AI已經(jīng)到來,,而且其影響不可否認(rèn)。隨著其能力和智能的日益人性化,,AI有潛力通過自主性解鎖無限的創(chuàng)新和增長(zhǎng)機(jī)會(huì),。AI助理要在無需人類持續(xù)監(jiān)督的情況下,實(shí)現(xiàn)獨(dú)立運(yùn)行,、決策和采取行動(dòng),,取決于一個(gè)關(guān)鍵因素:信任。
我們的研究發(fā)現(xiàn),,大多數(shù)(77%)受訪高管同意,,要想真正獲得AI的好處,需要在整個(gè)企業(yè)中建立堅(jiān)實(shí)的信任基礎(chǔ),,而更多(81%)的受訪者同意,,信任戰(zhàn)略必須與任何技術(shù)戰(zhàn)略同步發(fā)展。關(guān)于信任的文章有很多,,其中與這個(gè)話題特別相關(guān)的兩種信任分別是認(rèn)知(或理性)信任和情感信任,。那么,,這些信任類型會(huì)如何影響我們與AI的關(guān)系以及對(duì)AI自主性的接受程度呢?不妨想想教青少年開車的例子,。
對(duì)于許多父母來說,,教孩子開車是一個(gè)既令人興奮又令人焦慮的成人禮。在開始時(shí),,父母主要依賴情感信任——這是他們內(nèi)心固有的愛和信念,,認(rèn)為他們的孩子會(huì)盡最大努力保持安全和負(fù)責(zé)任駕駛。然而,,這種信任不足以讓他們冷靜地坐在副駕駛座位上,。他們還需要有認(rèn)知信任作為補(bǔ)充。隨著孩子通過學(xué)習(xí)道路規(guī)則,、做出明智決策和處理意外挑戰(zhàn)展示出自身的能力,,父母才會(huì)逐漸產(chǎn)生認(rèn)知信任。
同樣,,情感信任和認(rèn)知信任之間的相互作用,,反映了人類評(píng)估復(fù)雜系統(tǒng)的方式,包括AI系統(tǒng),。當(dāng)用戶認(rèn)為AI與他們的利益一致,、其目的符合倫理、具有同理心并且意圖可靠時(shí),,他們就會(huì)產(chǎn)生對(duì)AI的情感信任,。當(dāng)AI系統(tǒng)通過可解釋性和可衡量的結(jié)果證明其能力、一致性和透明度時(shí),,就能獲得用戶的認(rèn)知信任,。這些信任類型共同構(gòu)成了實(shí)現(xiàn)AI自主性及其變革潛力的基礎(chǔ)。
建立信任和實(shí)現(xiàn)自主性的核心是認(rèn)知數(shù)字大腦
構(gòu)建一個(gè)包含信息,、記憶,、語境和推理能力的認(rèn)知數(shù)字大腦,是AI下一個(gè)發(fā)展階段的核心,,即通過賦能個(gè)人,、企業(yè)、行業(yè)和政府來建立信任和實(shí)現(xiàn)自主性,。首先要有一個(gè)基礎(chǔ)模型,,為AI帶來廣泛的一般知識(shí)和內(nèi)置學(xué)習(xí)能力。通過讓AI系統(tǒng)學(xué)習(xí)特定部分業(yè)務(wù),,并支持用戶形成AI系統(tǒng)的偏好,、喜好和需求,它可以變得高度個(gè)性化和智能化,。
企業(yè)將分散的AI工作統(tǒng)一到一個(gè)數(shù)字大腦中,,可以整合工作流程、機(jī)構(gòu)知識(shí),、價(jià)值鏈,、社會(huì)互動(dòng)等。隨著系統(tǒng)變得越來越智能,,用戶越來越多,,它還會(huì)進(jìn)一步增強(qiáng)用戶的信任和自身的自主性。
自主性和信任革命
為了幫助企業(yè)領(lǐng)導(dǎo)者為即將到來的變革做好準(zhǔn)備,,埃森哲(Accenture)《技術(shù)展望2025》(Technology Vision 2025)報(bào)告探討了生成式和自主式AI在技術(shù)開發(fā),、客戶體驗(yàn)、現(xiàn)實(shí)世界和勞動(dòng)力等方面產(chǎn)生的連鎖反應(yīng),,將會(huì)掀起的商業(yè)格局重塑和信任革命,,并提出了四個(gè)關(guān)鍵趨勢(shì):
