一個(gè)周日下午,,我點(diǎn)開(kāi)了我們公司網(wǎng)站上聊天機(jī)器人的聊天記錄,它們平時(shí)負(fù)責(zé)“接待”訪客,,回答問(wèn)題,。
其中一份聊天記錄引起了我的注意,讓我笑出了聲,。
Juji(聊天機(jī)器人):在你的工作或個(gè)人生活中,,最大的痛點(diǎn)或挑戰(zhàn)是什么?
用戶(hù)A:你是男生嗎,?
Juji:我不是真人,。
用戶(hù)A:我的工作是在確保不被吃掉的情況下馴服老虎。
Juji:明白了,。如果哪天我變得更強(qiáng)大了,,我很愿意幫助你。
笑完之后,,我突然意識(shí)到,,優(yōu)秀的人工智能聊天機(jī)器人或許可以給我們?nèi)祟?lèi),尤其是領(lǐng)導(dǎo)者帶來(lái)一些重要的經(jīng)驗(yàn)教訓(xùn),。順著這個(gè)想法,,我進(jìn)一步觀察到優(yōu)秀的人工智能聊天機(jī)器人,尤其是那些擁有認(rèn)知智慧(人類(lèi)軟技能)的機(jī)器人,,能夠教領(lǐng)導(dǎo)者學(xué)會(huì)共情及其他方面的技能,。
謙卑
由于天生的不完美,優(yōu)秀的人工智能聊天機(jī)器人在設(shè)計(jì)時(shí)通常會(huì)被賦予更謙卑的語(yǔ)氣:它們會(huì)為自己的錯(cuò)誤道歉,,承認(rèn)缺乏某些知識(shí)等,。在我們?nèi)祟?lèi)的世界里,,每個(gè)人都是獨(dú)一無(wú)二的,但沒(méi)有人是完美的,。
如果領(lǐng)導(dǎo)者可以放下一點(diǎn)自負(fù),,帶上一點(diǎn)謙虛,承認(rèn)自己的缺點(diǎn)或錯(cuò)誤,,就將有助于領(lǐng)導(dǎo)者為團(tuán)隊(duì)成員樹(shù)立榜樣,,打造更團(tuán)結(jié)、更具同理心,、更和諧的職場(chǎng)文化,。
透明度
我們一直提倡不要把人工智能偽裝成真實(shí)的人類(lèi)。這樣做只會(huì)損害你的品牌,,讓客戶(hù)感覺(jué)受到了欺騙,,因?yàn)槿祟?lèi)很快就能夠發(fā)現(xiàn)人工智能的缺陷,意識(shí)到它根本不是真人,。此外,,科學(xué)研究表明,人工智能的自我暴露有助于提升對(duì)話的真實(shí)性,。因此,,優(yōu)秀的人工智能聊天機(jī)器人的訓(xùn)練目標(biāo)是對(duì)用戶(hù)保持透明,總是坦陳自己的能力和局限性,。
與之類(lèi)似,,如果領(lǐng)導(dǎo)者總是誠(chéng)實(shí)地面對(duì)自己的能力和局限,與團(tuán)隊(duì)誠(chéng)實(shí)地分享意見(jiàn)和感受,,就將有助于與團(tuán)隊(duì)建立可靠真實(shí)的關(guān)系,。這樣員工們會(huì)覺(jué)得做真實(shí)的自己是安全的、有保障的,、舒適的,,可以隨之培養(yǎng)出更緊密、更包容的工作文化,。
積極傾聽(tīng)
人們更愿意向機(jī)器敞開(kāi)心扉(當(dāng)需要透露敏感信息或自身想法時(shí)),,因?yàn)樗麄冇X(jué)得機(jī)器不會(huì)評(píng)判他們。
認(rèn)知型人工智能聊天機(jī)器人更進(jìn)一步,,積極傾聽(tīng)用戶(hù)的心聲,,比如說(shuō),重復(fù)用戶(hù)說(shuō)過(guò)的話,,充滿(mǎn)同理心地認(rèn)可用戶(hù)的感受,,讓用戶(hù)感覺(jué)得到了聆聽(tīng)。更重要的是,它們從不對(duì)用戶(hù)表達(dá)的內(nèi)容進(jìn)行評(píng)判,。同樣的,,如果組織的領(lǐng)導(dǎo)者能夠?qū)W會(huì)積極傾聽(tīng),摒棄自我意識(shí)和評(píng)判他人的沖動(dòng),,把注意力放在員工而不是自己的想法上,,就可以更好地與團(tuán)隊(duì)拉近關(guān)系,更好地理解并領(lǐng)導(dǎo)團(tuán)隊(duì),。
