KU Education Dean Says AI Forces Shift From Ancient Pedagogy
Ai4
Higher education must abandon centuries-old lecture models and adopt personalized AI instruction or risk obsolescence amid rapid technological evolution.
## Background
This video was published on YouTube by Ai4 on August 29, 2025. It features Rick Ginsberg, Dean of the School of Education and Human Sciences at The University of Kansas, discussing the urgent need and inherent challenges for higher education to transform its teaching, learning, and research practices in response to evolving generative AI technology.
## TL;DR
Rick Ginsberg argues that institutional change in higher education is remarkably difficult due to deep-seated cultural norms, pedagogical conservatism, and external financial and regulatory pressures. While most faculty currently use AI primarily for basic efficiency rather than pedagogical innovation, traditional educational structures—such as centuries-old curriculum frameworks and long-term research models—are increasingly mismatched with the rapid pace of technological advance. To navigate this shift, universities must establish proactive disclosure guidelines and evolve beyond passive lecture formats toward personalized learning models and student co-engagement.
## Atmosphere & style
The presentation balances lighthearted humor with a frank, pragmatic assessment of academic inertia, opening with relatable banter about speaking right before lunch. Ginsberg uses vivid metaphors, such as Cold War duck-and-cover drills and weight-loss struggles, to convey the deep psychological reluctance faculty feel when asked to alter long-standing teaching habits. The tone is intellectually grounded yet accessible, offering a realistic look at institutional resistance while maintaining a constructive vision for future educational design.
## Key takeaways
- [00:04](https://www.youtube.com/watch?v=YTvUPvozVMo&t=4s) Institutional change in education is exceptionally difficult due to deeply ingrained habits and rigid structural norms, often referred to as the 'grammar of schooling.'
- [00:04](https://www.youtube.com/watch?v=YTvUPvozVMo&t=4s) While generative AI is advancing rapidly, higher education institutions adapt at a far slower organizational pace.
- [03:42](https://www.youtube.com/watch?v=YTvUPvozVMo&t=222s) Asking university faculty to alter their long-standing teaching and research methods requires a profound psychological shift, comparable to a religious conversion.
- [05:58](https://www.youtube.com/watch?v=YTvUPvozVMo&t=358s) Academic resistance to AI is driven by a natural critical mindset, ethical concerns over algorithmic bias and hallucinations, and fears regarding student academic integrity.
- [05:58](https://www.youtube.com/watch?v=YTvUPvozVMo&t=358s) A fundamental mismatch exists between traditional 3-to-5-year academic research cycles (such as RCTs) and the rapid bi-monthly evolution of generative AI tools.
- [08:41](https://www.youtube.com/watch?v=YTvUPvozVMo&t=521s) Educators risk fallling into 'futurist presentism'—trying to prepare students for an unknown future job market using obsolete tools and legacy educational frameworks.
- [08:41](https://www.youtube.com/watch?v=YTvUPvozVMo&t=521s) Surveys indicate that while a majority of higher education faculty use AI, most employ it only for basic administrative efficiency and web searches rather than pedagogical innovation.
- [13:50](https://www.youtube.com/watch?v=YTvUPvozVMo&t=830s) Universities must proactively establish their own disclosure guidelines and ethical policies for AI rather than waiting for external guardrails.
- [13:50](https://www.youtube.com/watch?v=YTvUPvozVMo&t=830s) True educational transformation requires dismantling the passive, traditional lecture model in favor of interactive, human-centered skill development.
- [15:54](https://www.youtube.com/watch?v=YTvUPvozVMo&t=954s) Generative AI tools enable unprecedented personalized learning, allowing students to utilize AI as a thought partner and giving teachers visibility into real-time student workflows.
- [15:54](https://www.youtube.com/watch?v=YTvUPvozVMo&t=954s) Higher education must stop responding to challenges with 'more of the same' and instead radically rethink faculty roles while co-engaging students in system redesign.
## Detailed walkthrough
### [00:04](https://www.youtube.com/watch?v=YTvUPvozVMo&t=4s) Introduction and the Fundamental Reality of Institutional Change
Rick Ginsberg, Dean of the School of Education and Human Sciences at the University of Kansas and co-founder of the Center for Reimagining Education, opens his address with lighthearted banter about holding the tricky speaking slot right before lunch. Introducing his background in transforming teaching and learning within the K–12 space, he outlines the core thesis of his presentation: institutional change is exceptionally difficult. Referring to his book *Duck and Cover*, he draws a historical parallel to the Cold War duck-and-cover drills of the 1950s and 1960s, where schoolchildren were instructed to hide under wooden desks for protection against nuclear explosions. He uses this as a vivid metaphor for the absurd, deeply ingrained habits in education that persist despite being ineffective.
Ginsberg highlights the stark disconnect between rapid technological advancement and slow organizational adaptation. While generative AI is fundamentally reshaping society at breakneck speed, educational institutions remain bound by rigid structural norms—what scholars refer to as the "grammar of schooling." Referring to a recent *Forbes* article questioning whether college professors remain relevant in the AI age, he outlines five key themes for his talk: the inherent difficulty of personal and cultural change, the literature on resistance, the overwhelming uncertainty facing faculty, the historical trap of "presentism," and empirical data from global and domestic surveys on how educators are actually utilizing AI.
- **Speaker:** Rick Ginsberg, Dean of the School of Education and Human Sciences at the University of Kansas.
- **Key Metaphor:** Cold War "duck and cover" drills as an emblem of absurd, long-standing educational practices.
- **Deloitte Study Quote:** "Most organizations are moving at the speed of organizations, not at the speed of technology."
---
[03:42](https://www.youtube.com/watch?v=YTvUPvozVMo&t=222s) The Psychology of Change and "Religious Conversions" in Academia
To illustrate why adopting new practices is so difficult, Ginsberg compares institutional change to trying to lose weight. Though the core formula for weight loss is straightforward—lower caloric intake and raise metabolism—it requires upending personal habits, cultural norms, and daily impulses. Similarly, asking university faculty to alter how they teach and conduct research requires them to abandon methods they have spent their entire lives perfecting. He points out that while academics are often criticized for being politically liberal, their professional habits are intensely conservative; they continue to teach much like Socrates did thousands of years ago.
```
+-----------------------------------------------------------------+
| BARRIERS TO ADOPTION |
+-----------------------------------------------------------------+
| Cultural Norms ---> Deeply ingrained traditional pedagogy |
| Academic Mindset ---> Trained specifically to analyze/critique |
| External Stress ---> Budget cuts & policy constraints |
+-----------------------------------------------------------------+
```
Ginsberg uses a personal anecdote to emphasize the scale of this psychological shift, likening it to asking his wife—the daughter of a conservative Presbyterian minister—to convert to Catholicism on the spot. Changing foundational pedagogical methods is akin to asking scholars to undergo a religious conversion. Furthermore, this internal reluctance is compounded by external pressures, such as federal and state funding cuts, policy restrictions, and political debates over what educators can say in the classroom.
- **Pedagogical Conservatism:** Faculty often adhere strictly to traditional lecture formats despite holding progressive political views.
- **External Pressures:** Higher education faces mounting state and federal funding cuts alongside increased regulatory scrutiny.
---
[05:58](https://www.youtube.com/watch?v=YTvUPvozVMo&t=358s) Academic Resistance, Ethics, and the AI Research Paradox
Examining the extensive literature on resistance to change, Ginsberg notes that academics are uniquely trained to analyze and critique proposals rather than embrace swift implementation. This critical mindset naturally fosters skepticism and fear toward generative AI. He addresses common objections raised by faculty, including AI hallucinations, algorithmic bias regarding race and gender, and widespread worries about student cheating. Reframe the cheating issue, he quotes gun control advocates ("Guns don't kill, people kill; AI doesn't cheat, people cheat"), arguing that students turn to AI for dishonest shortcuts primarily when assignments are boring, poorly designed, or excessively difficult.
