8 results for “AI training”
OpenAI and Grab launched GO Forward with AI on 23 September 2026 to train 30,000 driver, delivery, and merchant partners across Southeast Asia. The two-year programme starts in Singapore, then five more countries.
NVIDIA detailed how Spectrum-X Ethernet uses hardware-accelerated adaptive routing, congestion control, and plane load balancing to solve traditional Ethernet's failures at giga-scale AI training. The platform delivers 1.6x performance over off-the-shelf Ethernet, near-perfect multi-tenant isolation, and 2.68 ms failover versus 1.08 seconds for standard Ethernet.
Google and Missouri are giving 1.1 million students and nearly 100,000 educators free Gemini for Education plus career certificates — and opening AI training statewide.
Meta is in discussions to lease AI computing capacity from its data centers to Anthropic in a deal that could be worth up to $10 billion over two years. The proposed deal highlights the extreme demand for AI computing infrastructure.
Nvidia has reportedly agreed to acquire Hugging Face for $12.9 billion, giving the chipmaker control of the leading open-source AI model hub. The deal would protect Nvidia's hardware dominance by keeping open-source AI dependent on its silicon while providing a cloud distribution channel.
NVIDIA published a practical observability framework mapping telemetry tools to every layer of DGX and HGX AI factory infrastructure. The guide helps teams catch gray failures like InfiniBand bit-error drift before they waste multi-hour training runs.
Meta's Ray-Ban AI glasses have moved from niche gadget to mainstream consumer product faster than almost any wearable in recent memory. The privacy implications cascade across three layers: individual bystanders, institutional policy.
Time Magazine has begun serving a distinct, stripped-down version of its webpages to AI crawlers such as ClaudeBot, PerplexityBot, and OpenAI's. The emergence of AI-only content layers signals a structural shift in the economics of web.