1. 二進(jìn)制大爆炸——基于生成式AI的編程助手將軟件開發(fā)人員的角色提升為系統(tǒng)工程師,這加速了業(yè)務(wù)的數(shù)字化,。企業(yè)現(xiàn)在開發(fā)自己的認(rèn)知系統(tǒng)和將AI嵌入其獨(dú)特DNA中的選擇,,將決定其在未來幾年的發(fā)展?jié)摿ΑF髽I(yè)必須重新構(gòu)思應(yīng)用架構(gòu),,轉(zhuǎn)向基于意圖的框架和代理系統(tǒng),。由于編程的民主化和數(shù)字大腦的創(chuàng)建,我們看到企業(yè)定制系統(tǒng)越來越多,。隨著這些系統(tǒng)變得更加強(qiáng)大,、適應(yīng)能力更強(qiáng)和更加個(gè)性化,它們?cè)鰪?qiáng)的能力將激發(fā)更大的信任,。越來越多的代理和多代理系統(tǒng)將管理整個(gè)流程和功能,。例如,AI代理正在優(yōu)化沃爾瑪(Walmart)的庫(kù)存管理,,簡(jiǎn)化萬豪(Marriott)的預(yù)訂流程,,并增強(qiáng)雀巢(Nestlé)的供應(yīng)鏈運(yùn)營(yíng),這展示了AI系統(tǒng)如何憑借其專業(yè)知識(shí),、能力和一致性提供可衡量的結(jié)果,。
2. 你的未來面貌——自主性更強(qiáng)的AI將徹底改變客戶互動(dòng)方式,創(chuàng)造高度相關(guān)和個(gè)性化的體驗(yàn),。企業(yè)可以通過定義其AI個(gè)性,,即通過同理心、理解和真實(shí)性與客戶產(chǎn)生共鳴的品牌情感體現(xiàn),,從而建立更深的情感聯(lián)系和忠誠(chéng)度,。這有助于在AI系統(tǒng)與其服務(wù)對(duì)象之間建立持久的、基于信任的關(guān)系,。例如,,一家全球護(hù)膚品牌正在構(gòu)建一個(gè)超級(jí)個(gè)性化引擎,,其角色類似于個(gè)人首席美容科學(xué)家。它向客戶學(xué)習(xí),,將傳統(tǒng)AI與皮膚掃描技術(shù)相結(jié)合,,提供高度個(gè)性化的體驗(yàn),并具有高度的情感信任,。
3. 當(dāng)大型語言模型擁有實(shí)體時(shí)——我們正在迎來一個(gè)新的機(jī)器人時(shí)代,,機(jī)器人能夠理解現(xiàn)實(shí)世界中幾乎無限數(shù)量的場(chǎng)景,并且可以自主采取安全,、準(zhǔn)確的行動(dòng),。機(jī)器人和人類將形成一種共生關(guān)系,技術(shù)成為值得信任的伙伴,,將放大人類的潛力,。這種關(guān)系離不開情感信任,因?yàn)槿祟愋枰嘈胚@些系統(tǒng)理解他們的意圖,,與他們的目標(biāo)一致,,并且能夠以安全、符合倫理和協(xié)作的方式運(yùn)行,。例如,,埃森哲與英偉達(dá)(Nvidia)合作,幫助凱傲集團(tuán)(KION Group)創(chuàng)建未來的倉(cāng)庫(kù),,由AI驅(qū)動(dòng)的機(jī)器人與人類無縫合作,,以更快、更安全和以更低的成本完成訂單,。
4. 新的學(xué)習(xí)循環(huán)——在人類與AI之間形成良性循環(huán),,幫助我們真正學(xué)習(xí)、成長(zhǎng),,并重新構(gòu)思實(shí)現(xiàn)前所未有的績(jī)效水平的新方法,。隨著人類日益依賴AI提供一致、準(zhǔn)確和透明的結(jié)果,,就能形成認(rèn)知信任,。透明度至關(guān)重要,能夠讓用戶清晰了解AI的決策過程,,并確??勺匪葚?zé)任。例如,,埃森哲正在通過自主化代理轉(zhuǎn)變其營(yíng)銷職能,,以幫助其更快創(chuàng)建和執(zhí)行更明智的營(yíng)銷活動(dòng)。營(yíng)銷人員與展現(xiàn)出可靠性和透明度的AI助理合作,能夠提高信任度,,同時(shí)提升創(chuàng)造力和效率,。
自主化AI的新時(shí)代為企業(yè)和個(gè)人提供了一個(gè)非凡的機(jī)會(huì)。通過構(gòu)建基于認(rèn)知和情感信任的系統(tǒng),,企業(yè)可以重新定義客戶互動(dòng),,促進(jìn)共同創(chuàng)造,并支持人類和AI的共同繁榮發(fā)展,。以信任為基礎(chǔ),AI才能真正具有無限的可能性,。(財(cái)富中文網(wǎng))