```
+-----------------------------------------------------------------+
| THE RESEARCH TIME MISMATCH |
+-----------------------------------------------------------------+
| Gold-Standard Research (RCTs) ---> Takes 3 to 5 years |
| Generative AI Evolution ---> Updates every 2 months |
+-----------------------------------------------------------------+
```
Beyond classroom dynamics, Ginsberg highlights a fundamental structural crisis in academic research: the mismatch between traditional methodology and rapid technological evolution. The gold standard of educational and scientific research—randomized controlled trials (RCTs) and rigorous experimental designs—typically requires three to five years to execute properly. However, generative AI models undergo major transformations every two months. This reality makes traditional long-term impact studies nearly obsolete before they are even published.
- **Common AI Criticisms:** Hallucination rates, training data bias, climate impact, and lack of source attribution.
- **Research Conflict:** Traditional 3–5 year randomized controlled trials cannot effectively evaluate technologies that iterate every few months.
---
[08:41](https://www.youtube.com/watch?v=YTvUPvozVMo&t=521s) Educational Presentism and the Realities of Current AI Use
Addressing the uncertainty surrounding AI's value, Ginsberg asks whether the technology will truly improve educational outcomes or simply make life more complex. He introduces the historical concept of "presentism"—evaluating past actions using modern standards, such as judging Abraham Lincoln's 19th-century colonization proposals through a 21st-century lens—and warns of a reverse "futurist presentism." Educators are attempting to prepare students for a future job market that does not yet exist using tools and frameworks designed for the present. As an example of systemic stagnation, he points out that high schools have taught the exact same four math subjects (Algebra, Geometry, Trigonometry, Calculus) in the exact same manner for over 120 years.
```
+-----------------------------------------------------------------+
| FACULTY AI ADOPTION (DATA HIGHLIGHTS) |
+-----------------------------------------------------------------+
| Digital Education Council Survey (Global): |
| * 61% of faculty use AI |
| * 88% of users employ it for basic tasks (e.g., search) |
| |
| US Research University Survey (Ongoing): |
| * 57% view AI positively |
| * 70% use AI for teaching support or efficiency |
+-----------------------------------------------------------------+
```
To ground the discussion in real-world practice, Ginsberg shares data from recent faculty surveys regarding actual AI integration:
- **Digital Education Council Survey (Global):** A survey of 1,600 faculty members across 100 universities revealed that while 61% reported using AI, 88% of those users employed it only for basic tasks like web searches or minor editing rather than transformative teaching.
- **US Research University Survey (Ongoing):** Early results from Ginsberg's own study of higher education deans and faculty show that 57% view AI positively, with 70% utilizing it for teaching assistance. However, adoption remains heavily focused on administrative efficiency rather than pedagogical innovation.
- **Systemic Stagnation:** High school mathematics curricula have remained largely unchanged for over 120 years.
- **Current Usage Pattern:** The vast majority of faculty use AI solely as an efficiency tool (editing, drafting, searching) rather than a transformative teaching mechanism.
---
[13:50](https://www.youtube.com/watch?v=YTvUPvozVMo&t=830s) The Path Forward and Redefining the Classroom Experience
In concluding the first portion of his presentation, Ginsberg notes that while AI is slowly filtering into higher education, true transformation remains rare. Institutions across the country are waiting for external guidelines or guardrails that may never arrive, meaning individual campuses and departments must establish their own policies on disclosure and ethical use. He stresses that external disruptions—much like the COVID-19 pandemic—will eventually force institutions to change how they operate, whether they feel prepared or not.
Answering a question from the audience regarding what a genuinely transformed higher education system would look like, Ginsberg explains that operational efficiency is merely the baseline. The true structural shift requires dismantling the traditional lecture model: a single instructor standing at the front of a room transmitting information to a passive audience. Higher education must shift toward interactive, human-centered skill development that AI cannot replicate.
- **Institutional Guidance:** Universities must proactively build clear disclosure guidelines for AI use rather than waiting for top-down regulatory frameworks.
- **Vision for Transformation:** Moving away from passive, didactic lectures toward active learning centered on uniquely human capabilities like innovation, critical questioning, and creativity.
[15:54](https://www.youtube.com/watch?v=YTvUPvozVMo&t=954s) Personalized Learning and Transforming the Educational Model
Generative AI unlocks a level of personalized learning that educators have only ever dreamed of in the past. Tools like ChatGPT's Study Mode are truly remarkable, offering college students their very own dedicated AI assistant to guide their studies. Furthermore, platforms like MagicSchool AI provide educators with visibility into how students interact with artificial intelligence in real time. For instance, a teacher can monitor a student's workflow and intervene with constructive guidance—advising a student not to simply ask the AI for direct answers, but to use the technology as a thought partner to help them figure out *how* to solve the problem themselves.
This shift necessitates a fundamental transformation in how education operates. A major frustration in the current landscape is that despite constant complaints about both K–12 and higher education, the standard institutional response is always to "do more of the same." As the speaker quips, "I'm not a mathematician, but I know $A + B = C$, and $C$ is 'we stink'." Persisting with the same traditional methods will only yield worse outcomes. Moving forward, higher education must radically rethink faculty roles, lean into personalized learning experiences, and directly co-engage students in designing these changes, as students are actively using AI today and will be the ones living with the future system.
- **ChatGPT Study Mode:** Functions as an individualized AI tutor designed specifically to assist college students with their learning process.
- **MagicSchool AI Capabilities:** Allows teachers to track student interactions with AI during class and actively steer them toward critical thinking rather than simple answer-seeking.
- **Core Recommendations:** Reconsider traditional faculty roles, embrace personalized instruction, and actively involve students in co-creating the future of education.
## Notable quotes
> [00:04](https://www.youtube.com/watch?v=YTvUPvozVMo&t=4s) Most organizations are moving at the speed of organizations, not at the speed of technology. — Rick Ginsberg
> [05:58](https://www.youtube.com/watch?v=YTvUPvozVMo&t=358s) Guns don't kill, people kill; AI doesn't cheat, people cheat — Rick Ginsberg
> [15:54](https://www.youtube.com/watch?v=YTvUPvozVMo&t=954s) I'm not a mathematician, but I know $A + B = C$, and $C$ is 'we stink'. — Rick Ginsberg
## Mentioned
Rick Ginsberg, University of Kansas, School of Education and Human Sciences, Center for Reimagining Education, Duck and Cover, Forbes, Deloitte, Digital Education Council, ChatGPT, MagicSchool AI
## Video description
Rick Ginsberg, Dean of the School of Education and Human Sciences, The University of Kansas
Abstracts
AI will dramatically change higher education as we know it. As generative AI and other forms of artificial intelligence continue to evolve, higher education must transform its practices to meet the needs of students and faculty, emerging workforce requirements, while supporting society's growing thirst for new knowledge and understanding. But significant challenges face any transformation, and this talk considers issues related to teaching, learning, and research practice confronting colleges and universities, along with the financial concerns this creates. The talk will examine the inherent resistance to change, coping with ethical issues, inequities, supports currently available to assist change, and the many challenges inherent in transforming teaching, learning, and research approaches. It shares perspectives from around the globe. Attendees will learn about the urgency of demands AI is prompting, and ways they can both support the coming transformation while demanding more of institutions of higher education.
