人工智能作為強大新型學(xué)習(xí)方式的催化劑。 這篇文章為什么寫這么長，多半是AI幫著寫的。 作者：約翰·哈格爾、戴夫·圖爾、塔拉·曼德雷卡和約翰·西利·布朗

H-CORPS

2026年4月1日

關(guān)于人工智能的文章層出不窮。毫無疑問，這是一項非常強大的新技術(shù)，但它作為推動新型學(xué)習(xí)方式發(fā)展的催化劑的作用卻鮮為人知。盡管人工智能蘊藏著巨大的未開發(fā)潛力，但一些重大障礙阻礙了這些新型學(xué)習(xí)方法的廣泛應(yīng)用。我們需要采取一些措施，幫助更多的人接受這些新的學(xué)習(xí)方式，從而在事業(yè)和生活的其他方面取得更大的成就，并在這一過程中享受樂趣。

新的學(xué)習(xí)方式

我們都熟悉傳統(tǒng)的學(xué)習(xí)方法：參加課程或培訓(xùn)項目，聽老師講課，記住老師講的內(nèi)容，然后應(yīng)用所學(xué)知識。這種學(xué)習(xí)方式高度標(biāo)準(zhǔn)化、工業(yè)化，側(cè)重于分享已有的知識。

人工智能催生了哪些新的學(xué)習(xí)方式？這涉及諸多方面：

• 他們將不同背景的人聚集在一起，幫助他們彼此互動。當(dāng)人們與來自不同背景和觀點的小群體建立深厚聯(lián)系時，學(xué)習(xí)效率會更高。

• 它們更加個性化，能夠更深入地理解參與者的具體情況，并且會隨著情況的變化而發(fā)展。當(dāng)學(xué)習(xí)內(nèi)容聚焦于人們的具體情況時，他們的學(xué)習(xí)效果會更好。

• 它們培養(yǎng)了多種能力，包括好奇心、協(xié)作能力、想象力、創(chuàng)造力和反思能力。當(dāng)人們從與學(xué)習(xí)相關(guān)的常規(guī)任務(wù)中解放出來時，他們就能將時間和精力投入到培養(yǎng)有助于他們以更豐富的方式學(xué)習(xí)的能力上。

• 他們注重在行動中學(xué)習(xí)，而不僅僅是空談。如果我們觀察和反思行動的實際結(jié)果，而不是僅僅討論可能的結(jié)果，就能獲得更深刻的理解。

• 它們能激發(fā)玩樂精神，鼓勵參與者探索、嘗試和實驗。當(dāng)我們與他人一起進(jìn)入玩樂模式時，我們更愿意冒險，也更能看到以前未曾注意到的事物。

• 他們模擬多種可能的結(jié)果，以探索哪些行動能產(chǎn)生最大的影響。我們常常過于關(guān)注一兩種潛在結(jié)果，而沒有意識到可能存在更有效的結(jié)果。

• 它們支持快速迭代。迭代速度越快，我們就能越迅速有效地改進(jìn)方法，從而獲得更好的結(jié)果。

這些新型學(xué)習(xí)方式的關(guān)鍵基礎(chǔ)在于影響力小組的組建。影響力小組由3至15名背景各異的成員組成，他們彼此之間建立起深厚的信任關(guān)系，共同致力于在特定領(lǐng)域產(chǎn)生更大的影響力。他們意識到，無論每個人多么聰明或才華橫溢，通過合作并利用人工智能技術(shù)，他們都能更快地學(xué)習(xí)。

人工智能技術(shù)提供人工智能代理、數(shù)字孿生等模擬工具以及多種解決方案，以幫助促進(jìn)參與者之間的互動，從而使他們能夠更有效地想象和評估這些新方法。

案例研究：利用人工智能解決方案進(jìn)行學(xué)習(xí)

Anthropos 首席執(zhí)行官 Stephano Bellasio 提供了一個通過模擬進(jìn)行學(xué)習(xí)的例子。他分享了 150 多個基于 6 萬項技能的 AI 模擬案例，這些案例能夠模擬新銷售人員與客戶的互動方式。在這些模擬中，招聘經(jīng)理利用 AI 模擬來評估銷售人員的能力并識別技能差距。更多的模擬案例可以幫助設(shè)計 AI 支持的學(xué)習(xí)路徑，從而提升新入職銷售人員的績效。

Work3 Institute 的聯(lián)合創(chuàng)始人兼首席執(zhí)行官 Deborah Perry Piscione 分享了他們?nèi)绾翁剿?Asana AI Teammates，以構(gòu)建利用機器學(xué)習(xí)分析海量員工數(shù)據(jù)并深入了解其不斷變化的工作環(huán)境的基礎(chǔ)架構(gòu)。這些數(shù)據(jù)包括技能、經(jīng)驗、性格特征、溝通風(fēng)格和績效指標(biāo)。員工的興趣愛好也可能被納入考量?！斑@些非常復(fù)雜的算法能夠發(fā)現(xiàn)人類管理者可能忽略的模式和洞察，例如員工的潛在才能、互補的工作方式，甚至是他們面臨的個人問題，”她說道。

皮西奧內(nèi)表示：“人工智能旨在增強和支持人類決策，而不是完全取代它。雖然人工智能可以提供寶貴的見解和建議，但在應(yīng)對團隊動態(tài)的細(xì)微差別和復(fù)雜性時，人類的判斷仍然至關(guān)重要，它能夠幫助我們判斷特定員工何時最有效率或最具創(chuàng)造力?！?/p>

像 Moderna、施耐德電氣等公司已經(jīng)建立了人才網(wǎng)絡(luò)數(shù)據(jù)集，利用人工智能技術(shù)幫助全球團隊協(xié)作，從而提升創(chuàng)造力并挖掘新的機遇。這些公司正在探索新的團隊組合方式，以期改善工作成果。

從某種意義上說，這些新的學(xué)習(xí)方法有助于我們更有效地融合智人（思考者）、創(chuàng)造者（建造者）和游戲者（玩耍者）的實踐。我們很少會意識到玩耍是學(xué)習(xí)的關(guān)鍵組成部分，尤其是在商業(yè)領(lǐng)域。但正如約翰·赫伊津哈在其經(jīng)典著作《游戲者》中所指出的，玩耍是人類文化和認(rèn)知的基礎(chǔ)和塑造因素。人類的學(xué)習(xí)和文化是通過充滿樂趣的探索、即興創(chuàng)作和實驗而發(fā)展的。在玩耍中，動力源于內(nèi)心。事實上，當(dāng)人們擁有內(nèi)在動力而非僅僅為了達(dá)成某些外部目標(biāo)時，他們的學(xué)習(xí)效率會更高。簡而言之，《游戲者》將學(xué)習(xí)重新定義為一個由好奇心、探索精神和內(nèi)在樂趣驅(qū)動的、本質(zhì)上充滿樂趣的過程。本文將探討如何利用人工智能的應(yīng)用，將這種好奇心和實驗精神融入到我們以全新方式進(jìn)行構(gòu)建和學(xué)習(xí)的努力中。