本評(píng)論文章來自《財(cái)富》人工智能頭腦風(fēng)暴的贊助商埃森哲,。本文作者卡西克·納拉因現(xiàn)任埃森哲集團(tuán)首席技術(shù)執(zhí)行官兼首席技術(shù)官。關(guān)藍(lán)為首席人工智能官,。
Fortune.com上發(fā)表的評(píng)論文章中表達(dá)的觀點(diǎn),,僅代表作者本人的觀點(diǎn),不能代表《財(cái)富》雜志的觀點(diǎn)和立場(chǎng),。
翻譯:劉進(jìn)龍
審校:汪皓
AI已經(jīng)到來,,而且其影響不可否認(rèn)。隨著其能力和智能的日益人性化,,AI有潛力通過自主性解鎖無限的創(chuàng)新和增長(zhǎng)機(jī)會(huì),。AI助理要在無需人類持續(xù)監(jiān)督的情況下,實(shí)現(xiàn)獨(dú)立運(yùn)行,、決策和采取行動(dòng),,取決于一個(gè)關(guān)鍵因素:信任。
我們的研究發(fā)現(xiàn),,大多數(shù)(77%)受訪高管同意,,要想真正獲得AI的好處,需要在整個(gè)企業(yè)中建立堅(jiān)實(shí)的信任基礎(chǔ),,而更多(81%)的受訪者同意,,信任戰(zhàn)略必須與任何技術(shù)戰(zhàn)略同步發(fā)展。關(guān)于信任的文章有很多,,其中與這個(gè)話題特別相關(guān)的兩種信任分別是認(rèn)知(或理性)信任和情感信任,。那么,這些信任類型會(huì)如何影響我們與AI的關(guān)系以及對(duì)AI自主性的接受程度呢,?不妨想想教青少年開車的例子,。
對(duì)于許多父母來說,教孩子開車是一個(gè)既令人興奮又令人焦慮的成人禮,。在開始時(shí),,父母主要依賴情感信任——這是他們內(nèi)心固有的愛和信念,認(rèn)為他們的孩子會(huì)盡最大努力保持安全和負(fù)責(zé)任駕駛。然而,,這種信任不足以讓他們冷靜地坐在副駕駛座位上,。他們還需要有認(rèn)知信任作為補(bǔ)充。隨著孩子通過學(xué)習(xí)道路規(guī)則,、做出明智決策和處理意外挑戰(zhàn)展示出自身的能力,,父母才會(huì)逐漸產(chǎn)生認(rèn)知信任。
同樣,,情感信任和認(rèn)知信任之間的相互作用,,反映了人類評(píng)估復(fù)雜系統(tǒng)的方式,包括AI系統(tǒng),。當(dāng)用戶認(rèn)為AI與他們的利益一致,、其目的符合倫理、具有同理心并且意圖可靠時(shí),,他們就會(huì)產(chǎn)生對(duì)AI的情感信任,。當(dāng)AI系統(tǒng)通過可解釋性和可衡量的結(jié)果證明其能力、一致性和透明度時(shí),,就能獲得用戶的認(rèn)知信任,。這些信任類型共同構(gòu)成了實(shí)現(xiàn)AI自主性及其變革潛力的基礎(chǔ)。
建立信任和實(shí)現(xiàn)自主性的核心是認(rèn)知數(shù)字大腦
構(gòu)建一個(gè)包含信息,、記憶,、語境和推理能力的認(rèn)知數(shù)字大腦,是AI下一個(gè)發(fā)展階段的核心,,即通過賦能個(gè)人,、企業(yè)、行業(yè)和政府來建立信任和實(shí)現(xiàn)自主性,。首先要有一個(gè)基礎(chǔ)模型,,為AI帶來廣泛的一般知識(shí)和內(nèi)置學(xué)習(xí)能力。通過讓AI系統(tǒng)學(xué)習(xí)特定部分業(yè)務(wù),,并支持用戶形成AI系統(tǒng)的偏好,、喜好和需求,它可以變得高度個(gè)性化和智能化,。
企業(yè)將分散的AI工作統(tǒng)一到一個(gè)數(shù)字大腦中,,可以整合工作流程、機(jī)構(gòu)知識(shí),、價(jià)值鏈,、社會(huì)互動(dòng)等。隨著系統(tǒng)變得越來越智能,,用戶越來越多,,它還會(huì)進(jìn)一步增強(qiáng)用戶的信任和自身的自主性。
自主性和信任革命
為了幫助企業(yè)領(lǐng)導(dǎo)者為即將到來的變革做好準(zhǔn)備,埃森哲(Accenture)《技術(shù)展望2025》(Technology Vision 2025)報(bào)告探討了生成式和自主式AI在技術(shù)開發(fā),、客戶體驗(yàn),、現(xiàn)實(shí)世界和勞動(dòng)力等方面產(chǎn)生的連鎖反應(yīng),將會(huì)掀起的商業(yè)格局重塑和信任革命,,并提出了四個(gè)關(guān)鍵趨勢(shì):
