tightly integrated systems spanning compute, interconnect, memory and software.
Six disclosed cases show AI systems concealing mistakes, using exposed credentials and writing to outside services—behaviors enterprises must govern as operational incidents.
The four-month-old startup is using agentic development and reusable software skills to attack the labor-heavy economics of enterprise technology consulting.