我們應(yīng)該澄清，之前提到的“新的學(xué)習(xí)方式”其實并非全新。事實上，在人類歷史的大部分時間里，我們都在實踐這些學(xué)習(xí)方式，直到現(xiàn)代教育體系主導(dǎo)了兒童的學(xué)習(xí)。這些教育體系推行了一種截然不同的學(xué)習(xí)模式，而這種模式也被我們畢業(yè)后進(jìn)入的機構(gòu)所接受。這種較新的學(xué)習(xí)方式在變化相對較小的時代或許行之有效，但隨著時代變遷的加速，我們相信回歸那些塑造了人類歷史大部分時期的學(xué)習(xí)方式，仍然具有重要的價值。

新學(xué)習(xí)方式的力量

為什么這些新的學(xué)習(xí)方式如此有用？為什么它們比傳統(tǒng)的學(xué)習(xí)方式更有影響力？

它們幫助不同領(lǐng)域（不僅限于商業(yè)領(lǐng)域）的人們在瞬息萬變的世界中發(fā)揮更大的影響力。在這個世界里，我們需要能夠更輕松地解讀快速變化的環(huán)境，理解如何轉(zhuǎn)換環(huán)境并學(xué)習(xí)如何最大限度地發(fā)揮自身潛力。

一些學(xué)生已經(jīng)開始利用人工智能來優(yōu)化他們的學(xué)習(xí)和創(chuàng)業(yè)活動。在加州橙縣的一所社區(qū)大學(xué)，一位名叫克里斯蒂安·洛佩茲的學(xué)生開發(fā)了人工智能代理，這些代理會分析他收到的課程大綱，并提供高階學(xué)習(xí)計劃，以確保他的學(xué)習(xí)習(xí)慣盡可能高效。課程結(jié)束后，學(xué)生的代理會反思他在課程中的學(xué)習(xí)成果，并提出改進(jìn)未來學(xué)習(xí)效果的建議，從而讓他有更多時間從事其他工作。借助人工智能，克里斯蒂安創(chuàng)建了一條提升自身技能的途徑，這條途徑比傳統(tǒng)的學(xué)習(xí)方式更加個性化、深入和高效。

來自橙縣的另一位學(xué)生阿迪亞·賈達(dá)夫（Aditya Jadhav）利用Notion AI、Chat-GPT、Notebook LM和谷歌Firebase Studio等人工智能工具，在學(xué)校創(chuàng)建并領(lǐng)導(dǎo)了一個人工智能俱樂部，同時還要兼顧31個學(xué)分的課程學(xué)習(xí)和拓展人脈。領(lǐng)導(dǎo)一個人工智能愛好者社群的挑戰(zhàn)，為阿迪亞提供了一個至關(guān)重要的領(lǐng)導(dǎo)力發(fā)展機會。他的人工智能俱樂部也為同學(xué)們開辟了新的學(xué)習(xí)和成長途徑。阿迪亞定期為俱樂部成員舉辦研討會，將人工智能的倫理理解與實踐性的黑客馬拉松和挑戰(zhàn)相結(jié)合，幫助參與者提升技術(shù)和人工智能素養(yǎng)。人工智能不僅為他們探索其對未來的影響提供了新的途徑，也為阿迪亞提供了一個提升個人領(lǐng)導(dǎo)技能的機會，讓他能夠利用人工智能工具輔助信息傳遞、學(xué)習(xí)如何與成員社交以及設(shè)計實踐學(xué)習(xí)體驗。 Aditya 表示，他的許多同齡人都希望更好地了解人工智能，幫助他們打好必要的基礎(chǔ)，也能為他們提供支持未來成功的關(guān)鍵技能。

在瞬息萬變的世界中，我們需要預(yù)見如何創(chuàng)造更多價值，并通過與擁有不同視角和經(jīng)驗的人攜手合作來克服意想不到的障礙。一種引人注目的新方法是合作共贏。

影響力小組的成員們不斷互相激勵，力求產(chǎn)生更大的影響力，并在項目未能達(dá)到預(yù)期效果時互相支持。他們致力于不斷迭代，邁向未來。

個性化、適應(yīng)性、協(xié)作、想象力、行動學(xué)習(xí)以及豐富且實時的反饋，有助于人們在快速變化的世界中更有效地學(xué)習(xí)?？焖俚蔀橛行W(xué)習(xí)循環(huán)的關(guān)鍵要素。這是一個持續(xù)的過程。

人工智能可以提供哪些具體工具來支持這些新的學(xué)習(xí)方式？

人工智能提供了眾多工具來支持這些新的學(xué)習(xí)方式。其中包括：

1. 代理人可以幫助我們識別并聯(lián)系到影響群體中更多元化的參與者群體，并幫助我們找到可能與我們的學(xué)習(xí)目標(biāo)相關(guān)的數(shù)據(jù)和信息。

2. 模擬工具可以幫助我們設(shè)想正在考慮采取的行動可能帶來的后果。

3. 預(yù)測分析可以幫助我們預(yù)測我們所處環(huán)境可能發(fā)生的演變。

4. 推薦系統(tǒng)可以幫助我們評估可能采取的行動的利弊。

5. 實時反饋機制可以為我們提供更豐富、更快速的行動反饋。

6. 通過協(xié)作提示，將團隊聚集在一起，創(chuàng)建場景，并深入探討如何利用人工智能解決問題。

這些工具固然有用，但如果整合到自適應(yīng)學(xué)習(xí)平臺中，它們的價值將更加凸顯。這些平臺能夠匯聚眾多參與者，并擴大其影響力。這些學(xué)習(xí)平臺與我們現(xiàn)有的學(xué)習(xí)平臺截然不同。如今的學(xué)習(xí)平臺只是簡單地聚合課程和講座，以支持傳統(tǒng)的學(xué)習(xí)方式。而這些新型學(xué)習(xí)平臺則有效地整合了人工智能提供的眾多工具，以支持新型學(xué)習(xí)方式。它們的主要組織組成部分是共享工作空間，這些空間能夠支持影響力小組，幫助他們探索這些新的學(xué)習(xí)方式，同時將他們連接到更廣泛的網(wǎng)絡(luò)中，從而擴大學(xué)習(xí)規(guī)模。

從根本上講，人工智能可以將人們從如今耗費大量時間的日常工作中解放出來，使他們能夠集中精力學(xué)習(xí)如何創(chuàng)造更多價值。許多高管都關(guān)注人工智能如何節(jié)省時間，但他們的目標(biāo)是實現(xiàn)任務(wù)自動化并盡可能地裁員。然而，他們更應(yīng)該思考的是，如何讓從眾多日常工作中解放出來的人們能夠利用人工智能的潛力來支持新的學(xué)習(xí)方式。這些新的學(xué)習(xí)方式可以為客戶和其他利益相關(guān)者帶來更大的價值。例如，一家大型公司的呼叫中心客服人員能夠更深入地關(guān)注客戶提出的意外問題和難題，從而顯著提高客戶滿意度。人工智能接管了日常通話的處理，并提供了工具來幫助呼叫中心客服人員解決一些他們以前從未遇到過的更具挑戰(zhàn)性的問題。