KU教育学院院长称AI倒逼陈旧教学法变革
高等教育必须摒弃沿袭数百年的讲授模式并采用个性化AI教学，否则将在飞速的技术演进中面临被淘汰的风险。
Rick Ginsberg认为，受根深蒂固的文化规范、教学保守主义以及外部财务与监管压力的影响，高等教育的体制性变革格外艰难。尽管目前大多数教师主要将AI用于提升基础效率而非教学创新，但沿袭数百年的课程框架与长期研究模式等传统教育结构，正日益脱节于快速的技术发展步伐。为应对这一转变，高校必须建立主动披露规范，摆脱被动讲授模式，转向个性化学习模式与学生共同参与。
堪萨斯大学教育与人文科学学院院长 Rick Ginsberg 指出，高校变革面临根深蒂固的制度惯性，教师在专业教学习惯上极为保守，改变传统教学法宛如要求其进行“宗教皈依”。 (3:42)
传统学术研究需要 3 到 5 年的随机对照试验（RCT），与生成式 AI 每两个月一次的技术迭代速度严重脱节，导致长周期评估在发表时已失去时效性。 (5:58)
全球调研显示 61% 的高校教师使用 AI，但高达 88% 的人仅将其用于搜索或文书等基础效率任务，并未实现深度的教学法创新。 (8:41)
高校必须主动制定 AI 披露准则与伦理规范，借助 ChatGPT Study Mode 和 MagicSchool AI 等工具推进个性化学习，摒弃被动讲座模式并让学生参与教学重构。 (15:54)
Reading summary
Rick Ginsberg, Dean of the School of Education and Human Sciences at the University of Kansas, argues that higher education is trapped between rapid technological disruption and profound institutional inertia. He contends that adapting to generative AI requires universities to dismantle centuries-old instructional habits in favor of human-centered, personalized learning models.
Cultural and Psychological Barriers to Reform (3:42)
Institutional change in education is hindered by deeply ingrained structural norms that Ginsberg compares to Cold War 'duck and cover' drills—persistent habits maintained despite clear inefficacy. Drawing on a Deloitte finding that organizations move at their own pace rather than the speed of technology, Ginsberg notes that changing academic practices resembles the behavioral difficulty of weight loss. While faculty are often perceived as politically liberal, he asserts their professional pedagogy is profoundly conservative, adhering to passive lecture methods reminiscent of Socrates.
Ginsberg compares asking faculty to alter their core pedagogical and research habits to demanding an immediate religious conversion. This internal professional reluctance is further aggravated by external operational stresses, including federal and state funding cuts, policy restrictions, and heightened legislative scrutiny over classroom speech.
Methodological Lags and Academic Skepticism (5:58)
Because academics are explicitly trained to critique and dissect proposals rather than execute rapid organizational rollouts, proposals to adopt generative AI encounter pervasive institutional skepticism. Faculty frequently point to technological shortcomings such as model hallucinations, algorithmic racial and gender bias, lack of source attribution, environmental costs, and academic dishonesty. Addressing cheating, Ginsberg argues that students primarily resort to unauthorized AI shortcuts when assignments are poorly designed, unengaging, or excessively difficult.
Beyond classroom ethics, Ginsberg identifies a structural contradiction in educational scholarship: the research time mismatch. Rigorous academic validation—such as randomized controlled trials (RCTs)—typically requires three to five years to design, execute, and publish. Because generative AI models iterate and upgrade every two months, conventional gold-standard research becomes obsolete before findings reach peer review.
Adoption Realities and the Presentism Trap (8:41)
Educational institutions remain prone to 'futurist presentism,' attempting to prepare students for an undefined future economy using outdated present-day frameworks. As an illustration of institutional stagnation, Ginsberg highlights that high school mathematics curricula (Algebra, Geometry, Trigonometry, and Calculus) have remained fundamentally unchanged for over 120 years.
Empirical survey data indicates that contemporary faculty adoption of AI remains predominantly administrative rather than transformational. A Digital Education Council survey of 1,600 faculty across 100 global universities found that while 61% used AI, 88% of those users deployed it only for routine tasks like web searches and basic text editing. Similarly, preliminary results from Ginsberg's survey of U.S. research university deans and faculty reveal that 57% view AI favorably and 70% utilize it for teaching assistance, yet implementation remains centered on operational efficiency rather than structural pedagogical change.
Toward Personalized, Active Learning Models (15:54)
To move beyond basic operational efficiencies, Ginsberg argues institutions must abandon the traditional one-to-many lecture model and cultivate capabilities that AI cannot easily replicate, including critical questioning, creativity, and collaborative innovation. Instead of waiting for top-down regulatory frameworks that may never materialize, individual universities and departments must proactively institute their own clear disclosure and ethical guidelines.
Emerging platforms demonstrate how AI can facilitate personalized learning at scale. Tools like ChatGPT's Study Mode provide dedicated student tutoring, while platforms such as MagicSchool AI enable instructors to monitor student interactions in real time. Rather than allowing students to simply extract answers, educators can intervene to ensure AI serves as a thought partner for problem-solving. Ginsberg concludes that institutions must redesign faculty roles and directly co-engage students in rebuilding educational structures rather than persisting with failing legacy practices.
