Fresh juice


Microsoft unveils powerful AI that creates lifelike human avatars

In a groundbreaking development, researchers at Microsoft have unveiled an artificial intelligence tool that possesses the remarkable ability to generate deeply realistic human avatars. However, the tech giant has refrained from providing a timeline for its public release, citing concerns over the potential misuse of this technology to facilitate the creation of deceptive deepfake content.



Dubbed VASA-1, short for "visual affective skills," this AI model can produce an animated video of a person speaking, complete with synchronized lip movements, using just a single image and an audio clip containing speech. This feat of technological wizardry has the potential to revolutionize numerous industries, from entertainment and education to healthcare and beyond.

However, as with any powerful technology, there are valid concerns surrounding its ethical application. Disinformation researchers have expressed apprehensions about the rampant misuse of AI-powered applications to create "deepfake" pictures, videos, and audio clips, particularly in the crucial context of the upcoming election year.

Recognizing the gravity of these concerns, the authors of the VASA-1 report, released this week by Microsoft Research Asia, have taken a firm stance. "We are opposed to any behavior to create misleading or harmful contents of real persons," they wrote. "We are dedicated to developing AI responsibly, with the goal of advancing human well-being."

In a responsible move, Microsoft has stated that it has "no plans to release an online demo, API, product, additional implementation details, or any related offerings until we are certain that the technology will be used responsibly and in accordance with proper regulations."

The researchers at Microsoft have highlighted the remarkable capabilities of VASA-1, which can capture a wide spectrum of facial nuances and natural head motions. "It paves the way for real-time engagements with lifelike avatars that emulate human conversational behaviors," they stated in their post.

Moreover, VASA-1 is not limited to conventional applications; it can work with artistic photos, songs, and even non-English speech, according to Microsoft.

While acknowledging the potential benefits of this technology, such as providing virtual teachers to students or therapeutic support to people in need, the researchers have reiterated their commitment to responsible development. "It is not intended to create content that is used to mislead or deceive," they emphasized.

Notably, the VASA-1 videos still exhibit "artifacts" that reveal their AI-generated nature, providing a safeguard against potential misuse. However, as the technology continues to evolve, it is crucial to establish robust ethical frameworks and regulations to ensure its responsible deployment.

The unveiling of VASA-1 has sparked discussions among experts and technology enthusiasts alike. Ben Werdmuller, the technology lead at ProPublica, expressed excitement at the prospect of someone using the tool to represent themselves in a virtual meeting, wondering aloud, "Like, how did it go? Did anyone notice?"

This development by Microsoft comes on the heels of OpenAI's recent reveal of a voice-cloning tool called "Voice Engine," which can essentially duplicate someone's speech based on a 15-second audio sample. However, OpenAI has also adopted a cautious approach, citing the potential for synthetic voice misuse.

The risks associated with deepfake technology have already manifested in real-world scenarios. Earlier this year, a consultant working for a long-shot Democratic presidential candidate admitted to creating a robocall impersonation of Joe Biden, sent to voters in New Hampshire, in an attempt to highlight the dangers of AI-powered disinformation.

As we stand at the precipice of a new era of AI-driven avatar creation, it is imperative that we strike a delicate balance between harnessing the transformative potential of this technology and safeguarding against its misuse. Microsoft's responsible approach to the release of VASA-1 is a commendable step in this direction, setting a precedent for other technology companies to prioritize ethical considerations in their pursuit of innovation.

The road ahead is paved with both extraordinary opportunities and formidable challenges. It is incumbent upon us, as a society, to engage in open and transparent dialogues, establish robust regulatory frameworks, and foster a culture of ethical AI development. Only then can we truly harness the power of technologies like VASA-1 to shape a future where innovation and integrity go hand in hand.

Share with friends:

Write and read comments can only authorized users