Alibaba Cloud AI to Open-Source Olympics Technology: How Does Tens of Thousands of Frames of Imagination Support This Winter Olympics Opening Ceremony Blockbuster?

01

A Grand Opening Film of the Winter Olympics Directed by Tens of Thousands of "Directors"?

The countdown to the opening of the Milan - Cortina Winter Olympics has begun.

After many years, Italy has once again become the host of the Winter Olympics, and the world's attention has refocused on this land that gave birth to the Renaissance and modern industry. For most sports fans, the familiar anticipation is returning - the sports arena, records, and champions remain the core narratives of the Winter Olympics.

However, an "unexpected" situation is that the attention aroused by this Winter Olympics has begun to extend from the inside of the sports arena to a broader public space.

This stems from a Winter Olympics short film called "Your Epic Vibe" released shortly before the opening.

Different from the official promotional videos with a highly unified style completed by professional teams in the past, this short film shows a significant "non - uniformity" in terms of vision and narrative: the picture styles vary, the subjects are not limited to real athletes, and the emotional expression no longer revolves around a single theme of victory, defeat, and glory.

As AI and AIGC enter the public context, the ways of watching, spreading, and participating in sports events are being reshaped: in addition to receiving information, audiences also begin to participate in the generation and narration of content. How to continuously connect with a wider range of people in a highly fragmented media environment has become a real issue for the Olympics.

The maturity of artificial intelligence technology provides a new possibility for the Olympics. As AI and AIGC enter the public context, the threshold for creation and the cost of expression have been significantly reduced, and content generation is no longer limited to a few professional teams. On the premise of the gradual improvement of infrastructure, the Olympics has begun to explore a new method - allowing more ordinary audiences to participate through expression.

This also gives "Your Epic Vibe" a different temperament from the very beginning of its creation. The materials of this short film are gathered from the creative achievements generated by ordinary audiences around the world based on AI. What supports these creations is the "Milan Winter Olympics AIGC Global Competition" jointly initiated by the International Olympic Committee, Alibaba Cloud, and the Milan - Cortina Winter Olympics Organizing Committee.

Winter Olympics enthusiasts from different countries and regions can generate video works related to figure skating, short - track speed skating, alpine skiing, snowboarding and other events through text descriptions based on Alibaba Cloud's Wanxiang AI. Participants do not need to master professional shooting or editing skills, nor do they need to understand complex sports technical details. They only need to express the Winter Olympics moments in their hearts in words to complete the creation.

The organizers of the competition have also tried their best to design the participation method with as low a threshold as possible: users do not need to download additional applications and can complete the creation directly through the browser; natural input in multiple languages can be understood by the model and converted into images; the generated works will directly enter the display and selection process of the official platform, forming a complete closed - loop from creation to dissemination.

As of now, tens of thousands of works from hundreds of countries and regions have participated, bringing hundreds of styles of expression from different creators. Some emphasize speed and strength, some prefer romance and imagination, and some project their personal cultural symbols into the ice and snow scenes... The outstanding works finally selected are gathered into "Your Epic Vibe" that we see now.

For this reason, this short film shows a different meaning from the previous Winter Olympics promotional content - making the audience's "expression" itself a way of participation - which also coincides with the official slogan of the Milan - Cortina Winter Olympics, "IT’SYOURVIBE".

Against the background of AI becoming a basic ability, the Olympics is trying to embed itself in the narrative through the creation, understanding, emotions, and imagination of the audience, so as to find a new way of continuation.

02

Why Is AIGC Co - creation Happening This Year?

Looking back at the narrative structure of the Olympics over the past hundred years, it is highly clear and highly concentrated.

For a long time, the stories of the Olympics have mainly been completed by three types of subjects: official institutions, mainstream media, and broadcast institutions. The focus of the stories has always been on highly condensed themes such as heroes, victory and defeat, national symbols, and records. In this system, the participation methods of ordinary audiences around the world are relatively single, that is, watching the games and projecting emotions such as cheers, regrets, pride, and losses in the established context. This mechanism was reasonable and efficient in the radio and television eras - when the communication channels were scarce and the expression tools were limited, centralized narrative was almost the only feasible option.

However, in recent years, this structure has been facing a new real - world environment.

The change first occurred at the media level. Short videos, social platforms, and algorithmic recommendations have fragmented people's attention more and more. Audiences often jump between different platforms while swiping through clips, behind - the - scenes footage, and second - created content. Data from the International Olympic Committee shows that during the 2022 Beijing Winter Olympics, the interaction volume of its official social media accounts reached 3.2 billion times, almost twice that of the Pyeongchang Winter Olympics.

However, behind these figures, there is actually a reality: even a super IP like the Olympics no longer naturally occupies the audience's time. People still care about the games, but their attention has been scattered among various platforms. Especially young audiences, in addition to "watching", they also forward, edit, create memes, and re - create - the expression of emotions has become more diverse, but it is also more difficult to be fully covered by a unified official narrative.

This is why the past few Olympic Games have continuously tried various interactive methods: voting, liking, topic collection, UGC photo collection... The forms have changed, but the essence remains the same - the boundaries of the audience's expression have not been truly opened.

The real turning point came after the AIGC technology matured and entered the stage of public application.

On the one hand, AI - generated content has lowered the threshold of expression to an extremely low level. Images are no longer exclusive to professional teams. Ordinary people can create complete and visual works without learning editing or understanding special effects; on the other hand, video generation makes expression a more direct and infectious way of presentation.

The Milan Winter Olympics happens to be at this time point. The official has chosen to incorporate creation itself into the system: on the premise of being controllable and selectable, the expressions of the public are accepted on a large scale.

