It is two years since generative AI came to mainstream prominence with the launch of ChatGPT.
Over that period, the technology has been suggested as a potential solution to almost every issue facing the sports industry. However, we are now seeing what fans and viewers want from AI.
Unsurprisingly, this differs from what the major tech companies have initially offered as public-facing use cases. Fans are responding to real-world events curated by machine learning to fit in with their own schedules and agendas.
Despite the repeated efforts from the likes of Meta, there is minimal appetite to engage with artificially created characters and little thrill in being on one side of what a celebrity interaction may resemble.
They want personalization, not personalities.
In 2024, we saw a mixture of high-profile successes and expensive failed experiments regarding fan-facing generative AI activations.
Arguably the most successful use case was Peacock’s coverage of the Summer Olympics in Paris.
The streaming service allowed its subscribers to state their preferences and preferred means of following the action before delivering a level of coverage tailored to each individual separately.
Every personalized recap was even voiced by an AI commentator trained on the voice of Al Michaels. The legendary Olympics broadcaster made the clips feel more premium and as though the broadcaster was directly talking to its audience one by one.
Note that there was no option to interact with Michaels, but his voice combined with the technology to provide a high-end concierge viewing experience at scale.
So successful was the AI-powered coverage that Peacock recorded its highest-ever share of TV viewership in August and received praise for its user-friendly delivery.
The rest of the industry has taken notice and rival broadcaster ESPN later announced that it would look to incorporate a similarly personalized approach to subscribers of its forthcoming streaming service.
Flagship news show SportsCenter was once must-see television for sports fans in the US.
In 2025 we will move towards a world where SportsCenter is no longer a national communal viewing experience. Every ESPN subscriber will see their version based on their implicitly or explicitly supplied preferences.
TV viewership will undergo the same seismic shifts that social media users have already lived through.
Social platforms learned long ago that users’ interests can differ from what they choose to follow.
This is the reason why almost every major app now defaults to an algorithmically sourced feed of content upon open. The platforms have simply learned what an individual is most likely to engage with more effectively than the users are able – or willing – to express themselves.
A fragmented broadcast landscape has effectively killed the notion of a listings guide which viewers peruse and then decide what they want to watch.
It will not be replaced with a digital equivalent.
Much like opening a social media app, the next era of viewership for sports fans will see broadcasters inherently understand their preferences and serve them directly – and individually.
Viewership is not guaranteed
In a world where viewership is dictated by algorithms; leagues, teams, and rights holders cannot rely on editorial oversight to ensure their individual sport is featured.
Your event may well be a ‘major’ or steeped in history and, as such, once commanded a spot at the top of the news bulletins or on the front page of the newspaper sports section. Advancing technology means fans no longer play by the same rules. They will tell you what they consider a ‘major’, or express it through their behavior patterns.
The popularity of individual sports will be more meritocratic based on how effectively they can capture attention and be relevant to modern audiences. The number of eyeballs on your individual sport will be dictated by how much you make people care rather than who you are.
Star power will come to the fore
The star name and celebrity culture which dominates social media could become relevant in the broadcast space.
Of course, high-profile athletes are already used to advertise live games on television. However, in the future, when broadcasters deliver personalized content packages, users will express a desire to watch certain players in addition to their favorite teams.
A casual soccer fan, for example, may engage with content featuring Lionel Messi with little idea of who he plays for. This scenario makes it more important for organizations to raise the profile of and associate themselves with star individuals to enable multiple routes into potential viewership.
It will also be incumbent upon teams to provide fans and casual observers with ongoing and multiple reasons to identify with them, rather than rely on an assumed inherent loyalty.
Choose between 'authentic' and ‘functional’
Recognizable faces may be ever more important, but Meta’s missteps can teach us lessons.
The company’s first round of AI celebrity chatbots attempted to replicate one-to-one fan interactions at scale artificially. This was unsuccessful and quickly shelved. Meta’s second attempt pivots to a place where celebrities act as AI tour guides while users fulfil their individual use cases. Athletes and legends have a part to play as sports embrace AI to create more engaging fan experiences, but they should be used to capture attention and facilitate a process.
Organizations can use the technology to make them say whatever they want in any language of their choosing, which creates unique global opportunities around user experience. Yet, if presenting an opportunity to engage with a player personally, there is no artificial workaround. Fans still demand and expect authenticity.