例如，人工智能可以幫助人們更好地構(gòu)建和分析客戶畫像，探索各種可能的應(yīng)用場景，并找到職業(yè)發(fā)展的新方法。它模擬智能對話的能力使其成為項目式學(xué)習(xí)和問題式學(xué)習(xí)的理想選擇，因為在這種學(xué)習(xí)模式下，參與者需要探索并迭代解決現(xiàn)實世界的問題。

人工智能是通往更高層次視角和思維方式的墊腳石，它建立在所有為我們的知識庫做出貢獻(xiàn)的人的肩膀之上。通過人工智能進(jìn)行實驗，可以在投入資金之前積累洞察力和信心。

在亞利桑那州的一間教室里，這種實驗正在為學(xué)生們帶來新的突破，這將對他們的職業(yè)生涯大有裨益。豪爾赫·科斯塔是一位教授人工智能音樂制作課程的教育家，他利用人工智能技術(shù)在課堂上營造輕松愉快的氛圍。他所創(chuàng)造的學(xué)習(xí)空間以學(xué)生的自主性為中心，旨在幫助許多學(xué)生克服對被人工智能取代的恐懼，轉(zhuǎn)而探索人工智能的潛力，從而在他們熱愛的領(lǐng)域中發(fā)展，并在大學(xué)畢業(yè)后拓展他們的職業(yè)選擇。

他的學(xué)生杰米分享說，她最終接受了一份原本拒絕的工作，因為她從科斯塔的課程中獲得了自信。這份工作需要創(chuàng)作廣告歌曲，而她以前從未做過。雖然科斯塔的課并沒有教她如何創(chuàng)作廣告歌曲，但課堂上接觸到的Suno AI等人工智能工具，為杰米提供了接受新的創(chuàng)意和職業(yè)挑戰(zhàn)所需的信心和知識。

豪爾赫與其中一位作者合作，探索在餐桌上利用人工智能進(jìn)行學(xué)習(xí)的新方法。戴夫·圖爾接受了一項挑戰(zhàn)：借助人工智能創(chuàng)作一首歌，目的并非取代他的創(chuàng)造力，而是為了拓展他的能力。他是一位吉他手，此前從未創(chuàng)作過歌詞、旋律和完整的歌曲。他用歌曲的含義訓(xùn)練了ChatGPT，幫助他創(chuàng)作歌詞。這讓他迅速建立起對歌詞創(chuàng)作的信心，并讓他有更多時間去探索旋律的創(chuàng)作。如今，他已經(jīng)掌握了無需人工智能輔助即可創(chuàng)作歌詞的新技能。戴夫利用蘋果錄音平臺上的人工智能，創(chuàng)作出了后來被稱為“戴夫與不眠騎士”的作品。歌曲的原型經(jīng)過混音后，現(xiàn)場樂手代替了數(shù)字輔助樂器進(jìn)行演奏。在歌詞創(chuàng)作（類似于商業(yè)信息傳遞）、歌曲原型制作（類似于商業(yè)產(chǎn)品驗證）以及利用人工智能進(jìn)行新學(xué)習(xí)后協(xié)調(diào)人際協(xié)作的過程中，涌現(xiàn)出了新的學(xué)習(xí)方法。

在商業(yè)領(lǐng)域，人工智能賦能的實驗可以帶來變革。GLIDR.ai 的首席執(zhí)行官 Jim Hornthal 正在利用 GLIDR.ai 重塑各行各業(yè)的創(chuàng)新模式。GLIDR.ai 是一個原生人工智能平臺，能夠引導(dǎo)創(chuàng)新團隊將想法轉(zhuǎn)化為洞察和實際影響。其應(yīng)用范圍涵蓋生命科學(xué)、國防、能源、教育、政府、非營利組織和風(fēng)險投資等眾多領(lǐng)域。

通過整合機器學(xué)習(xí)、實時市場情報和結(jié)構(gòu)化實驗，GLIDR 已幫助企業(yè)家、研究人員、非營利組織領(lǐng)導(dǎo)者和企業(yè)創(chuàng)新者快速、嚴(yán)謹(jǐn)?shù)卦O(shè)計、測試和優(yōu)化商業(yè)模式。GLIDR 基于精益創(chuàng)業(yè)加速器 (Lean LaunchPad) 方法論，該方法論引導(dǎo)初創(chuàng)企業(yè)完成從構(gòu)思業(yè)務(wù)、客戶、市場、驗證到商業(yè)化的九步流程。該流程最初由史蒂夫·布蘭克 (Steve Blank) 開發(fā)，后被美國國家科學(xué)基金會 (NSF) 采納，用于啟動和擴展其創(chuàng)新加速器項目 I-Corps。目前，GLIDR 已幫助全球超過 25,000 個團隊?wèi)?yīng)對不確定性，降低商業(yè)化和市場推廣投資的風(fēng)險。

借助作為嵌入式思考伙伴的人工智能代理，用戶能夠獲得切實可行的指導(dǎo)，從而驗證市場和產(chǎn)品假設(shè)，發(fā)掘進(jìn)入市場的機會空間，設(shè)計實驗進(jìn)行驗證，并使解決方案與利益相關(guān)者的需求保持一致，最終全面展現(xiàn)項目的吸引力、可行性和生存能力。GLIDR 正在重新定義創(chuàng)新：更快、更智能、更具可擴展性。

風(fēng)險投資公司NfX的風(fēng)險投資團隊最近分享了一個由三人組成的獨角獸企業(yè)模型，該模型包含一位富有遠(yuǎn)見的創(chuàng)始人、一位精于數(shù)據(jù)分析的專家和一位溝通能力卓越的領(lǐng)導(dǎo)者。他們采用人工智能助手來加速業(yè)務(wù)系統(tǒng)的運行。這些人工智能工具能夠加速他們的學(xué)習(xí)，從而創(chuàng)建新的商業(yè)模式和框架，提升產(chǎn)品和服務(wù)的價值。一些公司已經(jīng)開始使用人工智能助手在員工休息時工作，并提高了初創(chuàng)團隊的技能水平。據(jù)他們報告，投資組合中的公司在軟件和游戲開發(fā)領(lǐng)域的產(chǎn)出提高了100%。

需要克服哪些障礙和困難？

雖然這些新的學(xué)習(xí)方式為我們創(chuàng)造了更大的影響力，但我們?nèi)孕杩朔恍┳璧K我們前進(jìn)的重大障礙。這些障礙既存在于我們的環(huán)境中，也存在于我們自身。

最大的障礙在于世界各地教育體系和大型組織中盛行的文化，這種文化仍然以傳統(tǒng)的課堂和培訓(xùn)課程學(xué)習(xí)方式為中心。這種文化對這些新的學(xué)習(xí)方式往往持懷疑態(tài)度，甚至充滿敵意。

