"While GitHub presents the model as a new option for code generation, there is internal and external criticism about inadequate security controls, coercion of development teams, and inappropriate associations surrounding Grok. GitHub reports that it is adding a new language model to Copilot in collaboration with Elon Musk's AI company xAI. The model in question is Grok Code Fast 1, which will be available as a public preview for users of Copilot Pro, Pro+, Business, and Enterprise in Visual Studio Code."
"According to the official announcement, the rollout will be phased. For organizations with Business and Enterprise subscriptions, an administrator must first activate the model in the settings. For individual users, simply select Grok Code Fast 1 in the model selector. Additionally, the model can be accessed via Bring Your Own Key by adding your own xAI API key. GitHub is offering free access until September 2; after this date, regular terms and conditions will apply."
"The model is an extension of xAI's Grok family, which has previously come under fire for producing inappropriate output and controversial personas. Grok Code Fast 1 is tailored to programming tasks such as code generation and completion, and should therefore be better aligned with Copilot's objectives. GitHub emphasizes that all partner models are assessed according to the Responsible AI standards of parent company Microsoft and that Grok Code Fast 1 has undergone a full review, including automated evaluations and manual testing."
xAI's Grok Code Fast 1 is being added as a new language model option in GitHub Copilot, available in a phased public preview for Pro, Pro+, Business, and Enterprise users. Organizations must enable the model via admin settings while individuals can select it in the model selector; Bring Your Own Key with an xAI API key is supported. Free access is available until September 2, after which standard terms apply. The model targets programming tasks and underwent automated and manual evaluations under Microsoft's Responsible AI standards. Internal criticism cites superficial security checks, coercion of teams, and concerns about Grok's past inappropriate output and personas.
Read at Techzine Global
Unable to calculate read time
Collection
[
|
...
]