Ginsberg observes that higher education will inevitably face disruptive external pressures—comparable to the COVID-19 pandemic—that force adaptation. Whether institutions successfully evolve depends on their willingness to proactively replace long-standing instructional habits with active, personalized learning before external circumstances mandate it.
## Background
该视频由行业会议频道 Ai4 于 2025 年 8 月 29 日在 YouTube 平台发布。视频内容为主旨演讲，由堪萨斯大学教育与人类科学学院院长 Rick Ginsburg 主讲，围绕生成式 AI 对高等教育带来的剧烈冲击展开，系统分析了大学在教学、科研、组织惯性与伦理上面临的深层转型挑战。
## TL;DR
堪萨斯大学教育与人类科学学院院长 Rick Ginsburg 指出，高等教育受制于根深蒂固的体制惯性，变革阻力堪比让人改变宗教信仰，现行的应对常常流于“课桌防核弹”般的表面文章。尽管师生对 AI 的学术不端、算法偏差及传统科研范式失效心存戒备，但多项调查显示大部分高校教师仍仅将 AI 视作浅层查阅与文本润色工具，远未触及核心教学模式。他强调高校不能陷入“未来主义的现在主义”陷阱或被动等待现成规范，而应破除单向灌输式的讲台授课传统。大学应积极拥抱 AI 带来的个性化学习契机，重塑教师作为思维引导者的角色，并与学生协同共建面向未来的全新教育体系。
## Atmosphere & style
演讲整体呈现出直率犀利、生动幽默且切中要害的叙事风格，发言者依托三十余年的院长经历直言不讳地解构体制沉疴，毫无官僚作派。他精妙地穿插了“冷战防核演习”、“痛苦减肥”以及“劝说保守派妻子改宗天主教”等令人捧腹却无比贴切的比喻，将学院派的清高防御与保守本质描摹得入木三分。整场分享节奏明快、警省意味浓厚，既充斥着对高校“穿新鞋走老路”的清醒针砭，又在终局处激发起推动个性化学习与范式重构的紧迫感与行动力。
## Key takeaways
- [00:04](https://www.youtube.com/watch?v=YTvUPvozVMo&t=4s) 高等教育机构具备根深蒂固的保守与制度惯性（学校运作语法），其演进速度不仅远远落后于指数级迭代的 AI 技术，甚至低于普通商业组织。
- [03:41](https://www.youtube.com/watch?v=YTvUPvozVMo&t=221s) 要求大学教师重塑教学模式在心理和认知层面如同“改变宗教信仰”或“减肥”，颠覆了学者们穷尽一生建立的学术认同感与工作习惯。
- [03:41](https://www.youtube.com/watch?v=YTvUPvozVMo&t=221s) 高校除了面临内部教学模式的顽固阻力，还同时承受来自州政府与联邦层面对科研经费削减与学术话语管控的外部重压。
- [05:58](https://www.youtube.com/watch?v=YTvUPvozVMo&t=358s) 学术界本能的批判与怀疑态度放大了对 AI 幻觉、算法偏差、学术作弊及版权问题的担忧，但作弊的根源往往在于教学设计本身的乏味与失效。
- [05:58](https://www.youtube.com/watch?v=YTvUPvozVMo&t=358s) AI 模型的快速迭代使传统耗时 3 至 5 年的 RCT（随机对照试验）等实证科研范式面临时间尺度上的致命冲突与瘫痪危机。
- [10:08](https://www.youtube.com/watch?v=YTvUPvozVMo&t=608s) 教育界面临“未来主义的现在主义”陷阱，即试图为未知未来培养人才，却仍深陷于过去百年来固化陈旧的学科课程流水线中。
- [10:08](https://www.youtube.com/watch?v=YTvUPvozVMo&t=608s) 全球与全美高校调查显示，尽管超半数教师开始接触 AI，但近 90% 的应用仍局限于浅层的信息检索与文本润色，并未触及深层教学与科研范式变革。
- [13:50](https://www.youtube.com/watch?v=YTvUPvozVMo&t=830s) 大学必须主动建立 AI 使用与信息披露（Disclosure）规范，而不是消极等待尚不存在的外部指导方针。
- [13:50](https://www.youtube.com/watch?v=YTvUPvozVMo&t=830s) 高等教育必须彻底打破单向灌输式的“讲台讲授”旧范式，转向重构人机协作与深层互动的学习模式。
- [15:54](https://www.youtube.com/watch?v=YTvUPvozVMo&t=954s) 借助 ChatGPT 学习模式与 MagicSchool AI 等工具，AI 正让个性化学习落地，推动教师角色从直接灌输答案转变为引导思维路径。
- [15:54](https://www.youtube.com/watch?v=YTvUPvozVMo&t=954s) 教育界必须摒弃“在旧方法上加倍努力”的路径依赖，与学生协同共创（co-engage）重构未来的教育体系。
## Detailed walkthrough
### [00:04](https://www.youtube.com/watch?v=YTvUPvozVMo&t=4s) 变革之难与教育中的“蹲下并掩护”隐喻
堪萨斯大学（University of Kansas）教育与人类科学学院院长、重构教育中心（Center for Reimagining Education）联合创始人 Rick Ginsburg 开门见山地指出，在审视当前高等教育的变革时，核心命题非常简单直接——“改变真他妈的难（Change is really damn hard）”。为了说明教育系统根深蒂固的保守与惯性，他提到了自己与同事合著的《蹲下并掩护》（*Duck and Cover*）一书。该书以 20 世纪五六十年代冷战时期的核防空演习为隐喻：当时教育界设计了一套让孩子们钻到木制课桌底下、双手抱头贴膝的滑稽动作，宣称这能像乌龟壳一样抵御核弹爆炸，而事实上大家都心知肚明，一旦核弹落下便万事皆休。Ginsburg 借此指出，无论在 K12 还是高等教育领域，学校几十甚至上百年来都在重复着许多类似“课桌防核弹”般荒谬且低效的陈旧做法。
鲍勃·迪伦（Bob Dylan）的名曲《时代正在改变》（*The Times They Are a-Changin'*）精准印证了当下的 AI 与生成式人工智能浪潮，然而教育机构的演进步调却严重滞后。德勤（Deloitte）此前针对企业界的研究表明，“大多数组织是以组织的速度前进，而非以技术的速度前进”，而高校的响应速度甚至连普通组织的平均水平都达不到。教育学者提出的“学校运作语法”（Grammar of schooling）如同根深蒂固的底层结构，极难撼动。《福布斯》（*Forbes*）曾刊文探讨“AI 时代大学教授是否依然具有不可替代性”，虽然结论是肯定的，但理由绝非学者们过去自以为的那套教学模式。
```mermaid
flowchart TD
accTitle: 高等教育面对技术变革的速度脱节
subgraph SpeedGap["演进速率脱节"]
A["生成式 AI 技术演进<br/>（指数级光速迭代）"] -->|技术驱动| B["普通组织演进速度<br/>（商业组织响应）"]
B -->|层层滞后| C["高等教育/学校运作语法<br/>（数百年惯性，极度保守）"]
end
subgraph Metaphor["历史隐喻与现状"]
D["冷战演习隐喻：Duck and Cover<br/>形式主义应对根本危机"] -.-> C
end
```
- **演讲者身份**：Rick Ginsburg，堪萨斯大学教育与人类科学学院院长，担任院长职务已达 31 年，重构教育中心联合创始人。
- **引用文献与报告**：图书《Duck and Cover》；德勤（Deloitte）组织速度与技术速度对比报告；《Forbes》特稿《大学教授在 AI 时代是否依然重要》。
---
### [03:41](https://www.youtube.com/watch?v=YTvUPvozVMo&t=221s) 变革阻力、范式颠覆与宗教皈依般的挑战
Ginsburg 将推动大学教师改变教学模式比作“减肥”或是“改变宗教信仰”。减肥的原理人尽皆知——少吃多动、创造热量缺口，但面对各种诱惑时却极度痛苦，因为它挑战了个人的生活习惯与文化规范。大学教授们穷尽一生攻读学位、钻研学术，习惯了原有的教学与思考模式。要求他们采用 AI 重塑教学，无异于让他们全盘颠覆安身立命的职业根基。正如一位学者所言，教育工作者骨子里具有“鲜明的保守倾向”；尽管学界在政治议题上常被批评偏自由派，但在具体工作方式上却极度顽固守旧——“苏格拉底当年怎么教，我今天就怎么教”。
更具冲击力的类比是“宗教皈依”。Ginsburg 提到自己的妻子是一位非常保守的长老会牧师的女儿，如果他走上前去对妻子说：“亲爱的，虽然我很尊重你的信仰，但天主教才是真理，我们去追随教皇吧”，妻子必然会挑起眉毛、手叉着腰把他当成白痴看待。让大学教授彻底改变现有的教学体系，在心理和范式层面就像强迫他们完成一次宗教信仰的皈依一样困难。此外，高校还面临来自州政府与联邦层面的外部挑战，包括科研经费削减、学术话语管制等，使得变革环境更加雪上加霜。
- **行为与心理阻力**：要求学者重塑行为模式违背了他们进入学术界的初衷与长期形成的认同感。
- **外部环境压力**：除了内部教学与研究惯性，大学还承受着州级与联邦层面对科研经费削减和言论边界收紧的外部重压。
---
### [05:58](https://www.youtube.com/watch?v=YTvUPvozVMo&t=358s) 学院派批判本能、AI 伦理争议与科研范式危机
在组织变革层面，学者接受的核心专业训练就是“分析与批判”，这种特质让他们天然地站在怀疑和挑剔的立场审视任何新技术。面对以光速迭代的 AI，学术界充斥着恐惧与愤世嫉俗的情绪，并列举了大量负面痛点：幻觉（Hallucination）、算法偏差（Bias）、学术作弊、环境与能耗负担、版权与缺乏来源引用（Attribution）等。针对这些痛点，Ginsburg 进行了犀利的反驳与剖析：
- **关于幻觉与偏差**：媒体与政客每天都在产生幻觉；而 AI 输出的偏见本质上源于训练数据本身的性别、种族等偏见。
- **关于学术作弊**：正如拥枪派常说的“枪不杀人，人杀人”，AI 本身不作弊，是人在作弊。学生之所以作弊，是因为教学内容枯燥乏味、晦涩难懂或者教学方式失败。教师完全可以“用 AI 来对抗 AI 作弊”，设计具备抗 AI 能力的作业或利用验证链提示词（Chain of verification prompting）来追溯信息来源。
- **科研范式的致命冲击**：在学术研究中，随机对照试验（RCT）及准实验设计被视为黄金标准，通常需要耗时 3 至 5 年完成。然而 AI 模型每两个月就发生一次颠覆性迭代，当“实验干预变量（Treatment）”以光速变化时，基于传统周期的实证科研范式几乎陷入瘫痪。
