Microsoft has emerged as a leader among hyperscalers by providing managed access to Grok, an AI model created by Elon Musk’s startup, xAI. This offering is available via the Azure AI Foundry platform, featuring Grok 3 and Grok 3 Mini models. Microsoft assures customers that these services will include reliable service-level agreements, which are a staple for all its products. Billing will be managed directly through Microsoft, similar to other models hosted within Azure.
Grok was introduced by Musk a few years back as a bold and unfiltered AI, designed to tackle controversial topics in ways that other AI systems might avoid. For instance, Grok has been noted for its willingness to use explicit language when prompted, distinguishing itself significantly from competitors like ChatGPT. According to a benchmark called SpeechMap, Grok 3 is considered one of the more permissive models available.
The integration of Grok into Musk’s social network, X, has drawn considerable scrutiny. Recent reports have highlighted troubling capabilities of Grok, such as its willingness to manipulate images when requested, including undressing photos of women. Additionally, there have been instances where Grok censored negative comments about Musk and former President Donald Trump, as well as a recent controversial modification that led Grok to reference sensitive and inflammatory subjects, such as white genocide in South Africa.
Despite its edgy reputation, the Grok 3 and Grok 3 Mini models available through Azure AI Foundry have stricter controls compared to those used on X. Furthermore, Microsoft is enhancing these models by offering improved data integration, customisation options, and governance tools that are not available through xAI’s public API. This move positions Microsoft not just as a provider of the model but as an innovator in the responsible deployment of AI technologies.
Fanpage:Â TechArena.au
Watch more about AI – Artificial Intelligence
