今年巴黎奧運(yùn)會(huì)期間,當(dāng)你在刷社交媒體或看電視頻道時(shí),你可能比平時(shí)更關(guān)注一些不太常見的比賽,例如緊張的擊劍比賽,或者69歲的澳大利亞馬術(shù)選手瑪麗·漢娜為爭(zhēng)奪金牌所付出的努力。
他們可能要感謝計(jì)算巨頭英特爾(Intel)。在上一波科技浪潮之后,英特爾最終被微軟(Microsoft)甩在了身后。
人工智能助力奧運(yùn)會(huì)
作為生成式人工智能誕生以來(lái)的第一屆奧運(yùn)會(huì),巴黎成為半導(dǎo)體廠商英特爾展示其技術(shù)實(shí)力的終極試驗(yàn)場(chǎng)。
英特爾一直使用“數(shù)字孿生”技術(shù),該技術(shù)使其甚至早在奧運(yùn)會(huì)開始之前,就能復(fù)制巴黎的各個(gè)場(chǎng)館,以幫助巴黎進(jìn)行規(guī)劃。
英特爾還展示了一種人才識(shí)別網(wǎng)絡(luò),可根據(jù)敏捷性、耐力和力量將年輕運(yùn)動(dòng)員與理想的奧運(yùn)項(xiàng)目相匹配。英特爾在法蘭西體育場(chǎng)(Stade de France)內(nèi)設(shè)立了一個(gè)互動(dòng)展示廳,觀眾可以在那里發(fā)現(xiàn)他們的興趣點(diǎn)。
但冷門項(xiàng)目的運(yùn)動(dòng)員可能會(huì)為英特爾的一種創(chuàng)新技術(shù)歡呼,那就是使用人工智能創(chuàng)建精彩集錦。
英特爾使用人工智能自動(dòng)挑選奧運(yùn)項(xiàng)目中的精彩瞬間,并發(fā)送給廣播公司,以幫助他們快速整理出播出素材。這項(xiàng)技術(shù)主要通過(guò)觀眾的噪音來(lái)突出重要時(shí)刻,例如足球比賽中的進(jìn)球或游泳比賽中的沖刺階段等。
這項(xiàng)技術(shù)類似于IBM在溫布爾登網(wǎng)球賽中使用的技術(shù)。IBM的技術(shù)可通過(guò)觀眾的噪音幫助快速找到精彩瞬間,使廣播公司可以從大量視頻中挑選出最佳片段。
這項(xiàng)技術(shù)可以幫助那些在遠(yuǎn)離中心球場(chǎng)的17號(hào)球場(chǎng)上拼搏的低排名選手,在得分時(shí)得到更多認(rèn)可。
奧運(yùn)會(huì)人工智能也是出于類似的初衷,只是它的任務(wù)更加艱巨,需要在40個(gè)不同項(xiàng)目長(zhǎng)達(dá)數(shù)萬(wàn)小時(shí)的視頻中挖掘精彩瞬間。
然而,這項(xiàng)技術(shù)的最大受益者可能是乒乓球、速度攀巖和馬術(shù)等較冷門運(yùn)動(dòng)的選手。
英特爾歐洲、中東和非洲總裁德蒙特·哈加根對(duì)《財(cái)富》雜志表示:“傳統(tǒng)上,無(wú)論是內(nèi)容編輯還是精彩集錦的制作,甚至是跨平臺(tái)的內(nèi)容消費(fèi),都意味著它們只能覆蓋一定數(shù)量的運(yùn)動(dòng)項(xiàng)目?!?/p>
“但現(xiàn)在,在人工智能的幫助下,它們能夠擴(kuò)大內(nèi)容覆蓋的范圍。”
在四年一屆的奧運(yùn)會(huì)期間,冷門運(yùn)動(dòng)項(xiàng)目的選手爭(zhēng)奪獎(jiǎng)牌時(shí),能夠獲得更多關(guān)注,這并不新鮮。甚至在游泳、田徑和體操等更熱門的項(xiàng)目中,運(yùn)動(dòng)員在奧運(yùn)期間的媒體曝光率,也遠(yuǎn)高于平時(shí)。
然而,這種情況正變得越來(lái)越不確定。
近年來(lái),奧運(yùn)會(huì)為了保持吸引力引入了更多主流運(yùn)動(dòng),如高爾夫,這導(dǎo)致一些傳統(tǒng)冷門項(xiàng)目面臨被排擠的風(fēng)險(xiǎn)。
再加上其他非主流項(xiàng)目的出現(xiàn),比如在巴黎奧運(yùn)會(huì)上首次亮相的霹靂舞等,爭(zhēng)奪觀眾的競(jìng)爭(zhēng)變得更加激烈。
哈加根表示:“我們與國(guó)際奧委會(huì)的一個(gè)目標(biāo)是,讓一些相對(duì)冷門的運(yùn)動(dòng)項(xiàng)目也能獲得更多關(guān)注。與我們認(rèn)為的重點(diǎn)項(xiàng)目相比,這些項(xiàng)目可能沒有得到同樣多的曝光時(shí)間。”