```mermaid
flowchart TD
accTitle: AI 引发的学术界疑虑与范式危机
A["AI 技术冲击"] --> B["教学伦理层面的疑虑"]
A --> C["科研范式层面的危机"]
B --> B1["学生作弊与学术诚信问题"]
B --> B2["内容幻觉与算法固有偏见"]
B --> B3["版权及文献来源引用缺失"]
C --> C1["RCT 等黄金标准需耗时 3-5 年"]
C --> C2["AI 技术底层每 2 个月大幅迭代"]
C1 -.->|时间尺度彻底冲突| C2
```
---
### [10:08](https://www.youtube.com/watch?v=YTvUPvozVMo&t=608s) 未来的“现在主义”陷阱与高校 AI 应用现状调查
在探讨教育应该传授什么技能时，Ginsburg 引入了历史学概念“现在主义”（Presentism）——即用当下的现代认知去苛评和解释历史（例如因林肯曾提出将黑奴迁移安置而用当今标准指责其为种族主义者）。他警告说，教育界面临着一种**“未来主义的现在主义”（Futurist presentism）**陷阱：大家天天都在说“我们正在为尚未出现的工作培养学生”，但我们在设计课程时，却只能受限于当下的已知认知，根本无法准确预知未来的技术形态。诸如高中数学延续 120 年不变的“代数-几何-三角函数-微积分”流水线，大多数人步入社会后一生都未曾使用。面对 AI，大学更应聚焦于培养人类独有的特质，如创新思维、创造力、精准提问（Prompting）等。
为了解全球高等教育界对 AI 的真实应用情况，Ginsburg 对比了国际与全美调查数据：
- **数字教育委员会（Digital Education Council，总部位于新加坡）全球调查**：覆盖全球约 100 所大学的 1600 名跨学科教师。结果显示，虽然有 61% 的教师声称使用 AI，但其中 88% 仅用于类似 Google 搜索的浅层信息检索或简单改写，停留在日常提效层面，几乎没有任何颠覆性的教学与科研变革。
- **全美研究型大学教师调查（Ginsburg 团队开展）**：在受访者中，51% 对 AI 持积极看法，但其中三分之一认为实际落地面临巨大挑战；70% 在教学中有过 AI 使用经历，但绝大多数局限于改写润色和信息查询。
| 调查维度 | 新加坡数字教育委员会（DEC）全球调查 | Ginsburg 团队全美研究型大学调查 |
|:--- |:--- |:--- |
| **样本规模与范围** | 全球约 100 所高校，1600 名跨学科教师 | 全美顶尖研究型大学教师（预计回收 500-1000 份） |
| **使用率与态度** | 61% 教师正在使用 AI | 51% 态度积极，但其中 1/3 认为落地推行极难 |
| **核心应用场景** | 88% 的使用者仅用于日常检索与浅层操作 | 70% 用于辅助教学，主要局限于润色与查资料 |
| **变革深度** | 几乎完全停留在效率提升，未触及深层模式转变 | 极少出现教学范式层面的实质性重构 |
---
### [13:50](https://www.youtube.com/watch?v=YTvUPvozVMo&t=830s) 破除“讲台讲授”旧范式与被迫变革的临界点
高等教育引入 AI 的步伐极其迟缓，大部分教职员工陷入迷茫，苦苦等待尚不存在的指导方针与安全护栏（Guard rails），而这些规范实际上需要高校自身主动建立。Ginsburg 强调，高校必须尽快确立关于技术使用与**信息披露（Disclosure）**的规则。AI 目前尚未在绝大多数高等教育机构中催生出颠覆性改变，但正如 COVID-19 疫情或自然海啸一样，外部技术奇点将带来倒逼机制，最终迫使整个体系做出改变。
在问答环节回答“如果闭上眼睛构想高等教育利用 AI 实现真正变革的愿景，会看到什么”时，Ginsburg 明确指出：在行政运营层面，AI 可以大幅提升事务效率；但在核心教学领域，大学必须彻底打破“教师站在教室/屏幕前方进行单向灌输（Didactic relationship），学生在下方被动接收”的陈旧教学模式。教育必须从单向的知识传递，转向重构人机协作与深层互动的全新学习范式。
### [15:54](https://www.youtube.com/watch?v=YTvUPvozVMo&t=954s) AI 驱动的个性化学习与教育范式的根本变革
AI 为教育领域带来了前所未有的突破，使过去人们只能梦想的“个性化学习”（Personalized Learning）成为现实。像 ChatGPT 推出的学习模式（Study Mode）表现极其出色，其设计初衷就是为了服务大学生，让他们拥有属于自己的专属 AI 助手。以 MagicSchool AI 等平台为例，教师在课堂上能够实时掌握并监督所有学生如何与 AI 互动。这种透明度让教师的角色从单纯灌输答案转变为引导思维——比如教师可以及时提醒学生：“不要直接去要答案，而是让它帮助你思考如何得出答案。”
面对外界长期以来对 K-12 基础教育及高等教育的种种抱怨与批评，教育界最令人担忧的惯性反应往往是“按老一套加倍去做”（let's do more of the same）。演讲者幽默而犀利地打了个比方：“我虽然不是数学家，但我知道如果 $A + B = C$，而 $C$ 代表‘我们搞得一团糟’，那么做更多 $A + B$ 只会让我们搞得更糟。”因此，教育界必须彻底重新审视教师的角色定位，大力推进个性化学习，并让学生深度共同参与（co-engage）到这些新型教育模式的构建之中——因为学生才是当下的核心使用者，也是未来必须面对和适应这套新系统的人。
- **个性化工具实践**：ChatGPT 的学习模式与 MagicSchool AI 等工具赋予了学生专属智能伴学支持，同时让教师能够清晰洞察学生的 AI 使用过程。
- **引导式教学思维**：教学核心应从“获取现成答案”转变为“借助 AI 启发思考路径与解题过程”。
- **打破路径依赖**：坚决摒弃“用旧方法加大力度”的应对逻辑，必须从根本上重塑教与学的模式。
- **重构教育三要素**：重新定义教师角色、全面落地个性化学习，并与学生协同共创未来的教育体系。
## Notable quotes
> [00:04](https://www.youtube.com/watch?v=YTvUPvozVMo&t=4s) 改变真他妈的难（Change is really damn hard） — Rick Ginsberg
> [00:04](https://www.youtube.com/watch?v=YTvUPvozVMo&t=4s) 大多数组织是以组织的速度前进，而非以技术的速度前进 — Rick Ginsberg
> [03:41](https://www.youtube.com/watch?v=YTvUPvozVMo&t=221s) 苏格拉底当年怎么教，我今天就怎么教 — Rick Ginsberg
> [03:41](https://www.youtube.com/watch?v=YTvUPvozVMo&t=221s) 亲爱的，虽然我很尊重你的信仰，但天主教才是真理，我们去追随教皇吧 — Rick Ginsberg
> [05:58](https://www.youtube.com/watch?v=YTvUPvozVMo&t=358s) 枪不杀人，人杀人 — Rick Ginsberg
> [10:08](https://www.youtube.com/watch?v=YTvUPvozVMo&t=608s) 我们正在为尚未出现的工作培养学生 — Rick Ginsberg
> [15:54](https://www.youtube.com/watch?v=YTvUPvozVMo&t=954s) 不要直接去要答案，而是让它帮助你思考如何得出答案。 — Rick Ginsberg
> [15:54](https://www.youtube.com/watch?v=YTvUPvozVMo&t=954s) 按老一套加倍去做（let's do more of the same） — Rick Ginsberg
> [15:54](https://www.youtube.com/watch?v=YTvUPvozVMo&t=954s) 我虽然不是数学家，但我知道如果 A + B = C，而 C 代表‘我们搞得一团糟’，那么做更多 A + B 只会让我们搞得更糟。 — Rick Ginsberg
## Mentioned
Rick Ginsberg, 堪萨斯大学（University of Kansas）, 教育与人类科学学院（School of Education and Human Sciences）, 重构教育中心（Center for Reimagining Education）, 《Duck and Cover》, 鲍勃·迪伦（Bob Dylan）, 德勤（Deloitte）, 《福布斯》（Forbes）, 数字教育委员会（Digital Education Council）, ChatGPT, MagicSchool AI
## Video description
Rick Ginsberg, Dean of the School of Education and Human Sciences, The University of Kansas
Abstracts
AI will dramatically change higher education as we know it. As generative AI and other forms of artificial intelligence continue to evolve, higher education must transform its practices to meet the needs of students and faculty, emerging workforce requirements, while supporting society's growing thirst for new knowledge and understanding. But significant challenges face any transformation, and this talk considers issues related to teaching, learning, and research practice confronting colleges and universities, along with the financial concerns this creates. The talk will examine the inherent resistance to change, coping with ethical issues, inequities, supports currently available to assist change, and the many challenges inherent in transforming teaching, learning, and research approaches. It shares perspectives from around the globe. Attendees will learn about the urgency of demands AI is prompting, and ways they can both support the coming transformation while demanding more of institutions of higher education.