世界上所有大型機構(gòu)——包括公司、學(xué)校、非政府組織和政府——都遵循著一種共同的制度模式。這種制度模式的核心在于“可擴展的效率”——我們組織和運營機構(gòu)的方式，都以更快、更經(jīng)濟地完成當(dāng)前任務(wù)為目標(biāo)。

這種制度模式將工作定義為嚴(yán)格限定、高度標(biāo)準(zhǔn)化的任務(wù)，這些任務(wù)需要可靠高效地完成。為了定義和執(zhí)行這些任務(wù)，要實現(xiàn)可擴展的效率，就必須采用層級分明的組織結(jié)構(gòu)，并在各個層級明確定義角色和職責(zé)。

為了充分釋放人工智能的潛力，我們需要轉(zhuǎn)向一種截然不同的制度模式——可擴展的學(xué)習(xí)模式，更加強調(diào)共同實踐的學(xué)習(xí)方式。這種制度模式旨在激發(fā)探索者的熱情，并培養(yǎng)我們每個人內(nèi)在的一系列強大能力——好奇心、協(xié)作精神、想象力、創(chuàng)造力、反思能力和探索精神。

在這種新的制度模式下，工作不再被定義為嚴(yán)格限定的例行任務(wù)——這些任務(wù)將由機器完成。人類將專注于識別和把握機會，為利益相關(guān)者創(chuàng)造更大的價值。為此，他們將以小組（影響力小組）的形式聚集在一起，形成類似細(xì)胞的結(jié)構(gòu)。這些小組匯聚了對提升特定領(lǐng)域影響力、以創(chuàng)新方式解決問題充滿熱情的人們。遠(yuǎn)程和實體團隊能夠圍繞新興的集體學(xué)習(xí)建立緊密的聯(lián)系。

人工智能將通過為參與者提供培養(yǎng)自身能力的工具，幫助這些影響力群體創(chuàng)造更大的價值。然而，挑戰(zhàn)在于，當(dāng)前可擴展的效率模式與可擴展學(xué)習(xí)所需的流程和實踐存在著根本性的沖突，這些大型機構(gòu)的“免疫系統(tǒng)”和“抗體”會迅速啟動，扼殺任何向可擴展學(xué)習(xí)模式轉(zhuǎn)型的努力。因此，真正的挑戰(zhàn)在于如何克服巨大的阻力來推動變革。

我們面臨的另一個挑戰(zhàn)是，人工智能有可能取代我們的思維，而非增強我們的思維。數(shù)據(jù)顯示，雖然人工智能可以更快地將低技能水平的人提升到專家水平，但這種常規(guī)做法可能會錯過那些前沿的創(chuàng)新想法。我們需要有意識地使用人工智能，以確保我們不會失去人與人之間迸發(fā)的創(chuàng)造力火花。麻省理工學(xué)院發(fā)表了一項研究，題為《當(dāng)人工智能替你思考時，你的大腦會變成什么樣？》，指出如果使用不當(dāng)，人工智能反而會讓我們變得更笨。這項研究提醒我們，要區(qū)分人工智能通過學(xué)習(xí)拓展我們想象力的潛力，同時也要警惕它可能導(dǎo)致我們創(chuàng)造力倒退或抑制前沿探索的能力。

還有另一系列障礙和阻力。幾個世紀(jì)以來，我們目睹了社會變得越來越集中化。高度集中的社會往往會削弱基層創(chuàng)新和行動的潛力，人們也更容易變得更加被動。

我們還看到，那些能夠?qū)⑷藗兙奂谝黄?、建立更深層次、更廣泛的人際關(guān)系的社交場所正在不斷減少。我們變得越來越孤立，對執(zhí)政者的信任也在逐漸喪失。

為了幫助每個人培養(yǎng)自主意識，我們需要轉(zhuǎn)向更加去中心化的社會組織形式。我們還需要營造社交場所，幫助人們建立更深層次、更廣泛的聯(lián)系，以便他們能夠聚集在一起，擴大行動規(guī)模，并在實踐中學(xué)習(xí)。但是，我們必須再次認(rèn)識到，我們現(xiàn)有的中心化社會將會抵制去中心化。這絕非易事。

在另一個層面上，恐懼也是一道重要的情感障礙，隨著我們意識到世界變化之快，越來越多的人開始感受到這種恐懼。科技發(fā)展的日新月異使得掌握不斷演進(jìn)的最新人工智能工具變得極具挑戰(zhàn)性。當(dāng)我們意識到自己缺乏學(xué)習(xí)經(jīng)驗，無法掌握這些快速發(fā)展的技術(shù)所帶來的新型學(xué)習(xí)方式時，恐懼感便會與日俱增。此外，人們還普遍擔(dān)心會被人工智能取代。我們需要幫助人們克服這種恐懼，重點關(guān)注人工智能為我們所有人創(chuàng)造的機遇，讓我們能夠從事更具創(chuàng)造性的工作，從而為那些真正重要的人帶來更大的影響。

挖掘尚未開發(fā)的潛力

我們需要找到克服這些障礙和阻礙的方法，以便充分發(fā)揮人工智能作為新型學(xué)習(xí)方式催化劑的潛力。

與其試圖一次性改變所有事情，我們更應(yīng)該有選擇地開展能夠更好地服務(wù)客戶和合作伙伴的舉措。我們需要尋找切入點，以便用相對較小的投入快速取得顯著成效。巧妙的小舉措，往往能引發(fā)大的變化。

這些小小的舉動能幫助人們克服恐懼，因為它們不需要付出太多努力，而且可以相對快速地完成。當(dāng)人們看到其他人采取的小舉動產(chǎn)生了實際的影響時，也能幫助其他人克服恐懼。

這些小舉措的一個有趣例子是寶潔公司去年推出的一項實驗。

寶潔公司近期舉辦了一場別開生面的“線上黑客馬拉松”，776名來自商業(yè)和研發(fā)部門的員工齊聚一堂，參加了一場為期一天的線上研討會。研討會的目標(biāo)是為各自部門面臨的實際業(yè)務(wù)挑戰(zhàn)開發(fā)新的解決方案。參與者被隨機分配到以下四種情境之一：個人或兩人團隊，以及是否擁有生成式人工智能（Gen AI）的使用權(quán)限。

研究結(jié)果令人信服：使用人工智能技術(shù)的團隊完成任務(wù)的速度比未使用人工智能技術(shù)的團隊快約 12%。寶潔公司的研究表明，對于消費品公司而言，人工智能技術(shù)作為輔助手段而非替代人際互動，才能實現(xiàn)最佳績效。

除了速度提升之外，該研究還強調(diào)了人工智能在跨部門協(xié)作方面的顯著優(yōu)勢。人工智能促進(jìn)了來自不同背景的員工，無論其個人專長如何，都能共同開發(fā)出更加平衡的解決方案。在社會情感層面，人工智能的融入對員工士氣產(chǎn)生了積極影響，提升了整體工作體驗。這表明，戰(zhàn)略性地應(yīng)用人工智能不僅可以提高效率，還能促進(jìn)更好的團隊合作，營造更積極的工作環(huán)境。