“我們不知道這是否意味著這些運(yùn)動(dòng)會(huì)變得更受歡迎或有更多觀眾。我們需要隨著時(shí)間的推移進(jìn)行觀察,這也是我們的目的之一?!?/p>
哈加根重申,其目標(biāo)是增加項(xiàng)目的曝光度,而不是增加或減少任何特定項(xiàng)目的觀眾數(shù)量。
靈感乍現(xiàn)
隨著人工智能從案例研究進(jìn)入廣泛應(yīng)用,許多公司仍在思考如何將這項(xiàng)技術(shù)為己所用。
對(duì)于英特爾而言,奧運(yùn)會(huì)不只是其展現(xiàn)體育能力的機(jī)會(huì),也可以借此機(jī)會(huì)向客戶展示如何將人工智能應(yīng)用到不同領(lǐng)域。
“當(dāng)我們帶客戶參觀技術(shù)沙龍時(shí),你會(huì)看到人們當(dāng)場(chǎng)恍然大悟:‘我從未這樣想過(guò)?!?/p>
哈加根認(rèn)為,英特爾為奧運(yùn)會(huì)提供的解決方案,在奧運(yùn)會(huì)上有各種應(yīng)用場(chǎng)景,可以應(yīng)用于各行各業(yè)的企業(yè)面臨的日常問題。
“我認(rèn)為,企業(yè)可以從中學(xué)到的是,清楚你要如何應(yīng)用人工智能,知道它并非單一的技術(shù)。然后你要開始明確人工智能的用例,以及你期待它能創(chuàng)造哪些效益。”
與此同時(shí),如果人工智能可以幫助那些資金不足的業(yè)余運(yùn)動(dòng)員獲得關(guān)注,那么巴黎奧運(yùn)會(huì)不確定的遺產(chǎn),至少還有一個(gè)新的受益者。(財(cái)富中文網(wǎng))
譯者:劉進(jìn)龍
審校:汪皓
今年巴黎奧運(yùn)會(huì)期間,當(dāng)你在刷社交媒體或看電視頻道時(shí),你可能比平時(shí)更關(guān)注一些不太常見的比賽,例如緊張的擊劍比賽,或者69歲的澳大利亞馬術(shù)選手瑪麗·漢娜為爭(zhēng)奪金牌所付出的努力。
他們可能要感謝計(jì)算巨頭英特爾(Intel)。在上一波科技浪潮之后,英特爾最終被微軟(Microsoft)甩在了身后。
人工智能助力奧運(yùn)會(huì)
作為生成式人工智能誕生以來(lái)的第一屆奧運(yùn)會(huì),巴黎成為半導(dǎo)體廠商英特爾展示其技術(shù)實(shí)力的終極試驗(yàn)場(chǎng)。
英特爾一直使用“數(shù)字孿生”技術(shù),該技術(shù)使其甚至早在奧運(yùn)會(huì)開始之前,就能復(fù)制巴黎的各個(gè)場(chǎng)館,以幫助巴黎進(jìn)行規(guī)劃。
英特爾還展示了一種人才識(shí)別網(wǎng)絡(luò),可根據(jù)敏捷性、耐力和力量將年輕運(yùn)動(dòng)員與理想的奧運(yùn)項(xiàng)目相匹配。英特爾在法蘭西體育場(chǎng)(Stade de France)內(nèi)設(shè)立了一個(gè)互動(dòng)展示廳,觀眾可以在那里發(fā)現(xiàn)他們的興趣點(diǎn)。
但冷門項(xiàng)目的運(yùn)動(dòng)員可能會(huì)為英特爾的一種創(chuàng)新技術(shù)歡呼,那就是使用人工智能創(chuàng)建精彩集錦。
英特爾使用人工智能自動(dòng)挑選奧運(yùn)項(xiàng)目中的精彩瞬間,并發(fā)送給廣播公司,以幫助他們快速整理出播出素材。這項(xiàng)技術(shù)主要通過(guò)觀眾的噪音來(lái)突出重要時(shí)刻,例如足球比賽中的進(jìn)球或游泳比賽中的沖刺階段等。
這項(xiàng)技術(shù)類似于IBM在溫布爾登網(wǎng)球賽中使用的技術(shù)。IBM的技術(shù)可通過(guò)觀眾的噪音幫助快速找到精彩瞬間,使廣播公司可以從大量視頻中挑選出最佳片段。
這項(xiàng)技術(shù)可以幫助那些在遠(yuǎn)離中心球場(chǎng)的17號(hào)球場(chǎng)上拼搏的低排名選手,在得分時(shí)得到更多認(rèn)可。