[Source](https://www.youtube.com/watch?v=YTvUPvozVMo)
Source quotes
[0:04] Rick Ginsberg: 改变真他妈的难（Change is really damn hard）
[0:04] Rick Ginsberg: 大多数组织是以组织的速度前进，而非以技术的速度前进
[3:41] Rick Ginsberg: 苏格拉底当年怎么教，我今天就怎么教
[3:41] Rick Ginsberg: 亲爱的，虽然我很尊重你的信仰，但天主教才是真理，我们去追随教皇吧
[5:58] Rick Ginsberg: 枪不杀人，人杀人
[10:08] Rick Ginsberg: 我们正在为尚未出现的工作培养学生
[15:54] Rick Ginsberg: 不要直接去要答案，而是让它帮助你思考如何得出答案。
[15:54] Rick Ginsberg: 按老一套加倍去做（let's do more of the same）
[15:54] Rick Ginsberg: 我虽然不是数学家，但我知道如果 A + B = C，而 C 代表‘我们搞得一团糟’，那么做更多 A + B 只会让我们搞得更糟。
Source quotes
[0:04] Rick Ginsberg: Most organizations are moving at the speed of organizations, not at the speed of technology.
[5:58] Rick Ginsberg: Guns don't kill, people kill; AI doesn't cheat, people cheat
[15:54] Rick Ginsberg: I'm not a mathematician, but I know $A + B = C$, and $C$ is 'we stink'.
Original transcript
[0:04] Well, thank you all for being here. I
[0:06] appreciate it. You always want to be the
[0:08] speaker before lunch or before the end
[0:10] of the day, and here I am. Try to
[0:12] entertain you for a few minutes, and
[0:14] then we'll all have a chance to maybe
[0:15] talk and and go to lunch. Uh my name is
[0:17] Rick Ginsburg, the dean of school of
[0:19] education and human sciences at the
[0:21] University of Kansas. I'm also a
[0:24] co-founder of a center called the Center
[0:26] for Reimagining Education. have some of
[0:28] our great people that are here and we're
[0:30] working in that center in the K12 space
[0:32] to try to transform teaching and
[0:34] learning and it's hard and so the uh
[0:37] message that I have for you today um in
[0:40] examining what's happening in higher
[0:42] education is a simple one and that that
[0:44] message is that change is really damn
[0:45] hard and that's the beginning the middle
[0:48] and the end and I'm going to talk a
[0:49] little bit about why that is um our
[0:52] center and my work uh recently came from
[0:54] a book I wrote with a colleague called
[0:56] duck and cover not that many gray hairs
[0:58] in the audience. But many of you may
[1:00] remember uh back in the 50s and 60s we
[1:02] were worried about the Russians dropping
[1:04] bombs on us. So the the brilliant people
[1:06] in the education space came up with this
[1:08] plan to uh have the kids duck underneath
[1:10] their wooden desks and put their hands
[1:12] over their heads and their knees up to
[1:13] their chin because you know if a bomb
[1:15] went off you didn't have a shell. They
[1:17] told us like turtles do so you needed a
[1:20] wooden desk. And we all seen the
[1:22] mushroom clouds from Hiroshima. knew
[1:24] that if a bomb went off anywhere near
[1:26] us, it was time to cancel Christmas and
[1:28] we were done. And we used this notion of
[1:30] duck and cover as a metaphor for how in
[1:34] the education space, I would say K12 and
[1:36] higher education, although this book was
[1:38] K12, we didn't do higher education
[1:40] because we still work there. That will
[1:41] be when we retire. There's a lot of dumb
[1:43] stuff that we do and we've been doing it
[1:45] an awfully long time, hundreds of years
[1:47] sometimes. Um, a couple of quotes that I
[1:49] think capture what I'm talking about.
[1:52] you know the great song Bob Dylan the
[1:53] times they are a changing if anything in
[1:55] the world captures that it's what's
[1:58] happening in the AI and generative AI
[1:59] space I think that's why we're all here
[2:02] right there was a great study that came
[2:04] out from deote just last year focusing
[2:06] not on education but on businesses and I
[2:08] love this quote most organizations are
[2:11] moving at the speed of organizations not
[2:13] at the speed of technology and there's
[2:15] very few institutions of higher
[2:17] education that move at the speed of
[2:19] organizations so it's a real challenge
[2:21] for for us. And then finally, there's a
[2:23] great line that uh some scholars came up
[2:25] with years ago called the grammar of
[2:26] schooling. It's the way we do things in
[2:29] K12 and higher education. It's really
[2:31] hard to change grammar. So, I'm going to
[2:33] focus on uh on five things. There was a
[2:37] great piece in Forbes, you might have
[2:38] seen it about two weeks ago, and it
[2:40] said, "Are college professors still
[2:42] relevant in the age of AI?" That was the
[2:44] title of it. Um at the end they said yes
[2:48] but not for reasons that you may think
[2:50] about and maybe we can talk about that
[2:52] sometime. Five things I'm going to talk
[2:54] about. I'm going to talk about that
[2:55] change is hard. I want to talk about the
[2:58] whole concept of resistance to change
[3:00] and there's a literature on that
[3:01] encompassing all fields that all of us
[3:03] might work in. Um I want to talk about
[3:06] the uncertainty involved here and how
[3:08] that affects so many people. And the big
[3:10] question we get in academia is yeah but
[3:12] is it going to make things better? Will
[3:14] it work? Uh my argument is always yeah
[3:16] but it can make it seem worse. So why
[3:18] not try something new? Um the
[3:21] uncertainty in involved I mentioned a
[3:23] concept I want to share with you um from
[3:26] our historian friends called presantism
[3:28] and I have some data from a from a a
[3:31] worldwide survey on uh faculty member
[3:33] use and uh the beginnings of data from
[3:35] an American uh university survey on
[3:38] faculty use and that'll capture what
[3:41] we're getting at. So change is hard. How
[3:43] many here have ever tried to lose
[3:44] weight? It sucks, doesn't it? It's
[3:47] really hard. You got to change the way
[3:49] you think, the way you eat, the way you
[3:51] operate. Change is hard. Um, why is it
[3:54] so hard? Well, we're challenging our
[3:57] personal and cultural norms. I mean, I
[3:59] went into the speakers thing and had all
[4:00] those little cute cakes in there and
[4:02] everything. We all wanted to fill up our
[4:04] plates. I don't eat that stuff anymore.
