"Because these googlers are not teachers and not educators, they tend to want to give a history of something... It’s interesting but people don’t come out of it about what it is, how do I use it?"

This is a manifestation of the "curse of knowledge."

Supporting Evidence:

• Many subject matter experts, while possessing deep knowledge, may lack the pedagogical skills to effectively transfer that knowledge to others, often struggling with structuring content, engaging learners, and assessing understanding. (Training Industry Magazine, "Turning SMEs into Effective Trainers," trainingindustry.com)
• Research indicates that SMEs often focus on "what" to teach (content accuracy) rather than "how" to teach (instructional strategies), which can lead to ineffective learning experiences. (Journal of Workplace Learning, "Subject matter experts as workplace learning facilitators: A literature review," emerald.com/insight)

Professor Matthew Rascoff stated, "Cognitive science in education focuses on K-12. There’s a gap in adult learners." Professor Rob Urstein mentioned, “We know adults learn differently - but that’s all we know".

Supporting Evidence:

• While cognitive neuroscience has significantly advanced our understanding of learning and memory, its application to adult learning principles (andragogy) in corporate or higher education settings is still an emerging field, with more research focused on child and adolescent development. (Frontiers in Psychology, "Cognitive Neuroscience and Adult Learning: A Scoping Review," frontiersin.org)
• There is a recognized need to better integrate findings from cognitive neuroscience with adult learning theories to create more effective and evidence-based instructional practices for adults. (Adult Learning Quarterly, "Bridging the Gap: Cognitive Neuroscience and Adult Learning," adl.sagepub.com)

“I have always been envious of [the learning designer on our team] because he knows why certain things work the way they work... I wish I could combine those two”

The Google School for Leaders team expressed, "we have only 1 learning designer on our team but we wish we could have 3 of her." Instructional designers are often described as the "best kept secret in higher education."

Supporting Evidence:

• The role of instructional designers is often misunderstood and undervalued, despite their critical contribution to creating effective learning experiences, a situation reminiscent of the early days of UX design. (Chief Learning Officer, "The Evolving Role of the Instructional Designer," chieflearningofficer.com)
• Effective instructional design is crucial for learner engagement and knowledge retention, yet organizations often underestimate the specialized skills required, leading to underinvestment in this area. (Association for Talent Development (ATD), "The Value of Instructional Design," td.org)

The founder of UniversityNow observed learners “wanted to know they weren't alone.” A student from SI EDU, GSB confirmed, “It’s really nice to connect with someone in the organization.” MOOCs are seen as not working "because they don’t offer the community, social capital, connection required for skills learning”.

AI practitioners "Usually read blog posts and newsletters," "track Scholar citations," and follow the "firehose of information on social media." Adult learning often involves interacting with others and participating in group activities, leading to increased social engagement.

Supporting Evidence:

• While self-directed learning is promoted for its flexibility, many adult learners express a need for social interaction and community to stay motivated and deepen their understanding, a factor often missing in purely online, self-paced environments. (Computers & Education, "The role of social presence in online learning," sciencedirect.com)
• Research on Communities of Practice (CoPs) highlights the importance of social learning, where individuals learn by sharing experiences and insights within a group, suggesting that purely individual learning can be less effective. (Harvard Business Review, "The Power of Hidden Teams," hbr.org)

Testing recall, a principle of effective learning, is often absent in passive lecture-based corporate training.

Supporting Evidence:

• The Ebbinghaus forgetting curve demonstrates that learners forget a significant amount of information shortly after learning it if there is no attempt to retain it. Many corporate training programs lack spaced repetition and retrieval practice. (Journal of Applied Psychology, "Spacing and retrieval practice: A meta-analysis of the gold standards for learning," apa.org/pubs/journals/apl)
• Traditional "one-and-done" training events are often ineffective because they don't incorporate strategies to mitigate the forgetting curve, such as follow-up activities, microlearning, or performance support tools. (Performance Improvement Quarterly, "Beyond the Event: Aligning Training with the Forgetting Curve," onlinelibrary.wiley.com)

A Learning Designer at Google mentioned Intellum is "from the 90s" and that "HR is moving away from Intellum because they were quadruple charging." Another stated Intellum's UI is "user-friendly while manual, but internal efforts are so much heavier lifts." One facilitator mentioned, "we pay hundreds of thousands to LMS fees... hard to keep it up to date."