奧運(yùn)會(huì)人工智能也是出于類似的初衷,只是它的任務(wù)更加艱巨,需要在40個(gè)不同項(xiàng)目長(zhǎng)達(dá)數(shù)萬(wàn)小時(shí)的視頻中挖掘精彩瞬間。
然而,這項(xiàng)技術(shù)的最大受益者可能是乒乓球、速度攀巖和馬術(shù)等較冷門運(yùn)動(dòng)的選手。
英特爾歐洲、中東和非洲總裁德蒙特·哈加根對(duì)《財(cái)富》雜志表示:“傳統(tǒng)上,無(wú)論是內(nèi)容編輯還是精彩集錦的制作,甚至是跨平臺(tái)的內(nèi)容消費(fèi),都意味著它們只能覆蓋一定數(shù)量的運(yùn)動(dòng)項(xiàng)目?!?/p>
“但現(xiàn)在,在人工智能的幫助下,它們能夠擴(kuò)大內(nèi)容覆蓋的范圍?!?/p>
在四年一屆的奧運(yùn)會(huì)期間,冷門運(yùn)動(dòng)項(xiàng)目的選手爭(zhēng)奪獎(jiǎng)牌時(shí),能夠獲得更多關(guān)注,這并不新鮮。甚至在游泳、田徑和體操等更熱門的項(xiàng)目中,運(yùn)動(dòng)員在奧運(yùn)期間的媒體曝光率,也遠(yuǎn)高于平時(shí)。
然而,這種情況正變得越來(lái)越不確定。
近年來(lái),奧運(yùn)會(huì)為了保持吸引力引入了更多主流運(yùn)動(dòng),如高爾夫,這導(dǎo)致一些傳統(tǒng)冷門項(xiàng)目面臨被排擠的風(fēng)險(xiǎn)。
再加上其他非主流項(xiàng)目的出現(xiàn),比如在巴黎奧運(yùn)會(huì)上首次亮相的霹靂舞等,爭(zhēng)奪觀眾的競(jìng)爭(zhēng)變得更加激烈。
哈加根表示:“我們與國(guó)際奧委會(huì)的一個(gè)目標(biāo)是,讓一些相對(duì)冷門的運(yùn)動(dòng)項(xiàng)目也能獲得更多關(guān)注。與我們認(rèn)為的重點(diǎn)項(xiàng)目相比,這些項(xiàng)目可能沒有得到同樣多的曝光時(shí)間?!?
“我們不知道這是否意味著這些運(yùn)動(dòng)會(huì)變得更受歡迎或有更多觀眾。我們需要隨著時(shí)間的推移進(jìn)行觀察,這也是我們的目的之一?!?/p>
哈加根重申,其目標(biāo)是增加項(xiàng)目的曝光度,而不是增加或減少任何特定項(xiàng)目的觀眾數(shù)量。
靈感乍現(xiàn)
隨著人工智能從案例研究進(jìn)入廣泛應(yīng)用,許多公司仍在思考如何將這項(xiàng)技術(shù)為己所用。
對(duì)于英特爾而言,奧運(yùn)會(huì)不只是其展現(xiàn)體育能力的機(jī)會(huì),也可以借此機(jī)會(huì)向客戶展示如何將人工智能應(yīng)用到不同領(lǐng)域。
“當(dāng)我們帶客戶參觀技術(shù)沙龍時(shí),你會(huì)看到人們當(dāng)場(chǎng)恍然大悟:‘我從未這樣想過(guò)?!?/p>
哈加根認(rèn)為,英特爾為奧運(yùn)會(huì)提供的解決方案,在奧運(yùn)會(huì)上有各種應(yīng)用場(chǎng)景,可以應(yīng)用于各行各業(yè)的企業(yè)面臨的日常問題。
“我認(rèn)為,企業(yè)可以從中學(xué)到的是,清楚你要如何應(yīng)用人工智能,知道它并非單一的技術(shù)。然后你要開始明確人工智能的用例,以及你期待它能創(chuàng)造哪些效益?!?/p>
與此同時(shí),如果人工智能可以幫助那些資金不足的業(yè)余運(yùn)動(dòng)員獲得關(guān)注,那么巴黎奧運(yùn)會(huì)不確定的遺產(chǎn),至少還有一個(gè)新的受益者。(財(cái)富中文網(wǎng))
譯者:劉進(jìn)龍
審校:汪皓
As you scroll through social media or channel surfing the Paris Olympics this year, you might find yourself more invested than normal in a tense round of fencing or 69-year-old Aussie equestrian Mary Hanna’s bid at gold.