[4:06] But that's hard, isn't it? It's hard.
[4:09] It's very easy to lose weight. eat eat,
[4:11] you know, lower your caloric intake and
[4:14] increase your metabolism. That's that's
[4:16] all it takes. And it's really hard. So
[4:19] for faculty who spent their whole life
[4:21] studying and learning, etc., etc., etc.,
[4:24] we're asking them to to to change the
[4:26] way they think, we're asking them to
[4:28] change the way they behave for the very
[4:30] reasons they went into higher education
[4:31] in the first place. It's hard. Um u one
[4:36] writer uh described it this way, and I
[4:38] and I think he's right. He he says that
[4:41] uh educators have distinct conservative
[4:44] proclivities. And I love when people
[4:46] critique academics for being too
[4:48] liberal, which we probably are, but not
[4:50] to accord not about our work. We're not
[4:53] changing. Socrates did it. I'm doing it
[4:56] right. Um and higher education obviously
[4:59] has other challenges. Now, I don't want
[5:01] to go into those, but uh they're making
[5:03] my life miserable. And any of you in
[5:05] higher education know what I'm talking
[5:06] about. Whether it be at the state level
[5:08] or at the federal level in terms of
[5:10] cutting our grants and and and what we
[5:11] can say and what we can't say, etc. So,
[5:14] oh, let me go back. Um, read back. Yeah.
[5:18] I like to think of it like a religious
[5:20] conversion. My wife is the daughter of a
[5:22] very conservative Presbyterian minister.
[5:25] So, imagine me going up to her and
[5:27] saying, "Hey, honey. Look, I, you know,
[5:29] I really admire your faith, etc., etc.,
[5:31] But the Catholics have it right. You
[5:34] should don reference to the Pope where
[5:36] her eyebrow would go up like this and
[5:38] her hand would go on her hip and she'd
[5:40] look at me like she always looks at me
[5:41] about two or three times a day like I'm
[5:43] an idiot. And that's what we're asking
[5:45] faculty to do. Change their religion.
[5:49] Think about what we're asking in every
[5:51] field. What we're asking people to do,
[5:53] we're asking them to make a religious
[5:54] conversion. And that's not easy.
[5:58] Resistance to change. There's a huge
[6:00] literature on this all over the all over
[6:02] the uh uh I I think in in all fields in
[6:05] the business world and etc. has some
[6:08] students in Taiwan that run AI
[6:10] upskilling companies and this is what
[6:11] they're writing about um about this. But
[6:15] think about academics. What are we
[6:17] trained to do? To analyze and critique.
[6:20] That's what we're trained to do. That's
[6:22] what we do well. That is our work. Let's
[6:26] change. Let's let's reform. Let's look
[6:28] at something differently. Changing that
[6:30] is hard. Um there's lots of fear, lots
[6:33] of cynicism, and I think AI makes this
[6:36] uh u more challenging than anything ever
[6:38] in the past because as you all know,
[6:39] it's all changing at warp speed. I'm
[6:42] going to talk a little later. I It's
[6:43] down here. I'll hold off. Um lots of
[6:45] point of of negativity. Well, there's
[6:46] hallucinations.
[6:48] I mean, I watch Fox News and MSNBC and
[6:51] you know, they they hallucinate every
[6:52] damn day.
[6:55] um biases in the output. You know, if
[6:57] the data that that that the uh
[6:59] generative AI uh bots are using have
[7:02] biases in it, then the output have
[7:04] biases in it as well. And I think a lot
[7:07] of that has been uh in terms of gender
[7:09] and race and things of that. You you
[7:10] feel it here in this field uh for
[7:12] certain. Um uh ethical concerns uh
[7:16] cheating, that's all we hear about. Kids
[7:18] are going to cheat. I'm going to quote
[7:20] my friends who are gun advocates. Guns
[7:22] don't kill, people kill. AI doesn't
[7:25] cheat. People cheat. Why do kids cheat?
[7:28] Because what we're teaching them is
[7:29] boring or they don't get it or it's too
[7:32] hard and we've got to get better at what
[7:34] we're doing. That's why kids are
[7:36] cheating. And it, truth be told, AI does
[7:38] make it easier to cheat. There's lots of
[7:40] bots that offer AI resistant
[7:42] assignments. You should go use AI to fix
[7:44] AI. And that would be a great way, I
[7:46] think, to go about doing it. Um, of
[7:48] certainly the climate impact is real. uh
[7:50] we can't solve that in this room, but
[7:52] that's that's a certainly a fair
[7:54] criticism. Um lots of academics are
[7:56] upset about the lack of attribution to
[7:58] sources that happens with the use of of
[8:00] of generative AI. That's true. But you
[8:03] know what? If you get something, you
[8:04] want the sources, ask it. It'll give it
[8:06] to you. And there's, you know, chain of
[8:08] verification, prompting, and things like
[8:10] that you can utilize to to find things
[8:12] out as well. The last thing is I think
[8:14] the whole research world is going to
[8:15] have to change. The gold standard of
[8:17] research is randomized control trials.
[8:20] they you know conservatively take three
[8:21] to five years to do well. AI changes
[8:23] every two months. How can we do the gold
[8:27] standard of research whether
[8:29] experimental quasi experimental design
[8:31] or randomized control trials at a time
[8:34] when the treatment changes like that we
[8:37] can't do it. So lots of challenges for
[8:39] us in higher education lots of
[8:41] uncertainty. And I get this all the time
[8:43] and it's a fair question but is it going
[8:44] to make what we do better? And I don't
[8:46] know. I I I I I kind of think if you do
[8:49] it well, yes, and if you don't do it
[8:50] well, no. But the reality is that's a
[8:53] fair question. It's true about research.
[8:55] It's true about a lot of things. It can
[8:56] make your life more efficient. We all
[8:58] know that. But if we're talking about
[9:00] transformative kinds of changes, which
[9:02] is what I think about, then I think that
[9:04] uh it it it gets difficult. Uh it
[9:07] requires for all of our faculty lots of
[9:09] new learning and and uh different kinds
[9:12] of approaches. all of our campuses are
[9:14] going to have to uh not only upskill us
[9:16] but invest money and and create
[9:19] guidelines and guard rails and all of
[9:21] those kinds of things and it's hard. Um
[9:24] and I don't know how many times in the
[9:25] K12 space we hear this. I've got 5 years
[9:27] before I retire. Can I just like do what
[9:29] I've always done and that's what we get
[9:32] a lot in the academy as well. We get
[9:34] lots of questions from constituents and
[9:36] parents and and uh our political leaders
[9:40] and students. I mean, there are still
[9:41] places that are banning uh iPhones in
[9:43] schools. This is more powerful than all
[9:46] the computers that put people on the
[9:47] moon, and we don't want our kids to have
[9:49] that in our hands. Are we out of our
[9:50] minds? We can't figure out how to use
[9:52] these in ways that they don't become
[9:54] distractions. I mean, this is the kind
[9:56] of stuff that we face at all levels. Uh
[9:58] I I think um and of course, there's a
[10:01] lot of costs that will be associated
[10:03] with this in higher education, and
[10:05] that's always an issue for us. Um and I
[10:08] don't know if there are any historians
[10:09] here. Historians have a term in in their
[10:12] bag of tricks called presantism. So
[10:15] what's presantism? That's when you try
[10:17] to interpret the present,
[10:21] the past, I'm sorry, in terms of our
[10:23] understandings today. The best example
[10:25] of that is the people who call Lincoln a
[10:27] racist. Because if you know what Lincoln
[10:29] thought about doing, he thought about
[10:31] maybe moving the slaves to a state or
[10:33] maybe then sending it back to Africa,
[10:35] which was wildly liberal in the 1860s
[10:38] and late 1850s. But people have looked
[10:40] at that today and say, "Aha, he wasn't
[10:43] the the great um freer of the slaves. He
[10:46] was a racist." That's presantism.
