Latest 6G and AI news, analysis, and industry updates.
OpenAI’s GPT-Realtime-2 moves voice from a transcription layer to a native streaming reasoning interface for agents, raising the real competitive bar to latency, interruption handling, and reliability. The shift mirrors telecom QoS: model intelligence is still required, but session quality, error recovery, and trust boundaries will decide whether voice agents can run production workloads.
OpenAI, AMD, Broadcom, Intel, Microsoft, and NVIDIA launched the MRC open network protocol to cut GPU idle time in AI training clusters by automatically rerouting traffic across multiple paths, signaling that data-center networking is becoming a decisive layer of AI infrastructure competition.
OpenAI and Oracle have canceled the planned expansion of the Abilene data center from 1.2 GW to 2.0 GW, exposing a widening gap between AI capacity forecasts and the physical, financial, and power realities of gigawatt-scale buildouts.
A comprehensive analysis of network automation adoption levels across the world's largest telecom operators, from basic scripting to AI-driven autonomy.
The evolution of self-organizing network technology from basic SON to AI-driven fully autonomous networks, and what it means for operators.