Home News At the US Open, IBM serves up AI-generated tennis commentary and draw analysis

At the US Open, IBM serves up AI-generated tennis commentary and draw analysis

by WeeklyAINews
0 comment

Head over to our on-demand library to view classes from VB Rework 2023. Register Right here


Again in Could, IBM doubled down on its AI efforts with the announcement on the firm’s annual Think conference of its new Watsonx product platform, which supplies a foundational mannequin library that can be utilized to fine-tune pretrained fashions for enterprise software growth.

Now, the corporate is serving up what it hopes is a generative AI ace: For the primary time, it’s providing AI-generated audio tennis highlights for all matches through the two-week-long U.S. Open Tennis Championships, in addition to AI-powered evaluation to find out the projected issue of participant attracts and potential opponents.

Greater than 700,000 individuals head to Flushing Meadows, New York, every year to look at the very best tennis gamers on the earth compete, whereas greater than 10 million tennis followers around the globe comply with the match by means of the U.S. Open app and web site. And, for 3 a long time, IBM has been working with the United States Tennis Association on creating digital experiences for tennis followers.

IBM’s information operations bunker

The trouble begins within the basement-level IBM information operations heart at Arthur Ashe Stadium, the place tens of millions of knowledge factors are captured and analyzed. There are sometimes 56 information factors collected for each single level of a tennis match.

IBM is utilizing gen AI fashions constructed, skilled and deployed with Watsonx, and working throughout a hybrid cloud infrastructure from Crimson Hat OpenShift, to generate detailed audio narration and captions to accompany U.S. Open spotlight movies at unprecedented scale — for each match within the singles draw, throughout all 17 courts.

See also  Zuckerberg says Meta building generative AI into all its products, recommits to 'open science'

As well as, IBM debuted its Watsonx-powered AI Draw Evaluation that makes use of each structured and unstructured information to challenge the extent of benefit or drawback of all gamers within the singles draw. Every participant receives an IBM AI Draw Evaluation firstly of the match, which will probably be up to date day by day because the match progresses and gamers are eradicated. Each draw is ranked, permitting followers to click on into particular person matches and see the projected issue of their draw and potential opponents.

Beforehand, the USTA couldn’t cowl highlights of all matches

Kirsten Corio, chief business officer on the USTA, instructed VentureBeat that with 128 males and 120 ladies enjoying singles within the U.S. Open — in addition to doubles, juniors and wheelchair tennis matches — the group couldn’t cowl the highlights of many of the matches all through the match.

“Relying on what number of writers you’ve, you may solely do just a few matches at a time,” she mentioned. “The opposite matches would simply have stats and scores, however no commentary, so these tales are untold.”

So the USTA and IBM started to consider the way to scale match protection by combining stats and tales with gen AI.

“How might we use the information and expertise to really write highlights that may be dependable and correct sufficient?” mentioned Corio.

Corio added that the USTA desires of together with AI-generated highlights in several languages sooner or later. “We might love to do this in Spanish, to scale extra engagement,” she mentioned. “That’s the pure subsequent step.”

See also  Mistral AI bucks release trend by dropping torrent link to new open source LLM

Questions for IBM and USTA about AI hallucinations, information management

Whereas the USTA has been partnering with IBM on its expertise efforts for many years, in the case of at present’s superior AI functions, Corio identified that having the ability to management the information and the ecosystem is essential.

The USTA makes use of its personal curated, official information, “however there are lots on the market who peddle in unofficial information,” she defined. “We’re not but certain what the downstream results of that might be, so we’re truly placing collectively just a few completely different job forces throughout the corporate post-U.S. Open, to dig into how can it profit us? How can we defend in opposition to any potential battle?”

A extra of-the-moment concern is AI hallucinations — however in a presentation within the IBM Knowledge Heart beneath Arthur Ashe Stadium, an IBM spokesperson instructed VentureBeat that the corporate is doing human-in-the-loop high quality checks on its AI Commentary.

“We’re hoping over time we will cut back the necessity for human QA, however we do verify every spotlight clip, to make it possible for the commentary is stable,” the spokesman mentioned.

Source link

You Might Be Interested In
See also  Valued at $1B, Kai-Fu Lee's LLM startup unveils open source model

You may also like

logo

Welcome to our weekly AI News site, where we bring you the latest updates on artificial intelligence and its never-ending quest to take over the world! Yes, you heard it right – we’re not here to sugarcoat anything. Our tagline says it all: “because robots are taking over the world.”

Subscribe

Subscribe my Newsletter for new blog posts, tips & new photos. Let's stay updated!

© 2023 – All Right Reserved.