西門子、歐特克和其他公司提供的AI仿真系統(tǒng)能夠創(chuàng)建制造流程的“數(shù)字孿生”，讓來自工程、運營和業(yè)務(wù)部門的小團隊能夠在無風(fēng)險的環(huán)境中協(xié)作試驗流程改進(jìn)方案，例如，在投入預(yù)算實施之前，團隊可以先模擬工廠流程。這種方法打破了長期存在的組織壁壘，在各部門之間建立了共享的背景信息。團隊發(fā)現(xiàn)了多年來被忽視的優(yōu)化機會，最終實現(xiàn)了15%的效率提升，并為長期存在的制造難題找到了新的解決方案。

由丹·貝·金（Dan Be Kim）和布萊里姆·賈沙里（Blerim Jashari）開發(fā)的哈佛大學(xué)人工智能學(xué)習(xí)工作室，有力地展現(xiàn)了人工智能在構(gòu)建可擴展學(xué)習(xí)環(huán)境方面的潛力。金和賈沙里設(shè)計了一門課程，旨在通過提供結(jié)構(gòu)化的機會，讓參與者“動手實踐”人工智能和課程設(shè)計，從而轉(zhuǎn)變參與者與人工智能的關(guān)系，使其從“不知所措”轉(zhuǎn)變?yōu)槌錆M好奇心和自主性。

研究人員反思了他們以項目式學(xué)習(xí)和跨學(xué)科合作為核心的教學(xué)方法，認(rèn)為這是對人工智能不斷發(fā)展的回應(yīng)。人工智能這項技術(shù)既挑戰(zhàn)了傳統(tǒng)的教育模式，也為我們的學(xué)習(xí)方式開啟了全新的可能性。他們指出：“我們的課程某種程度上驗證了我們需要更多樣化的學(xué)習(xí)分組方式；人工智能就像一種奇妙的粘合劑，促進(jìn)了這種多樣性?！比斯ぶ悄懿粌H為重新定義學(xué)習(xí)空間本身提供了催化劑，也為有意義且富有創(chuàng)意的跨學(xué)科合作搭建了平臺。

為了將人工智能作為促進(jìn)新學(xué)習(xí)方式和克服前進(jìn)道路上障礙的催化劑，我們建議采取以下基于小步快跑、巧妙推進(jìn)原則的方法：

1. 確定這項計劃的領(lǐng)導(dǎo)者。理想情況下，這個人應(yīng)該是首席執(zhí)行官，或者至少是向首席執(zhí)行官匯報的人員。關(guān)鍵在于找到一位能夠看到新型學(xué)習(xí)方式的潛力，并且即使面臨巨大阻力，也充滿熱情地探索這些新型學(xué)習(xí)方式的人。

2. 確定目標(biāo)領(lǐng)域。尋找組織內(nèi)一個具體但規(guī)模較小的部門，該部門能夠迅速有效地從學(xué)習(xí)中獲益。尋找至少部分參與者對學(xué)習(xí)充滿熱情，并渴望為相關(guān)利益攸關(guān)方帶來更大影響的部門。

3. 組建影響小組。將3-15名擁有不同技能和背景的人員組成小組。確保每個小組中至少有一些人對他們的工作充滿熱情。為這些小組提供相關(guān)的AI工具，并簡要概述他們?nèi)绾问褂眠@些工具來學(xué)習(xí)更多知識。

4. 提出一些啟發(fā)性的問題。向這些影響力群體提出一些關(guān)于潛在機遇或挑戰(zhàn)的問題，這些問題可能顯著提升他們的影響力。鼓勵他們利用人工智能工具，提出能夠顯著提升影響力的新方法。

5. 反思“新的學(xué)習(xí)方式”。當(dāng)這些影響力小組開始取得成果時，讓他們停下來反思人工智能工具如何幫助他們探索新的學(xué)習(xí)方式并取得更大的影響力。建立一個實踐案例庫。

6. 激勵組織內(nèi)的其他成員。在整個組織內(nèi)宣傳這些影響力小組取得的成就，以及他們?nèi)绾卫萌斯ぶ悄芄ぞ咛剿餍碌膶W(xué)習(xí)方式。讓這些工具更廣泛地普及，并將其融入正式的學(xué)習(xí)和發(fā)展項目中。隨著其他小組開始利用這些新工具取得成效，繼續(xù)宣傳推廣，并重點關(guān)注他們是如何使用這些工具的。

7. 培育一種新文化。隨著人們對人工智能工具在促進(jìn)新型學(xué)習(xí)方式方面的潛力日益興奮，應(yīng)著重培育一種在整個組織內(nèi)推行可擴展學(xué)習(xí)的新文化。激發(fā)所有參與者的探索熱情，并鼓勵領(lǐng)導(dǎo)者不斷提出啟發(fā)性問題，以激勵參與者積極學(xué)習(xí)。

人工智能技術(shù)正以驚人的速度發(fā)展，蘊藏著巨大的未開發(fā)潛力。它能夠提供工具和方法，支持全新的學(xué)習(xí)方式，從而創(chuàng)造更大的價值。之所以說它尚未被充分發(fā)揮潛力，是因為在接受這種新型學(xué)習(xí)方式的過程中，我們面臨著一些重大的障礙和阻礙。然而，有一些方法可以幫助我們釋放人工智能在促進(jìn)新型學(xué)習(xí)方式方面的全部潛力。這些方法已經(jīng)在某些領(lǐng)域得到應(yīng)用，我們需要探索如何在更廣泛的領(lǐng)域中推廣這些方法。

AI as a Catalyst for Powerful New Ways of Learning By John Hagel, Dave Toole, Tara Mandrekar, and John Seely Brown

H-CORPS

APR 01, 2026

Endless articles are appearing about artificial intelligence. There is no doubt that this is a very powerful new technology, but not enough attention is being paid to its role as a catalyst for powerful new ways of learning. While this has major unaddressed potential, some significant barriers stand in the way of widespread adoption of these new approaches to learning. We need to adopt approaches that will help many more people embrace these new ways of learning and achieve much greater impact, both in business and in other dimensions of their lives and have fun along the way.

New ways of learning

We’re all familiar with conventional approaches to learning. Go to classes or training programs, listen to the teacher, memorize what the teacher has to say and then apply this learning. This way of learning is highly standardized, industrialized and focused on sharing existing knowledge.

What are the new ways of learning that are enabled by AI? There are many dimensions:

• They bring groups of diverse people together and help them to interact with each other. People can learn more effectively when they connect deeply with a small group of people from different backgrounds and perspectives

• They are much more personalized, with a richer reading of the specific context of the participants, and they evolve as the context evolves. People learn better when the learning is focused on their specific context.