They may have Intel, the computing giant eventually left behind by Microsoft in the wake of the last big tech wave, to thank.
The AI Olympics
As the first Olympics since the advent of generative AI, Paris serves as the ultimate testbed for semiconductor maker Intel to demonstrate its tech capabilities.
The group has been using “digital twins,” which allows it to create replicas of the various stadiums across Paris to help with planning before the Olympics even started.
It has also been showcasing a talent identification network, which pairs young athletes with their ideal Olympic sport based on agility, endurance, and power. The group has an interactive showroom at the Stade de France arena where spectators can identify their niche.
But one innovation professional athletes from less popular sports might be cheering on is Intel’s use of AI to create highlights.
Intel is using AI to automatically pick out highlight reels for Olympic sports, which go to broadcasters to help them quickly pull together a broadcast package. The technology mainly relies on the noise of crowds to highlight key moments, for example a shot on goal in football or the closing stages of swimming races.
The technology is similar to that implemented by IBM at Wimbledon, where crowd noise helps find big plays quickly so broadcasters can pick the best bits from a trove of footage.
That might help lower-ranked players, who are warring it out on Court 17, far from the cries of center court, for example, get more recognition for a winning point.
The aim of the Olympics AI is similar, but under the much more strenuous conditions of digging through tens of thousands of hours of footage across 40 different sports.
Its payoff, however, might be greatest for athletes in lower-profile sports like table tennis, speed climbing and equestrian.
“Traditionally, both the editing of that and the highlights, but also just the consumption of that across the different platforms, meant that they could only get to a certain amount of sports,” Dermot Hargaden, Intel’s EMEA president, told Fortune.
“But now, with the assistance that they’re getting, they’re able to widen that lens much more.”
It’s not new that obscure sports get a profile boost every four years as their athletes fight for a medal. Even at the more popular end, swimmers, track & field competitors, and gymnasts enjoy coverage levels of magnitude higher than outside the Olympic cycle.
However, this is increasingly less assured.
As the Olympics has introduced more mainstream sports, namely golf, to the calendar in recent years in a bid to maintain relevance, there is a risk of the more traditional, obscure sports being pushed out of the mix.
Add to that the arrival of other alternative sports, like Breaking’s debut at the Paris Olympics, and the fight for eyeballs becomes all the tougher.
“One of the objectives that we had with the IOC was that we’d have accessibility to perhaps some of the lesser in popularity sports, which maybe didn’t get as much time as what we might consider to be the key event,” added Hargaden.
“We don’t know whether it means that that sport gains more popularity or has increased viewership. We’ll have to see over time, but that was part of the intent.”
Hargaden reiterated that the goal was to increase accessibility rather than drive viewership to or from any particular sport.
Light bulbs going off
As the AI wave moves from case studies to wider rollouts, many companies are still determining how the tech will serve them.
For Intel, the games aren’t just a way to showcase its sporting acumen but also to show clients how they can adopt AI for various operations.
“As we bring customers through that tech salon, just in that room, you can see light bulbs going off with respect to, ‘I never thought of it that way.’”
Hargaden sees Intel’s solutions for the Olympics as analogous to everyday problems businesses face in all sectors, given the variety of use cases at the games.
“I think that’s what that’s transferable to businesses, is know what you’re going to do with it, know that it’s not one thing. And then you start to get really, really clear on the use case and what the benefit is that you’re trying to achieve.”
In the meantime, if AI can help turn the spotlight to part-time athletes in poorly-funded sports, Paris’s uncertain Olympic legacy will have at least one more benefactor.