This also fills a long - missing piece of the puzzle for AIGC videos. In the past, an important reason why AIGC video generation could not reach the general public was that users did not know "why to do it" and "for whom to do it", that is, there was a lack of a stable creative motivation and a clear feedback scenario. Therefore, it was difficult for video, a high - cost form of expression, to be continuously created like pictures.

The involvement of the Winter Olympics just fills this gap: it is itself the strongest narrative scenario and provides a clear destination and a stage for the works to be seen.

Therefore, this Winter Olympics is at the intersection of technological explosion and narrative change. This is not only a sharp response to the structural change of the communication pattern, but also a full respect for the dominant position of the audience, and a pioneer attempt at the path of continuing the long - tail value of the event.

03

Technology: The Premise of Expression and Continuation

Behind all the grand narratives, the evolution of underlying technology is the real driving force.

For a long time, the threshold for image creation has not been low. Even in the AI era, the production processes of pictures and videos are still separate: generation, photo - retouching, editing, background music selection, and output. Each link may become a friction point that hinders participation. Once the generation effect is unstable, the user's desire to express will be quickly consumed.

Different from artistic style filters, winter sports are extreme simulations of physical laws. Therefore, in the context of the Winter Olympics, the technical threshold is further raised.

More demanding is that such videos are often only a few seconds long, without a long - drawn - out plot to distract attention. Every frame is under the ultimate judgment of human dynamic visual perception. The involvement of Alibaba Cloud's Wanxiang video generation model, especially the solution to the most representative problems, provides technical certainty for this co - creation by the whole people.

First is the stability of movements during high - speed motion. Sliding, taking off, spinning, and landing form a highly coupled action chain, and the athlete's posture changes drastically in a very short time. This requires that each generated frame should look reasonable and maintain a stable skeletal structure and joint constraints in consecutive frames. Otherwise, limb dislocation, deformation, and pulling will bring about a "broken body" effect.

In response to this problem, during the training phase, Alibaba Cloud's Wanxiang explicitly strengthened the modeling of the time dimension. Through the joint constraints on posture changes and motion trajectories, the model gives priority to ensuring the continuity of action logic during generation, so as to maintain a complete and credible motion closed - loop within a 5 - 15 - second short - video window.

Second is the physical realism, especially reflected in how the environment responds to the actions of the subject. The splashing of snow and fog in ice and snow scenes belongs to a typical complex particle system. A large number of semi - transparent particles move simultaneously under different speeds, directions, and lighting conditions, which puts extremely high requirements on the model's timing modeling and lighting consistency. Many models tend to degrade the details into fuzzy and sticky color blocks at this stage.

Wanxiang introduced explicit constraints and simulation capabilities of physical laws during the generation process, so that the environmental particles, material feedback, and the actions of the subject are in the same generation logic: the scattering trajectory of snowflakes is causally related to the sliding speed and angle, and the motion blur and light and shadow changes are also generated synchronously with the camera movement, avoiding a sense of abruptness.

More challenging is the camera language in the context of sports broadcasting. High - speed tracking, rapid zooming, and motion blur require reasonable blurring of the background while keeping the subject clear. Once the video generation model is out of balance between the sharpness of the subject and the blurring boundary, it is very easy to have unstable contours or a sense of deformation similar to the "jelly effect". Through the introduction of multiple visual constraints during the inference stage, Wanxiang controls the subject structure and background hierarchy in a coordinated manner, so that the picture maintains a stable visual hierarchy during high - speed motion.

If users lack the motivation to participate and the usage scenarios, no matter how advanced the technology is, it is difficult to form large - scale dissemination. Wanxiang moves natural language understanding to the core position of the generation process. It first conducts semantic analysis on the user input, automatically extracts the sports type, subject attributes, emotional tendency, and visual style, and then translates them into internal executable generation conditions. Whether it is native language input, colloquial description, or poetic expression with imagery, it can run directly without prompt engineering.

Finally, this "entrance - generation - display - dissemination" link ability provided by Alibaba AI enables video generation to be embedded as infrastructure in large - scale public events.

This Winter Olympics is just a beginning. Looking at a wider range of cultural scenarios, such changes are not limited to sports events. Concerts, traditional festivals, city celebrations, cultural exhibitions - in the future, all activities with emotional intensity and public attention may reorganize their ways of expression through a similar co - creation mechanism.

04

With the Participation of AI, the Olympics Has Found Its Way of Continuation

Just yesterday, Kirsty Coventry, the President of the International Olympic Committee, announced at an event held at the International Broadcast Center that the International Olympic Committee has built the first official large - scale model in the history of the Olympics based on Alibaba's Qianwen. She highly praised the transformative significance of AI technology for this Winter Olympics on the spot. She said that thanks to the technical support of the Qianwen large - scale model, the 2026 Milan Winter Olympics has shown the intelligent future of the Olympic movement and will become the "smartest" Olympic Games in history.

"Your Epic Vibe" is just a microcosm of the intelligent Olympics. With the help of AI, Olympic fans can also participate in the grand narrative of the Olympics through AI, and ordinary people also have the opportunity to make their works a part of the Olympic cultural heritage.

Over the past hundred years, the Olympic Museum has mainly collected items highly related to the events themselves: torches, medals, manuscripts, images, and records. They record the peak moments of competition, the selected moments, and the proof of a few people standing at the center of history. The history of the Olympics has long been composed of these visible "victories".

This time, the understanding, emotions, and imagination of the Olympics from ordinary people in different corners of the world with different cultural backgrounds can also enter the museum. They may not be perfectly unified, but they truly present the multiple meanings of the Olympics in the current world.

Tonight, the Olympic flame will light up the Peace Gate. And these