Supporting Evidence:

• Many Learning Management Systems (LMS) are criticized for being content repositories rather than dynamic learning environments, often lacking features for true engagement, social learning, or easy content updating. (ELearning Industry, "Is Your LMS a Learning Enabler or a Glorified Filing Cabinet?," elearningindustry.com)
• The administrative overhead and inflexibility of some corporate learning tools can make it difficult for L&D teams to quickly create, update, and deploy relevant content, hindering their ability to keep pace with business needs. (Chief Learning Officer, "The Trouble with Learning Tech," chieflearningofficer.com)

A Training Program Manager at Google noted slide decks "refer to stuff at the beginning of 2024. It’s not fresh, not accurate." "The most intimidating part is how do I write my content, when do I need to break this up." A Learning Designer stated they are "creating an e-learning about... good LOs and good assessments... That’s the hardest part for non learning folks."

Missing components include: scaffolding, relevance, discoverability, assessments, feedback, engagement, tailoring, concision, coaching, practice, data tracking, freshness.

Supporting Evidence:

• Scalability in learning design often leads to a trade-off with personalization and interactivity. Creating truly adaptive and engaging experiences at scale remains a significant challenge. (Educational Technology Research and Development, "The challenge of scaling effective personalized learning," link.springer.com)
• Many digital learning tools struggle to effectively implement formative assessment and provide meaningful, individualized feedback in a timely manner, crucial for effective learning. (Review of Educational Research, "A review of feedback models and their application to digital learning environments," journals.sagepub.com/home/rer)

"This space is changing so rapidly... All the knowledge I had about infrastructure all went out the window." "Everything is unified under LLM paradigm - everything is a text."

US labor market: 8.6M occupational shifts (2019-2022), 50% more than prior 3 years. Automation may impact 30% of hours worked by 2030 (McKinsey).

Supporting Evidence:

• The World Economic Forum's "Future of Jobs Report" consistently highlights the accelerating pace of technological change and its impact on skills, predicting a large percentage of core skills will change for most jobs. (World Economic Forum, "Future of Jobs Report," weforum.org)
• The "half-life of skills" concept suggests that skill value depreciates over time, shortening rapidly in tech fields, necessitating continuous learning. (Deloitte Insights, "The half-life of skills: Reskilling for the future of work," deloitte.com/insights)

One Google researcher mentioned to meet new challenges, "Google hires MDs." For a project, hiring a professor "brought many students over." "Whether it's us vs OpenAI vs Apple vs Google vs Meta. We also have to compete in that market."

At age 26, 59% of engineering/CS grads work in field; by age 50, only 41% do.

Supporting Evidence:

• Many companies prefer to "buy" talent (hire externally) rather than "build" talent (train internally) when facing skills gaps, especially for in-demand tech roles, contributing to intense competition. (SHRM, "Build vs. Buy: Weighing the Talent Options," shrm.org)
• The "war for talent" is persistent, particularly in tech, where companies prioritize attracting experienced professionals over upskilling current workforce, often leading to higher costs and turnover. (McKinsey Quarterly, "Winning the war for talent," mckinsey.com)

Reports suggest NASA lost the ability to go to the moon due to knowledge depreciation and lack of transfer as experienced personnel retired.

Supporting Evidence:

• Knowledge loss due to employee attrition and retirement is a significant risk, leading to reduced productivity, loss of competitive advantage, and costly errors if critical tacit knowledge is not captured and transferred. (Journal of Knowledge Management, "Strategies for managing knowledge loss: a comparative study," emerald.com/insight)
• Failure to invest in knowledge management and internal training can result in the loss of "organizational memory," making it difficult to learn from past experiences or maintain core competencies. (Sloan Management Review, "The High Cost of Lost Knowledge," smr.mit.edu)

Professor Matthew Rascoff stated, “MOOCS don’t work because they don’t offer the community, social capital, connection required for skills learning”.

Google MOOCs on ML are an example of corporate learning at scale.

Supporting Evidence:

• Despite initial hype, MOOCs have notoriously low completion rates, often attributed to learner isolation, lack of engagement, and insufficient support structures that foster community. (Science, "The MOOC Generalization: It's Time for a MOOC Intervention," science.org)
• Research highlights that social presence and interaction are critical for online learning success; the predominantly self-paced, individual nature of many MOOCs fails to adequately address these needs. (The Internet and Higher Education, "Community inquiry in MOOCs: A blended learning approach," sciencedirect.com)

Professor Matthew Rascoff: "I don't think the evidence supports the idea that [personalized/adaptive learning] is the Holy Grail... personalizing learning around learning styles has absolutely no evidence behind it, zero."

