Latency measures the delay between an AI request and the resulting response. Interactive generative systems often track time to first token, which reflects how quickly output begins, and time to last token, which captures total completion time. Latency comes from model computation, queueing, network transfer, retrieval, tool calls, safety checks, and application logic. It varies with prompt length, output length, model size, hardware, batching, cache use, and current load. Optimizing only average latency can hide poor experiences, so teams also monitor percentile values and separate each stage of the request. The right target depends on whether the product is conversational, analytical, offline, or safety-critical.
Meta plans to begin manufacturing its in-house Iris AI chip in September as it aims to double computing capacity to 14 gigawatts next year and cut its reliance on Nvidia.
OpenAI released GPT-Live, full-duplex voice models that listen and speak at once and hand hard questions to GPT-5.5, replacing ChatGPT’s Advanced Voice Mode globally.
Meta will build its first Canadian data center, a 1-gigawatt AI facility in Alberta costing about C$13 billion, powered largely by a dedicated natural gas plant.
OpenAI will release its GPT-5.6 models, Sol, Terra and Luna, to the public on July 9 after clearing US government review far faster than the 30-day window the rules envisioned.
Apple expanded its partnership with Broadcom in a multi-year deal worth more than $30 billion, its largest US manufacturing commitment, to produce over 15 billion American-made chips.
Meta released Muse Image, a free AI image generator, but a feature that lets users turn any public Instagram user’s photos into AI images, opt-out by default, has sparked backlash.
Elon Musk’s xAI has rebranded to SpaceXAI, folding its Grok chatbot and X into SpaceX and signaling that AI, not rockets, is now central to the space company’s strategy.
Microsoft cut about 4,800 jobs, roughly 2.1% of its workforce, hitting Xbox and commercial sales hardest, the latest tech layoffs to intensify fears about AI and work.
Nvidia’s Kyber rack for its 2027 Rubin Ultra chips has slipped more than a year to 2028 over a hard-to-manufacture circuit board, SemiAnalysis says, with no proven fallback.
Meta quietly released Pocket, an app that lets people build small interactive games from text prompts and share them in a TikTok-style feed, built on its Gizmo acquisition.