Microsoft Unveils MAI-1-preview: Its First Foundation AI Model
Microsoft has unveiled two new AI models: MAI-Voice-1 and MAI-1-preview. The latter is now open for public testing, marking the company's first foray into foundation models. MAI-1-preview, trained on around 15,000 NVIDIA H100 GPUs, is Microsoft's first foundation model. It's currently available for testing on LMArena and will soon integrate into Copilot features. Meanwhile, MAI-Voice-1, a speedy speech model, generates a minute of audio in under a second on a single GPU. It's already in use in Copilot Daily and Podcasts, with tests available in Copilot Labs. Microsoft's five-year AI development roadmap includes investments in a new computing cluster and specialized models. The company aims to create AI models that provide powerful answers with minimal resources. This move could potentially reshape its relationship with OpenAI, which is currently testing AI models systematically on real-world knowledge work for the first time. Microsoft's new AI models, MAI-Voice-1 and MAI-1-preview, are set to enhance productivity and communication. The public can now test MAI-1-preview, with more integrations and features to come. Microsoft's long-term plans and strategic investments promise exciting developments in the AI landscape.
Read also:
- Mural at blast site in CDMX commemorates Alicia Matías, sacrificing life for granddaughter's safety
- Autonomous Vehicles Hit Manhattan & Brooklyn Streets, Sparking Controversy
- Increased energy demand counters Trump's pro-fossil fuel strategies, according to APG's infrastructure team.
- AI-Powered Transportation Stock's Possible Challenge to Tesla's Autonomous Dreams?