讀出言外之意
與心理學(xué)家一樣,,認(rèn)知型人工智能聊天機(jī)器人的另一項(xiàng)重要技能是通過(guò)動(dòng)態(tài)分析用戶(hù)的對(duì)話文本,自動(dòng)推斷用戶(hù)未明確表達(dá)的需求,、興趣點(diǎn)和個(gè)性特征,,從而讀出言外之意。
2000多年前,,古羅馬的政治家西塞羅曾經(jīng)說(shuō)過(guò):“如果你想說(shuō)服我,,你就必須思考我的思想,感受我的感情,,說(shuō)出我的語(yǔ)言,。”得力的領(lǐng)導(dǎo)者應(yīng)該學(xué)會(huì)如何從與員工的對(duì)話中讀出言外之意,,與他們的感受和想法共情,,利用這些見(jiàn)解為他們?cè)诼殘?chǎng)上提供最大幫助,。
現(xiàn)在的聊天機(jī)器人擁有先進(jìn)的人類(lèi)軟技能,,例如積極傾聽(tīng)和理解言外之意,按照全球各地領(lǐng)導(dǎo)人的說(shuō)法,,這也是最重要的領(lǐng)導(dǎo)技能,。
隨著人類(lèi)不斷發(fā)展人工智能,認(rèn)知型人工智能也可以作為人類(lèi)自身行為的一面鏡子,,幫助人類(lèi)建立更高效能,、更包容的關(guān)系。如果領(lǐng)導(dǎo)者擁有這類(lèi)軟技能,,在領(lǐng)導(dǎo)團(tuán)隊(duì)時(shí)就更具同理心,、能夠積極傾聽(tīng)、提升透明度,,將打造一個(gè)更快樂(lè),、更高效、更包容的職場(chǎng),。(財(cái)富中文網(wǎng))
米歇爾·周,,博士,是Juji公司(Juji, Inc.)的聯(lián)合創(chuàng)始人及首席執(zhí)行官,,該公司主要生產(chǎn)具有認(rèn)知能力的易用型人工智能助手,,她還是美國(guó)計(jì)算機(jī)學(xué)會(huì)交互式智能系統(tǒng)學(xué)報(bào)(ACM Transactions on Interactive Intelligent Systems)的主編,。在創(chuàng)立Juji之前,米歇爾在IBM 公司的Research/Watson負(fù)責(zé)前沿技術(shù)的研發(fā),,包括IBM RealHunter和Watson Personality Insights,。米歇爾畢業(yè)于哥倫比亞大學(xué)(Columbia University),獲計(jì)算機(jī)科學(xué)博士學(xué)位,,還是美國(guó)計(jì)算機(jī)學(xué)會(huì)杰出科學(xué)家(ACM Distinguished Scientist),。
Fortune.com上的評(píng)論文章僅代表作者個(gè)人觀點(diǎn),不代表《財(cái)富》雜志的觀點(diǎn)和立場(chǎng),。
譯者:Agatha
一個(gè)周日下午,,我點(diǎn)開(kāi)了我們公司網(wǎng)站上聊天機(jī)器人的聊天記錄,它們平時(shí)負(fù)責(zé)“接待”訪客,,回答問(wèn)題,。
其中一份聊天記錄引起了我的注意,讓我笑出了聲,。
Juji(聊天機(jī)器人):在你的工作或個(gè)人生活中,,最大的痛點(diǎn)或挑戰(zhàn)是什么?
用戶(hù)A:你是男生嗎,?