Example: Newton, a company that "promised every student a robot tutor in the sky. It became kind of a joke..."

Supporting Evidence:

• Many edTech tools focus on features over pedagogical effectiveness, sometimes replicating outdated methods or overemphasizing simplistic assessments. (Educational Psychologist, The Cambridge Handbook of Multimedia Learning)
• The "learning styles" myth, despite being debunked, continues to influence some edTech designs, highlighting a disconnect between product development and learning science. (Journal of Educational Psychology, "Learning Styles: Concepts and Evidence," apa.org/pubs/journals/edu)

The "factory model" critique of schooling is often an ahistorical rhetorical device. Audrey Watters argues it's rhetoric used by engineers to sell edTech. The Prussian factory school, often cited, actually preceded American factories and aimed at effective, large-scale, tax-funded schooling.

Supporting Evidence:

• Historians argue the "factory model" narrative is an oversimplification and misrepresentation, used as a rhetorical tool for reform or new technologies. (History of Education Quarterly, "Revisiting the 'Factory Model' of Schooling," cambridge.org/core/journals/history-of-education-quarterly)
• The critique often ignores democratic and standardizing impulses that shaped early public education, which aimed for universal access. (American Journal of Education, "The Manufactured Crisis: Myths, Fraud, and the Attack on Public Education," journals.uchicago.edu/toc/aje/current)

Professor Rob Urstein's "primary interest area is adult education & certifications for those who didn't have access to education." Professor Matthew Rascoff: "Cognitive science in education focuses on K-12. There’s a gap in adult learners."

US schools projected $7.8B on K-12 edtech software.

Supporting Evidence:

• While K-12/higher ed edTech markets get attention, adult learners' specific needs in professional development/lifelong learning are less catered to, or addressed with repurposed solutions. (UNESCO Institute for Lifelong Learning, "Embracing a culture of lifelong learning," uil.unesco.org)
• Growing demand for edTech for reskilling/upskilling adult workforce, yet many platforms may not align with adult learning principles or practical needs. (OECD, "Getting Skills Right: Future-Ready Adult Learning Systems," oecd.org/future-of-work/)

A 2024 NBER working paper (w33662) indicates that leadership skills can be assessed by AI agents.

Supporting Evidence:

• AI-powered adaptive learning systems can tailor educational content and pacing to individual student needs, potentially leading to more efficient and effective learning outcomes. (Computers & Education, "A meta-analysis of the effects of adaptive learning on students’ learning outcomes," sciencedirect.com)
• AI can create sophisticated simulations and role-playing scenarios for practicing and assessing complex skills like decision-making and leadership in a safe environment. (Journal of Educational Computing Research, "Using artificial intelligence to assess soft skills in simulated environments," journals.sagepub.com/home/jec)

An AI Powered Orgs Stanford course by Professor Amir Goldberg suggests AI can help with 3 of the 4 organizational learning phases (search, inference, retention, but not action).

Supporting Evidence:

• AI tools can enhance organizational learning by improving knowledge discovery (search), facilitating data analysis (inference), and personalizing learning paths for knowledge retention. (Knowledge Management Research & Practice, "Artificial intelligence in organizational learning: A review and research agenda," tandfonline.com/loi/tkmr20)
• AI-driven knowledge management systems can help organizations capture, codify, and disseminate knowledge, supporting learning and retention phases. (MIS Quarterly, "Artificial Intelligence and the Future of Work: A Research Agenda on Organizational Learning," misq.org)

Professor Matthew Rascoff is "more skeptical... about personalized learning, adaptive learning. That's a fad... no evidence." He argues for "communalized learning, collective learning... that's actually the Holy Grail."

Supporting Evidence:

• While AI tutors/chatbots offer scalable support, research comparing them to human tutors or well-designed collaborative learning often shows limitations in fostering deep conceptual understanding or complex problem-solving. (Review of Educational Research, "The Effectiveness of Intelligent Tutoring Systems: A Meta-Analytic Review," journals.sagepub.com/home/rer)
• Push for extreme personalization in AI learning, sometimes based on flawed concepts like learning styles, may overlook benefits of social and collaborative learning crucial for communication, teamwork, and diverse perspectives. (Educational Researcher, "The Myth of Learning Styles, and the Dangers of an Old Idea," journals.sagepub.com/home/edr)