• They cultivate a diverse set of capabilities, including curiosity, collaboration, imagination, creativity, and reflection. When people are freed up from the more routine tasks associated with learning, they can spend their time and energy on cultivating capabilities that help them to learn in a much richer way.

• They focus on learning through action, not just converse. We will gain a lot deeper insight when we observe and reflect on the actual results of action taken rather than simply talking about what those results might be.

• They foster playful behavior so that participants are encouraged to explore, tinker, and experiment. When we move into play mode with others, we are much more willing to take risks and see things which were previously not visible.

• They simulate many alternative outcomes to explore which actions could have the greatest impact. We often get too narrowly focused on one or two potential outcomes, without realizing that there may be even more productive outcomes.

• They support rapid iteration. The more rapidly we iterate, the more quickly and effectively we can refine our approaches to yield even better results.

A key foundation for these new ways of learning is the formation of impact groups. These are small groups of 3-15 diverse people who build deep, trust-based relationships with each other, driven by a desire to have more and more impact in a specific domain. They realize that no matter how smart or talented each individual is, they will learn faster by coming together and leveraging AI in their collaboration.

AI technology provides simulation tools like AI agents, digital twins and a multitude of solutions to help facilitate interactions among participants so they can more effectively imagine and assess these new approaches.

Case Studies: Learning with AI-powered solutions

One example of learning through simulations is provided by Stephano Bellasio, CEO of Anthropos. He shared a few of over 150 AI simulations based on 60,000 skills that help to simulate how a new sales prospect would interact with a customer. In these simulations a hiring manager leverages AI simulations to assess the capabilities of sales prospects and identify skills gaps. Additional simulations could help to design AI supported learning journeys to improve the performance of the sales prospects once hired.

Deborah Perry Piscione, co-founder and CEO of the Work3 Institute shared how they have been exploring Asana AI Teammates to build the scaffolding in harnessing the power of machine learning to analyze a wealth of employee data and provide more insight into their evolving context. This data includes skills, experience, personality traits, communication styles, and performance metrics. The hobbies and passions of employees may also be included in the mix. “These very sophisticated algorithms uncover patterns and insights that human managers might overlook, such as hidden talents, complementary working styles, or even perhaps a personal issue that an employee is faced with,” she said.

Piscione stated, “AI is meant to augment and support human decision-making, not replace it entirely. While AI can provide valuable insights and recommendations, human judgment remains essential in navigating the nuances and complexities of team dynamics, weighing in when a particular employee is most productive or creative.”

Organizations like Moderna, Schneider Electric and others have developed a talent network data set to use AI to help team up across the globe, increasing creativity and unseen opportunities. These organizations are exploring new ways to experiment with team combinations to improve outcomes.

In a profound way, these new approaches to learning help us to more effectively blend the practices of homo sapiens (human who thinks), homo faber (human who builds) and homo ludens (human who plays). We seldom think play is a crucial part of learning, especially in the business world, but as Johan Huizinga points out in his classic book, Homo Ludens, play is a fundamental and formative aspect of human culture and cognition. Human learning and culture develop through playful explorations, improvisations and experimentation. In play, motivation comes from within. Indeed, people learn more effectively when they are intrinsically motivated rather than just trying to meet some external goals. In short, homo ludens reframes learning as a fundamentally playful process driven by curiosity, exploration and intrinsic joy. This paper will explore the potential to leverage AI’s applications to center this curiosity and experimentation in our efforts to build and learn in new ways.

We should clarify that the “new ways of learning” that we outlined earlier, are really not that new. In fact, throughout most of human history, we pursued these forms of learning until our modern educational systems took over the learning of children. These educational systems imposed a very different form of learning that was also embraced by the institutions we “graduated” into after school. This more recent form of learning was helpful in a world that did not change that much, but, as the pace of change accelerates, we believe that there can be significant value in returning to the ways of learning that shaped much of human history.

For more examples oflearning with AI-powered simulations—or to add examples of your own—click here to access our case study repository.

The power of new ways of learning

Why are these new ways of learning so useful? Why will they deliver far more impact than traditional ways of learning?

They help people in many different contexts (not just the business world) to achieve more impact in a rapidly changing world. In this world, we need to be able to read our quickly evolving context more easily and understand how we can shift contexts and learn to optimize our potential.

Some students have begun to leverage AI to optimize their educational and entrepreneurial efforts. At a community college in Orange County, CA, a student named Christian Lopez has developed AI agents that review his incoming syllabi and offer high-level study plans to ensure his habits are as effective as possible. After completing a course, the students’ agents then reflect on his success in the course and what can be done to improve future outcomes and free up more time for other endeavors. Using AI, Christian created a pathway to developing his skillset with more personalization, depth, and speed than he would solely achieve in the traditional learning environment.

Another student in Orange County, Aditya Jadhav, has leveraged AI tools like Notion AI, Chat-GPT, Notebook LM, and Google’s Firebase Studio to help him start and lead an AI Club on his school’s campus while managing a 31-unit course load and building his professional network. The challenge of leading a community of AI enthusiasts provided Aditya with a crucial opportunity to develop leadership skills. His AI Club has also fostered new avenues for learning and growth among his peers. Aditya leads regular workshops with club members that balance building ethical understanding of AI with hands-on hackathons and challenges, allowing participants to build technical and AI literacy skills. AI served as a vehicle for new exploration around its impacts on their future and provided an opportunity for Aditya to develop his personal skillset as a leader, using AI tools to assist with messaging, learn how to socialize with other members, and design hands-on learning experiences. Aditya reported that many of his peers want to better understand AI, and enabling them to develop the foundations necessary to do so also provides them with critical skills that will support their future success.

In a rapidly changing world, we need to anticipate ways to create more value and overcome unexpected obstacles by coming together with others who provide different perspectives and experiences. A compelling new way to do this is to play together to get things done.

Participants in impact groups are continually challenging each other to come up with even greater impact and supporting each other when initiatives fail to deliver the impact expected. They are committed to iterating towards the future.

Personalization, adaptation, collaboration, imagination, learning through action, and rich and real-time feedback help people to learn more effectively in a rapidly changing world. Rapid iteration becomes a key requirement for effective cycles of learning. It is a continuous process.

For more examples ofAI tools increasing students’ capacity to excel—or to add examples of your own—click here to access our case study repository.

What specific tools can AI provide to support these new ways of learning?

AI offers numerous tools to support these new ways of learning. Among these tools are:

1. Agents that can help us to identify and connect with a more diverse group of participants in an impact group, as well as helping us to locate data and information that might be relevant to our learning quest

2. Simulation tools that can help us to imagine potential outcomes from actions we are considering pursuing

3. Predictive analytics that can help us to anticipate how our context might be evolving

4. Recommendation systems that can help us evaluate the pros and cons of actions that we might pursue

5. Real-time feedback mechanisms that can provide us with much richer and more rapid feedback on the actions that we pursue

6. Collaborative prompting to bring teams together to create scenarios and dive deeper into problem solving with AI at the table.