1. 二進(jìn)制大爆炸——基于生成式AI的編程助手將軟件開發(fā)人員的角色提升為系統(tǒng)工程師,,這加速了業(yè)務(wù)的數(shù)字化。企業(yè)現(xiàn)在開發(fā)自己的認(rèn)知系統(tǒng)和將AI嵌入其獨(dú)特DNA中的選擇,,將決定其在未來幾年的發(fā)展?jié)摿?。企業(yè)必須重新構(gòu)思應(yīng)用架構(gòu),轉(zhuǎn)向基于意圖的框架和代理系統(tǒng),。由于編程的民主化和數(shù)字大腦的創(chuàng)建,,我們看到企業(yè)定制系統(tǒng)越來越多。隨著這些系統(tǒng)變得更加強(qiáng)大,、適應(yīng)能力更強(qiáng)和更加個(gè)性化,它們?cè)鰪?qiáng)的能力將激發(fā)更大的信任,。越來越多的代理和多代理系統(tǒng)將管理整個(gè)流程和功能,。例如,AI代理正在優(yōu)化沃爾瑪(Walmart)的庫(kù)存管理,,簡(jiǎn)化萬豪(Marriott)的預(yù)訂流程,,并增強(qiáng)雀巢(Nestlé)的供應(yīng)鏈運(yùn)營(yíng),這展示了AI系統(tǒng)如何憑借其專業(yè)知識(shí),、能力和一致性提供可衡量的結(jié)果,。
2. 你的未來面貌——自主性更強(qiáng)的AI將徹底改變客戶互動(dòng)方式,創(chuàng)造高度相關(guān)和個(gè)性化的體驗(yàn),。企業(yè)可以通過定義其AI個(gè)性,,即通過同理心、理解和真實(shí)性與客戶產(chǎn)生共鳴的品牌情感體現(xiàn),,從而建立更深的情感聯(lián)系和忠誠(chéng)度,。這有助于在AI系統(tǒng)與其服務(wù)對(duì)象之間建立持久的、基于信任的關(guān)系,。例如,,一家全球護(hù)膚品牌正在構(gòu)建一個(gè)超級(jí)個(gè)性化引擎,其角色類似于個(gè)人首席美容科學(xué)家,。它向客戶學(xué)習(xí),,將傳統(tǒng)AI與皮膚掃描技術(shù)相結(jié)合,提供高度個(gè)性化的體驗(yàn),,并具有高度的情感信任,。
3. 當(dāng)大型語言模型擁有實(shí)體時(shí)——我們正在迎來一個(gè)新的機(jī)器人時(shí)代,機(jī)器人能夠理解現(xiàn)實(shí)世界中幾乎無限數(shù)量的場(chǎng)景,并且可以自主采取安全,、準(zhǔn)確的行動(dòng),。機(jī)器人和人類將形成一種共生關(guān)系,技術(shù)成為值得信任的伙伴,,將放大人類的潛力,。這種關(guān)系離不開情感信任,因?yàn)槿祟愋枰嘈胚@些系統(tǒng)理解他們的意圖,,與他們的目標(biāo)一致,,并且能夠以安全、符合倫理和協(xié)作的方式運(yùn)行,。例如,,埃森哲與英偉達(dá)(Nvidia)合作,幫助凱傲集團(tuán)(KION Group)創(chuàng)建未來的倉(cāng)庫(kù),,由AI驅(qū)動(dòng)的機(jī)器人與人類無縫合作,,以更快、更安全和以更低的成本完成訂單,。
4. 新的學(xué)習(xí)循環(huán)——在人類與AI之間形成良性循環(huán),,幫助我們真正學(xué)習(xí)、成長(zhǎng),,并重新構(gòu)思實(shí)現(xiàn)前所未有的績(jī)效水平的新方法,。隨著人類日益依賴AI提供一致、準(zhǔn)確和透明的結(jié)果,,就能形成認(rèn)知信任,。透明度至關(guān)重要,能夠讓用戶清晰了解AI的決策過程,,并確??勺匪葚?zé)任。例如,,埃森哲正在通過自主化代理轉(zhuǎn)變其營(yíng)銷職能,,以幫助其更快創(chuàng)建和執(zhí)行更明智的營(yíng)銷活動(dòng)。營(yíng)銷人員與展現(xiàn)出可靠性和透明度的AI助理合作,,能夠提高信任度,,同時(shí)提升創(chuàng)造力和效率。