Juji:我不是真人,。
用戶(hù)A:我的工作是在確保不被吃掉的情況下馴服老虎。
Juji:明白了,。如果哪天我變得更強(qiáng)大了,,我很愿意幫助你。
笑完之后,,我突然意識(shí)到,,優(yōu)秀的人工智能聊天機(jī)器人或許可以給我們?nèi)祟?lèi),尤其是領(lǐng)導(dǎo)者帶來(lái)一些重要的經(jīng)驗(yàn)教訓(xùn),。順著這個(gè)想法,,我進(jìn)一步觀察到優(yōu)秀的人工智能聊天機(jī)器人,尤其是那些擁有認(rèn)知智慧(人類(lèi)軟技能)的機(jī)器人,,能夠教領(lǐng)導(dǎo)者學(xué)會(huì)共情及其他方面的技能,。
謙卑
由于天生的不完美,優(yōu)秀的人工智能聊天機(jī)器人在設(shè)計(jì)時(shí)通常會(huì)被賦予更謙卑的語(yǔ)氣:它們會(huì)為自己的錯(cuò)誤道歉,,承認(rèn)缺乏某些知識(shí)等,。在我們?nèi)祟?lèi)的世界里,每個(gè)人都是獨(dú)一無(wú)二的,,但沒(méi)有人是完美的,。
如果領(lǐng)導(dǎo)者可以放下一點(diǎn)自負(fù),帶上一點(diǎn)謙虛,承認(rèn)自己的缺點(diǎn)或錯(cuò)誤,,就將有助于領(lǐng)導(dǎo)者為團(tuán)隊(duì)成員樹(shù)立榜樣,,打造更團(tuán)結(jié)、更具同理心,、更和諧的職場(chǎng)文化,。
透明度
我們一直提倡不要把人工智能偽裝成真實(shí)的人類(lèi)。這樣做只會(huì)損害你的品牌,,讓客戶(hù)感覺(jué)受到了欺騙,,因?yàn)槿祟?lèi)很快就能夠發(fā)現(xiàn)人工智能的缺陷,意識(shí)到它根本不是真人,。此外,,科學(xué)研究表明,人工智能的自我暴露有助于提升對(duì)話的真實(shí)性,。因此,,優(yōu)秀的人工智能聊天機(jī)器人的訓(xùn)練目標(biāo)是對(duì)用戶(hù)保持透明,總是坦陳自己的能力和局限性,。
與之類(lèi)似,,如果領(lǐng)導(dǎo)者總是誠(chéng)實(shí)地面對(duì)自己的能力和局限,與團(tuán)隊(duì)誠(chéng)實(shí)地分享意見(jiàn)和感受,,就將有助于與團(tuán)隊(duì)建立可靠真實(shí)的關(guān)系,。這樣員工們會(huì)覺(jué)得做真實(shí)的自己是安全的、有保障的,、舒適的,,可以隨之培養(yǎng)出更緊密、更包容的工作文化,。
積極傾聽(tīng)
人們更愿意向機(jī)器敞開(kāi)心扉(當(dāng)需要透露敏感信息或自身想法時(shí)),,因?yàn)樗麄冇X(jué)得機(jī)器不會(huì)評(píng)判他們,。
認(rèn)知型人工智能聊天機(jī)器人更進(jìn)一步,,積極傾聽(tīng)用戶(hù)的心聲,比如說(shuō),,重復(fù)用戶(hù)說(shuō)過(guò)的話,,充滿(mǎn)同理心地認(rèn)可用戶(hù)的感受,讓用戶(hù)感覺(jué)得到了聆聽(tīng),。更重要的是,,它們從不對(duì)用戶(hù)表達(dá)的內(nèi)容進(jìn)行評(píng)判。同樣的,,如果組織的領(lǐng)導(dǎo)者能夠?qū)W會(huì)積極傾聽(tīng),,摒棄自我意識(shí)和評(píng)判他人的沖動(dòng),把注意力放在員工而不是自己的想法上,就可以更好地與團(tuán)隊(duì)拉近關(guān)系,,更好地理解并領(lǐng)導(dǎo)團(tuán)隊(duì),。
讀出言外之意
與心理學(xué)家一樣,認(rèn)知型人工智能聊天機(jī)器人的另一項(xiàng)重要技能是通過(guò)動(dòng)態(tài)分析用戶(hù)的對(duì)話文本,,自動(dòng)推斷用戶(hù)未明確表達(dá)的需求,、興趣點(diǎn)和個(gè)性特征,從而讀出言外之意,。
2000多年前,,古羅馬的政治家西塞羅曾經(jīng)說(shuō)過(guò):“如果你想說(shuō)服我,你就必須思考我的思想,,感受我的感情,,說(shuō)出我的語(yǔ)言?!钡昧Φ念I(lǐng)導(dǎo)者應(yīng)該學(xué)會(huì)如何從與員工的對(duì)話中讀出言外之意,,與他們的感受和想法共情,利用這些見(jiàn)解為他們?cè)诼殘?chǎng)上提供最大幫助,。
現(xiàn)在的聊天機(jī)器人擁有先進(jìn)的人類(lèi)軟技能,,例如積極傾聽(tīng)和理解言外之意,按照全球各地領(lǐng)導(dǎo)人的說(shuō)法,,這也是最重要的領(lǐng)導(dǎo)技能,。
隨著人類(lèi)不斷發(fā)展人工智能,認(rèn)知型人工智能也可以作為人類(lèi)自身行為的一面鏡子,,幫助人類(lèi)建立更高效能,、更包容的關(guān)系。如果領(lǐng)導(dǎo)者擁有這類(lèi)軟技能,,在領(lǐng)導(dǎo)團(tuán)隊(duì)時(shí)就更具同理心,、能夠積極傾聽(tīng)、提升透明度,,將打造一個(gè)更快樂(lè),、更高效、更包容的職場(chǎng),。(財(cái)富中文網(wǎng))
米歇爾·周,,博士,是Juji公司(Juji, Inc.)的聯(lián)合創(chuàng)始人及首席執(zhí)行官,,該公司主要生產(chǎn)具有認(rèn)知能力的易用型人工智能助手,,她還是美國(guó)計(jì)算機(jī)學(xué)會(huì)交互式智能系統(tǒng)學(xué)報(bào)(ACM Transactions on Interactive Intelligent Systems)的主編。在創(chuàng)立Juji之前,,米歇爾在IBM 公司的Research/Watson負(fù)責(zé)前沿技術(shù)的研發(fā),,包括IBM RealHunter和Watson Personality Insights,。米歇爾畢業(yè)于哥倫比亞大學(xué)(Columbia University),獲計(jì)算機(jī)科學(xué)博士學(xué)位,,還是美國(guó)計(jì)算機(jī)學(xué)會(huì)杰出科學(xué)家(ACM Distinguished Scientist),。
Fortune.com上的評(píng)論文章僅代表作者個(gè)人觀點(diǎn),不代表《財(cái)富》雜志的觀點(diǎn)和立場(chǎng),。
譯者:Agatha
On a Sunday afternoon, I was going through chat transcripts produced by our chatbot, which sits on our company’s website, greeting visitors and answering their questions.