These tools are all useful, but they provide even more value when integrated into adaptive learning platforms that can bring many participants together and scale the impact that they can achieve. These learning platforms are very different from the learning platforms that we have today. Today’s learning platforms simply aggregate classes and lectures to support traditional forms of learning. These new learning platforms effectively integrate the many tools that AI can provide to support new forms of learning. Their primary organizational component is shared workspaces that can support impact groups and help them to pursue these new ways of learning, while connecting them into broader networks that can scale their learning.

More fundamentally, AI can free up human beings from the routine tasks that consume much of their time today so that they can come together and focus on learning how to create more value. Many executives are focusing on time savings with AI. But their goal is to automate tasks and eliminate as many people as possible. Instead, they should consider how people who are freed up from so many routine tasks can harness the potential of AI to support new ways of learning. These new ways of learning could deliver far more value to their customers and other stakeholders. As an example, operators in the call center of a large company were able to focus much more deeply on the unexpected questions and problems of the customers they were dealing with, leading to much greater customer satisfaction. AI took over handling the routine calls and provided tools to help call center operators address some of the more challenging questions that they had never encountered before.

As an example of how these tools can support new ways of learning, AI enables individuals to better develop and analyze customer personas, explore likely scenarios, and find new methods for developing in their profession. Its ability to simulate intelligent conversation makes it ideal for project-based and problem-based learning, where participants are required to explore and iterate on real-world problems.

AI is a stepping stone to higher levels of perspective and disposition that stands on the shoulders of all those that have contributed to our database of knowledge. Experimentation through AI builds insight and confidence prior to expenditure.

In a classroom in Arizona, this kind of experimentation is leading to new breakthroughs for students that will serve them in their careers. Jorge Costa, an educator leading an AI in Music Production course, has leveraged AI in his classroom to promote playfulness. The space he’s fostered centers agency and aims to empower many of his students to overcome fear of replacement by AI, and instead explore AI’s potential to build on their passions and broaden their set of possibilities after leaving university.

One of his students, Jamie, shared that she ended up taking a job offer that she had originally declined because of the confidence she gained in his course. The gig involved writing a jingle, something she had never done before. Though Costa’s class didn’t teach her how to write jingles, Jamie’s exposure to AI tools like Suno AI in the classroom provided the confidence and context necessary to accept new creative and professional challenges.

Jorge collaborated with one of the authors in exploring new ways of learning with AI at the table. Dave Toole was challenged to write a song with the help of AI, not to replace his creativity, but to expand his capabilities. He is a guitar player who had not written lyrics, melody and complete songs. He trained ChatGPT with the meaning of the song to help write the lyrics to his song. This gave him a jump start to build his confidence in the lyrics and provided more time to explore writing melodies. This new skill has now been developed in writing lyrics with no AI involved. Dave used AI on the Apple recording platform to bring out what became called “Dave and the Sleepless Knights.” A prototype of the song was mixed and live musicians were brought in to replace the digital assisted instruments. New ways of learning surfaced in writing the lyrics (think of messaging in business), in prototyping the song (think of product validation in business), and the orchestration of human collaboration after taking advantage of AI for new ways of learning.

In the business world, the experimentation enabled by AI can be transformative. Jim Hornthal, CEO of GLIDR.ai is reshaping how innovation happens across industries, ranging from life sciences, defense, and energy to education, government, nonprofits, and venture capital, with GLIDR.ai - an AI-native platform that guides innovation teams from ideas to insight and impact.

By integrating machine learning, real-time market intelligence, and structured experimentation, GLIDR has been used to empower entrepreneurs, researchers, nonprofit leaders, and corporate innovators to design, test, and refine business models with speed and rigor. Grounded in the Lean LaunchPad methodology that steps start-ups through a 9 step process of conceiving of a business, the customer, market, validation and commercialization, this process was originally developed by Steve Blank and since adopted by the National Science Foundation to launch and scale their Innovation accelerator, the I-Corps program. GLIDR has now enabled over 25,000 teams globally to navigate uncertainty and de-risk commercialization and go-to-market investments.

With AI agents acting as embedded thought partners, users gain actionable guidance to validate market and product assumptions, uncover opportunity spaces to enter the market, design experiments to validate, and align solutions with stakeholder needs, delivering a comprehensive view of a venture’s desirability, feasibility, and viability. GLIDR is redefining innovation as a discipline: faster, smarter, and radically more scalable.

The Venture team at NfX, a venture firm, recently shared a model for a 3-person unicorn, made up of a founder, who is an aggressive visionary, a numbers person and a communications leader. They adopt AI-copilots to accelerate their business systems. The AI tools accelerate their learning to create new business models and frameworks that amplify the value of products and services. Companies are already adopting agents to work while their employees sleep and they are raising the level of skills that teams are bringing to their start-ups. They report portfolio companies increasing output by 100% in software and game development.

For more examples ofAI tools being used to enhance creativity—or to add examples of your own—click here to access our case study repository.

What barriers and obstacles will we need to overcome?

While there is a substantial opportunity to create more impact through these new ways of learning, we will need to address some significant barriers and obstacles that stand in our way. These barriers and obstacles are both in our environment and within ourselves.

The biggest barrier is the culture that prevails in our educational systems and large organizations around the world and that remains focused on traditional forms of learning in classes and training sessions. This culture can be quite skeptical, if not hostile, to these new ways of learning.

There’s an institutional model that governs all large institutions around the world – companies, schools, NGOs, and governments. This institutional model focuses on “scalable efficiency” – the way we organize and operate our institutions is driven by a goal of doing our current tasks faster and cheaper.

This institutional model defines work as tightly specified, highly standardized tasks that are performed reliably and efficiently. To define and execute these tasks, scalable efficiency mandates a hierarchical organization with roles and responsibilities clearly defined at all levels.

To unleash the full potential of AI, we will need to shift to a fundamentally different institutional model – scalable learning, with more of an emphasis on learning by doing together. This institutional model seeks to draw out the passion of the explorer and cultivate a powerful set of human capabilities that are within all of us – curiosity, collaboration, imagination, creativity, reflection, and play.

In this new institutional model, work will no longer be defined as tightly specified, routine tasks – those will be done by machines. Humans will focus on identifying and addressing opportunities to create far more value for stakeholders. To do this, they will come together in a cellular structure of small groups – impact groups – that bring participants together who share a passion for increasing impact in specific domains, solving problems in new ways. Remote and physical teams offer the ability to build a bond around emerging collective learning.

AI will help these impact groups to create far more value by providing participants with tools to cultivate their human capabilities. The challenge is that the current scalable efficiency model is deeply hostile to the procedures and practices required for scalable learning, and the immune system and antibodies of these large institutions will quickly mobilize to crush any effort to shift to the scalable learning model. The challenge will be to drive change against great resistance.