自主化AI的新時(shí)代為企業(yè)和個(gè)人提供了一個(gè)非凡的機(jī)會(huì),。通過構(gòu)建基于認(rèn)知和情感信任的系統(tǒng),,企業(yè)可以重新定義客戶互動(dòng),促進(jìn)共同創(chuàng)造,,并支持人類和AI的共同繁榮發(fā)展,。以信任為基礎(chǔ),,AI才能真正具有無限的可能性。(財(cái)富中文網(wǎng))
本評(píng)論文章來自《財(cái)富》人工智能頭腦風(fēng)暴的贊助商埃森哲,。本文作者卡西克·納拉因現(xiàn)任埃森哲集團(tuán)首席技術(shù)執(zhí)行官兼首席技術(shù)官,。關(guān)藍(lán)為首席人工智能官。
Fortune.com上發(fā)表的評(píng)論文章中表達(dá)的觀點(diǎn),,僅代表作者本人的觀點(diǎn),,不能代表《財(cái)富》雜志的觀點(diǎn)和立場(chǎng)。
翻譯:劉進(jìn)龍
審校:汪皓
AI is here, and its impact is undeniable. With its increasingly human-like capabilities and intelligence, AI has the potential to unlock limitless opportunities for innovation and growth, largely through autonomy. Realizing the promise of AI agents—to operate independently, make decisions, and take actions without constant human oversight—depends on one critical factor: trust.
Our research finds that most (77%) of the executives we surveyed agree the true benefits of AI require a solid foundation of trust across the enterprise, while even more (81%) agree that trust strategy must evolve alongside any technology strategy. While much has been written about trust, two types are especially relevant in this conversation: cognitive (or rational) trust and emotional trust. But how do these types of trust influence our relationship with AI and comfort with autonomy? Consider teaching a teenager to drive.
For many parents, it’s a rite of passage that evokes both excitement and anxiety. At the beginning, parents rely heavily on emotional trust—the inherent bond of love and belief that their child will try their best to be safe and responsible. Yet, this trust is not enough to let them sit calmly in the passenger seat. It must be complemented by cognitive trust, which develops over time as the child demonstrates competence by learning the rules of the road, making sound decisions, and handling unexpected challenges.