One transcript caught my eye and made me laugh.
Juji (the chatbot): What is the top pain point or challenge in your line of work or personal life?
User A: Are you a boy?
Juji: I’m not a person
User A: My job is to tame tigers without being eaten.
Juji: Got it. Once I become more powerful, I’d love to help you out.
After the laughs wore off, I suddenly realized that good A.I. chatbots could perhaps teach us humans, especially leaders, important lessons. Exploring this thought further, I observed five things that good A.I. chatbots, especially those with cognitive intelligence (human soft skills) can teach leaders about empathy and more.
Humility
Due to their imperfections, good A.I. chatbots are often designed with a humble tone: They apologize for their mistakes or admit a lack of certain knowledge. In our human world, everyone is unique, but no one is perfect.
If leaders can shed a little ego and carry a sense of humility, acknowledging their shortcomings or mistakes, it would help set examples for team members, promoting a more collaborative, empathetic, and harmonious workplace culture.
Transparency
We always advocate not to disguise A.I. as a real person. Doing so will only hurt your brand and make your audience feel cheated because humans will quickly discover the flaws in A.I. and realize it is not human after all. Moreover, scientific research shows A.I. self-disclosures promote authentic conversations. Thus, good AI chatbots are trained to be transparent with users and always disclose their capabilities and limitations.
Analogously, if leaders are always truthful about their abilities and limitations, and share honest opinions and feelings with their teams, it will help build trustworthy and authentic work relationships. As a result, employees will feel safe, secure, and comfortable being their genuine selves and foster a more connected, inclusive work culture.
Active listening
People are more willing to open themselves up to machines (when it comes to disclosing sensitive information or ideas generation) because they feel that machines do not judge them.
Cognitive A.I. chatbots take a step further by actively listening to their users, for example, repeating what their users have said and acknowledging users’ feelings empathetically to make users feel heard. More importantly, cognitive A.I. chatbots never place judgments on what users have expressed. Likewise, if organization leaders can practice active listening, remove egos and judgments from the room, and pay attention to their employees instead of their own thoughts, they could better connect with, understand, and lead their teams.
Reading between the lines
Like a psychologist, another important skill of cognitive A.I. chatbots is to read between the lines by dynamically analyzing a user’s conversation text and automatically inferring the user’s unspoken needs, implicit interests, and personality traits.
Cicero said more than 2000 years ago, “If you wish to persuade me, you must think my thoughts, feel my feelings, and speak my words.” Effective leaders should learn how to read between the lines from conversations with employees, empathizing with their feelings and thoughts and using those insights to best help them in the workplace.
Chatbots are now powered with advanced human soft skills, such as active listening and reading between the lines, which are also the most important leadership competencies, according to leaders around the world.
As humans continuously advance A.I. technologies, cognitive A.I. could also serve as a mirror for our own behavior–and help build more productive and inclusive relationships among humans. Leaders who embrace similar soft skills and lead with empathy, active listening, and transparency will build a much happier, more productive, and more inclusive workplace.
Michelle Zhou, Ph.D. is co-founder and CEO of Juji, Inc., the maker of accessible, cognitive artificial intelligence (A.I). assistants, and editor-in-chief for ACM Transactions on Interactive Intelligent Systems (TiiS). Prior to starting Juji, Michelle led and managed the research and development of cutting-edge technologies at IBM Research/Watson, including IBM RealHunter and Watson Personality Insights. Michelle received a Ph.D. in Computer Science from Columbia University and is an ACM Distinguished Scientist.
The opinions expressed in Fortune.com commentary pieces are solely the views of their authors and do not reflect the opinions and beliefs of Fortune.