Microsoft has announced a significant update to its intelligent assistant, Copilot, which now integrates the advanced image generation capabilities of OpenAI's GPT-4o model.
This move comes as part of a series of updates the company has been working on for Copilot, following months of buzz surrounding ChatGPT-4o's impressive image creation abilities, especially those that gained popularity for their 'Ghibli' artistic style.
According to Microsoft, "the image generation upgrade in Copilot, powered by the 4o model, allows for the creation of richer, more detailed images."
With this update, Copilot performs significantly better at refining images based on a previous one. And now, you can even upload your own photo to use as a starting point, offering a greater degree of creative control and flexibility.
Impressive Capabilities for Users
The integration of OpenAI's GPT-4o model brings a wide array of new image creation possibilities to Microsoft Copilot.
Among these capabilities, generating realistic images, modifying existing creations, and transforming the style of an existing image stand out.
Furthermore, the model offers the ability to display accurate and legible text within images, and to execute complex instructions efficiently.
It's worth noting that the company began rolling out the GPT-4o image generation tool to "Microsoft 365 Copilot" users (intended for businesses) last month, and now this update is reaching the consumer version of Microsoft Copilot.
With this, Microsoft Copilot now boasts image generation capabilities that surpass other Microsoft tools like "Microsoft Designer" and "Image Creator from Designer," which rely on OpenAI's older DALL-E technology.
To start experiencing these new capabilities for free, head to copilot.microsoft.com or use the assistant app on Android and iOS.
Race Against Time and Partnership Challenges
The question remains regarding Microsoft's ability to fully capitalize on the widespread interest in 'Ghibli'-style image generation techniques, especially given that the integration of GPT-4o into Copilot came after a period of anticipation.
Some observers believe this situation once again reflects how Microsoft finds itself unable to fully and swiftly leverage its multi-billion-dollar partnership with OpenAI.
In this context, Mustafa Suleyman, CEO of Microsoft AI, had expressed his dissatisfaction that OpenAI doesn't always provide immediate access to its advanced AI models, making it challenging for the tech giant to integrate these technologies across its vast range of products.
To counter any future challenges that may arise from this partnership, Microsoft is actively developing its own proprietary AI models within the company.
On the other hand, with these improvements, the tech giant is vigorously striving to position Copilot as a comprehensive AI assistant that strongly competes with major players in this field, such as OpenAI itself and Google with its Gemini model.
The most prominent goal is to enable users to create visual content directly within Microsoft 365 applications, such as Word, Excel, and Outlook, without needing to resort to external design tools.