Similarly, this interplay between emotional and cognitive trust reflects how humans evaluate complex systems, including AI. Emotional trust in AI arises when users perceive it as aligned with their interests, ethical in its purpose, having empathy, and reliable in its intentions. Cognitive trust is earned when AI systems prove their competence, consistency, and transparency through explainability and measurable outcomes. Together, these forms of trust are the foundation upon which AI’s autonomy—and its transformative potential—can be realized.
A cognitive digital brain at the center of trust and autonomy
Building a cognitive digital brain—which includes information, memory, context, and reasoning—is central to AI’s next stage, enabling trust and autonomy by empowering individuals, businesses, industries, and governments. It starts with foundational models that bring broad general knowledge and a built-in ability to learn. By teaching AI systems about specific parts of the business and allowing users to shape its preferences, likes, and needs, it can become deeply personalized and intelligent.
Enterprises that unify scattered AI efforts into a digital brain can integrate workflows, institutional knowledge, value chains, social interactions, and more. As the system grows smarter and more people use it, it also reinforces trust and autonomy.
Autonomy and the trust revolution
To help business leaders prepare for the transformative journey ahead, the Accenture Technology Vision 2025 report explores the business reinvention—and trust revolution—that will happen as generative and autonomous AI ripple across dimensions of technology development, customer experience, the physical world, and the workforce with four key trends:
1, The Binary Big Bang – Gen AI-based coding assistants elevate the role of software developers to systems engineers, which accelerates the digitization of business. The choices organizations make now in developing their own cognitive systems and embedding AI into their enterprise’s unique DNA will shape their potential for years to come. Application architectures must be reimagined, shifting toward an intention-based framework and agentic systems. We see a rise in custom systems across enterprises because of this democratization of coding and the creation of digital brains. As these systems become more capable, adaptive and personal, they will inspire greater trust through enhanced competence. More and more, agentic and multi-agent systems will manage entire processes and functions. For example, AI agents are optimizing inventory management for Walmart, streamlining booking processes for Marriott, and enhancing Nestlé’s supply chain operations—demonstrating how AI systems can deliver measurable results with expertise, competence, and consistency.
2, Your Face, in the Future — More autonomous AI will radically transform customer engagement, creating highly relevant and personalized experiences. Businesses can create deeper emotional connections and loyalty by defining their AI personality, an emotional embodiment of the brand that resonates with customers through empathy, understanding, and authenticity. This can help build enduring, trust-based relationships between AI systems and the people they serve. For example, a global skincare brand is building a hyper-personalization engine that acts like a personal chief beauty scientist. It learns from customers, combining traditional AI with skin scan technology to deliver highly personalized experiences with a high degree of emotional trust.
3, When LLMs Get Their Bodies — We are entering a new era for robotics with the ability to understand a nearly infinite number of scenarios in physical world and take safe, accurate actions autonomously. Robots and humans will form a symbiotic relationship, where technology becomes a trusted partner in amplifying human potential. This requires emotional trust, as humans need to feel confident that these systems understand their intentions, align with their goals, and operate safely, ethically, and collaboratively. For example, in partnership with Nvidia, Accenture is helping KION Group create the warehouse of the future, where AI-driven robots work seamlessly with people to fulfill orders faster, safer and at a lower cost.
4, The New Learning Loop — Creating a virtuous cycle between people and AI helps us to truly learn, grow, and reimagine new ways to achieve unprecedented levels of performance. Cognitive trust develops as people rely on AI’s proven ability to deliver consistent, accurate, and transparent results. Transparency is essential, providing users with clear insights into AI’s decision-making processes and ensuring accountability. For example, Accenture is transforming its marketing function with autonomous agents that help create and execute smarter campaigns faster. By pairing marketers with AI agents that demonstrate reliability and transparency, the collaboration enhances trust while unlocking new levels of creativity and efficiency.
This new age of autonomous AI presents an extraordinary opportunity for businesses and individuals alike. By building systems grounded in cognitive and emotional trust, organizations can redefine customer engagement, foster co-creation, and empower both people and AI to thrive together. With trust as the foundation, the possibilities of AI become truly limitless.
This commentary is from Accenture, a sponsor of Fortune Brainstorm AI. Karthik Narain is group chief executive, technology, and chief technology officer of Accenture. Lan Guan is chief AI officer.
The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not necessarily reflect the opinions and beliefs of Fortune.