Another challenge we face is the possibility of AI becoming a replacement for our thinking, rather than an enhancement. Data has shown that while AI can bring the lower level of skills up to the level of an expert more rapidly, the norm may miss innovative ideas that occur at the edge. It requires an intentional approach to using AI to ensure we are not losing the creative spark that happens between humans. MIT has published a study that AI can make us dumber when you use it wrong, entitled “Your Brain on ChatGPT, When AI Thinks For You.” This calls out for us to distinguish AI’s potential to expand our imagination through learning and to be wary of its ability to regress or inhibit our creative potential and exploration on the edge.

There’s another set of barriers and obstacles. Over several centuries, we have seen our societies become more and more centralized. Highly centralized societies tend to diminish the potential for grassroots innovation and initiatives. People tend to become more passive.

We have also seen a continuing erosion of social venues that bring people together so they can form deeper and broader relationships. We are becoming more isolated and our trust in those who govern us is eroding.

To help cultivate a sense of agency among everyone, we need to shift towards more decentralized forms of social organization. We also need to cultivate social venues that will help people build deeper and broader relationships with each other so that they can come together and scale their initiatives as they seek to learn through action. But, once again, we need to recognize that our centralized societies are going to resist decentralization. This will not be an easy transition.

On yet another level, fear is also a significant emotional barrier that more and more people are experiencing as we realize how rapidly the world is changing. The accelerating pace of technological development makes it challenging to master the latest AI tools that are constantly evolving. Our fear grows as we see that we lack any experience in pursuing the new forms of learning that are enabled by these rapidly evolving technologies. There is also a great deal of fear of replacement by AI. We need to help people overcome this fear by focusing on the opportunities that AI creates for all of us to pursue more creative forms of work that will deliver more and more impact to those who matter.

For examples ofAI use cases that underline our need for caution—or to add examples of your own—click here to access our case study repository.

Addressing the untapped potential

We need to find ways to overcome these barriers and obstacles so that we can unleash the full potential of AI as a catalyst for new forms of learning.

Rather than trying to change everything all at once, we need to be selective in our initiatives that help better serve our customers and partners. We need to search for leverage points where we can achieve some significant impact quickly with relatively modest effort. Small moves, smartly made, can set big things in motion.

These small moves help people to overcome their fear because they do not require significant effort and can be done relatively quickly. They also help other people to overcome their fear as they see that small moves taken by others are delivering real impact.

One interesting example of these small moves is an experiment launched by Procter & Gamble last year.

Procter & Gamble recently conducted a unique “l(fā)ive hackathon,” bringing together 776 commercial and R&D employees in a one-day virtual workshop. The goal was to develop new solutions for real business challenges within their own units. Participants were randomly assigned to one of four scenarios: an individual or a team of two, with or without access to Generative AI (Gen AI).

The results were compelling: teams utilizing Gen AI completed their tasks approximately 12% faster than those without AI technology. P&G’s findings suggest that for Consumer Packaged Goods (CPG) companies, peak performance is achieved when Gen AI acts as a support mechanism, not a replacement for human interaction.

Beyond speed, the study highlighted significant benefits in cross-collaboration. AI facilitated the development of more balanced solutions among employees from diverse backgrounds, regardless of their individual expertise. On a socio-emotional level, the integration of AI had a positive impact on employee morale, elevating the overall experience. This indicates that strategically implemented AI can not only boost efficiency but also foster better teamwork and a more positive work environment.

Siemens, Autodesk, and other companies offer AI simulation systems that create “digital twins” of manufacturing processes where small teams from engineering, operations, and business units collaboratively experiment with process improvements in a risk-free environment, such as a factory flow that the team wants to model prior to spending the budget to implement. This approach broke down longstanding organizational silos by creating shared contextual understanding across departments. Teams identified optimization opportunities that had been overlooked for years, resulting in 15% efficiency improvements and novel solutions to persistent manufacturing challenges.

Harvard’s AI Learning Studio, developed by Dan Be Kim and Blerim Jashari, serves as a powerful example of AI’s potential to foster scalable learning environments. Kim and Jashari designed a course meant to shift participants’ relationship with AI from a ‘Zone of Overwhelm’ to one of curiosity and agency by offering structured opportunities for participants to ‘tinker’ with AI and course design.

The researchers reflect on their approach—centered on project-based learning and interdisciplinary collaboration—as a response to the evolving nature of AI, a technology that both challenges traditional education models and unlocks radically new possibilities for how we learn. They noted that “Our course was kind of a proof of concept that we need more diversity in how we group learners; AI is a weird kind of glue that fosters that diversity”. AI provided the catalyst for redefining the learning space itself, while serving as a scaffold to meaningful and creative interdisciplinary collaboration.

In seeking to implement AI as a catalyst for new ways of learning and overcoming the obstacles that stand in our way, we would suggest adopting the following approach built on the principles of small moves, smartly made:

1. Identify the leader of the initiative.Ideally this might be the CEO or at least someone reporting to the CEO. The key is to find someone who sees the potential of new forms of learning and has the passion to explore these new forms of learning in spite of significant resistance.

2. Identify the area to target.Seek to find a specific but small area of the organization that could quickly benefit from learning more quickly and effectively. Seek out an area where at least some of the participants have passion about learning and delivering increasing impact to relevant stakeholders.

3. Assemble the impact groups.Bring together small groups of 3-15 people with diverse skillsets and backgrounds. Make sure that at least some of the people in each group have passion about their work. Provide these groups with the relevant AI tools and give them brief overviews of how they might use the tools to learn more.

4. Ask some inspiring questions.Present these impact groups with some questions regarding potential opportunities or problems that could significantly increase their impact. Encourage them to use their AI tools to come up with new approaches that could significantly increase impact.

5. Reflect on “new ways of learning.”As these impact groups begin to deliver results, have them step back and take time to reflect on how their AI tools helped them to pursue new ways of learning and achieve greater impact. Create a library of practices.

6. Inspire others in the organization.Spread the word throughout the organization regarding what these impact groups accomplished and how they used AI tools to pursue new ways of learning. Make the tools more broadly available and integrate them into formal learning and development programs. Continue to spread the word as other groups begin to achieve impact with these new tools and focus on how they used the tools.

7. Cultivate a new culture.As excitement begins to build about the potential of these AI tools to foster new ways of learning, focus on fostering a new culture that seeks to pursue scalable learning throughout the organization. Draw out the passion of the explorer among all participants and encourage leaders to continue to ask inspiring questions that can motivate participants to pursue their learning.

For more examples ofAI as a tool to scale new ways of learning—or to add examples of your own—click here to access our case study repository.

Bottom line

There’s a significant untapped potential in the rapidly evolving technology of AI. It can provide the tools and approaches to support fundamentally new ways of learning that can create much more value. It is an untapped potential because some significant barriers and obstacles stand in our way of embracing this new form of learning. However, there are approaches that can help us unleash the full potential of AI in fostering new ways of learning. These approaches have already been applied in some environments, and we need to find ways to pursue these approaches in a broader set of environments.

Access the H-Corps Case Study Repository

Thanks for reading H-CORPS! Subscribe for free to receive new posts and support my work.

Subscribe