[10:49] They've done the same thing with Dr.
[10:50] Seuss's writing recently and things
[10:52] along those lines. I think a problem
[10:54] that we have in the academy is that as
[10:57] we who are looking to the future are
[10:59] trying to figure out how to use this,
[11:01] we're trying to figure out how to use it
[11:03] knowing what we know now and we don't
[11:05] know what it's going to be in the
[11:07] future. How many times have we all heard
[11:09] we're training kids for jobs that don't
[11:11] exist yet. So what the hell should we
[11:12] train them on? We have no idea. It's a
[11:15] form of futurist presentism that worries
[11:17] me. So we have to think about what are
[11:19] the human kinds of skills that to this
[11:22] point anyway AI can't do that perhaps we
[11:25] should be training people on whether it
[11:27] be being innovative or being creative or
[11:29] figuring how to prop the right way or
[11:30] whatever it might be. That's a real
[11:32] challenge for us. And so the the reality
[11:35] is is that uh we don't know what skills
[11:37] and knowledge should be taught. Um I was
[11:40] just talking to somebody. We're talking
[11:41] about math and we've been teaching the
[11:43] same four subjects in math in high
[11:44] school for about 120 years and probably
[11:47] about the same way as we've taught it
[11:49] the last you know you take your algebra
[11:51] geometry trigonometry if you're smart
[11:52] calculus we still do it and then ask
[11:54] people how much have they used that
[11:56] short of you engineers most people will
[11:57] tell you never in my life not once right
[12:00] so that's what the kind of thing that we
[12:02] have to think about what is it that we
[12:04] can do we humans can provide the human
[12:07] touch um where are we today. Real
[12:11] quickly, there's a great group. If
[12:13] you're in higher education, I I I urge
[12:15] you to look into it. They're out of
[12:16] Singapore. It's called the Digital
[12:18] Education Council. It's about 100
[12:20] universities. And some suppliers, like
[12:22] in this exhibit room here, are trying to
[12:23] help universities bring AI into their
[12:26] space. They did a survey late last fall
[12:29] in December, I think, of about,600
[12:33] faculty members from around the world.
[12:35] And what did they find? uh 61% said
[12:38] they're using it, but 88% of that 61
[12:40] basically said no, they use it for
[12:43] simple things. They could they could
[12:44] have used Google for
[12:47] using it in transformative ways. We're
[12:48] not changing again mostly to support
[12:50] current practice or to uh um be more
[12:55] efficient. Uh we're in the midst of
[12:56] doing a survey of uh faculty members in
[12:59] in in higher education. The the the um
[13:01] digital educational survey was every
[13:03] field. Every field. The survey that I'm
[13:05] doing is just people. I've been a dean
[13:07] 31 years. So, uh um uh I I I know a lot
[13:11] of people in the space. We're doing it
[13:12] from with the research universities
[13:14] around the country. And we we we're
[13:16] still open. So, we expect to get between
[13:18] 500 a,000 responses. 517% see it as
[13:22] positive. Uh but a third see it. A third
[13:26] of that positive see it as a real
[13:27] challenge in order to implement. Uh 70%
[13:30] were using it for teaching, but mostly
[13:32] it's in limited use. They'll look
[13:34] something up. They'll help it rewrite
[13:36] something that they've written. Things
[13:38] like many of us use it for. Um, and it's
[13:41] all mostly about efficiency. I'll I'll
[13:43] close with this and then we can have
[13:45] some I don't know if I'll have time for
[13:47] questions, but certainly I can stick
[13:48] around if everybody wants to chat. Uh,
[13:50] AI is certainly entering higher
[13:52] education. Uh, we just had a prior
[13:54] speaker uh talking about what they're
[13:56] doing at UCLA as an example. Change is
[13:59] coming really, really, really slowly.
[14:01] Everybody's kind of waiting for the
[14:03] guard rails or the guidelines or
[14:04] whatever and they're not really coming.
[14:06] We have to develop them ourselves. Um
[14:09] the universities need to provide a lot
[14:11] more guidance that are faculty because
[14:13] most are lost. We've got a couple of uh
[14:16] folks here that work with faculty on our
[14:17] campus and everybody and we've joined
[14:19] this digital education council where
[14:21] everybody is grappling with the same
[14:23] thing to help guide us in doing this. Um
[14:26] uh the big thing I think in in our space
[14:30] for kids, for faculty, and for uh and
[14:33] for researchers is we got to think about
[14:36] um what guide rails we're going to allow
[14:38] or not allow. And everybody has to get
[14:40] into the world of disclosure. Here's
[14:42] what I used it for. And that's that's
[14:44] going to be a big big jump for us. Um to
[14:47] date, I would say that it's entering,
[14:49] but it's not entering higher education
[14:51] in in transformative ways in very many
[14:54] places. Um I know the work that we're
[14:56] doing in the K12 space there's maybe two
[14:58] or three places around the country we're
[15:00] aware of they're doing things kind of
[15:02] like the way we're doing it. So the
[15:03] question remains how will I AI change
[15:07] higher education practice. I do think
[15:10] that technology can create crisis much
[15:14] like co did much like other things have
[15:16] done like tsunamis do elsewhere you got
[15:18] to change the way you behave that's what
[15:20] the great writers on change talk about
[15:22] it's going to force us to change how
[15:24] that will be I I don't know thank you
[15:26] all and have a good lunch
[15:30] okay so the question was if I could
[15:32] close my eyes and I would look at the
[15:34] way higher education would use it in
[15:36] transformative ways what would I see
[15:38] well on the operations side, it can make
[15:40] things a whole lot more efficient. So
[15:41] that's that's and that's true in any
[15:43] business, whatever. But we've got to get
[15:45] out of this model of a teacher in the
[15:48] front of a classroom, whether online or
[15:49] face to face, and a dactic relationship
[15:52] with passive people working there. What
[15:54] AI does that we've never had before is
[15:57] create the possibility for personalized
[15:59] learning that we've never we've dreamed
[16:00] about in the past. I mean Chachi PT's
[16:03] study mode is remarkable and it was
[16:05] written for college students and they
[16:07] have their own AI assistant. So there if
[16:10] you go through magic school AI for
[16:12] example and you have a class in front of
[16:14] you the teacher can see what all the
[16:15] kids are doing with AI. So Maggie can
[16:17] say to uh to Lisa Lisa don't get the
[16:20] answers have that help you think about
[16:22] how to get the answers. I think we have
[16:24] to transform the way we've always done
[16:26] it. What bothers me is we hear
[16:28] complaints about K12 and higher
[16:29] education all the time and our response
[16:31] to it is always let's do more of the
[16:33] same and that's what we do. Well, I'm
[16:36] not a mathematician but I know but a and
[16:38] b a plus b equals c and c is we stink.
[16:41] Doing more of a plus b means we'll stink
[16:42] more. So, I think we have to reconsider
[16:46] faculty roles, more personalized
[16:49] learning, and co-engage co- uh co-engage
[16:52] students in the development of all this
[16:54] because they're the ones that are using
[16:55] it now, and they're the ones that are
[16:57] going to have to live with whatever we
[16:58] come up with. Thank you.
