English
Ask anything

November 1, 2026 · Artificial intelligence

Future of education: what is missing from Anthropic’s S1 initial public offering prospectus

Reda Sadki
Future of education: what is missing from Anthropic’s S1 prospective

Anthropic’s prospectus describes AI in education as tutoring for students and paperwork help for teachers. The research and debate of the past two years point to bigger questions: whether students learn or only perform, what a degree certifies when AI can produce the coursework, how professions will train their next experts, and who will own the systems that schools rent.

Anthropic’s S-1, the prospectus for its public offering, promises “providing individually tailored education” and a student learning mode “designed to strengthen critical thinking”, and reports no learning outcomes for either. In the first article, I examined what the filing says about education, workforce and human capital, and why it says so little. This article sets out what the filing leaves out.

I draw on three bodies of work that I have found most useful in my own practice, supporting a network of more than 80,000 health and humanitarian workers in 137 countries (Sadki, 2026h) and building AI co-workers to support their learning (Sadki, 2025a).

  • The first is the experimental evidence that separates performance from learning.
  • The second is the economist Jason Potts’s account of what universities sell, which explains why AI threatens their business model and not only their exams.
  • The third is George Siemens’s two years of reporting from inside university leadership, which shows how institutions are responding.

I add my own work on how health workers become experts.

Studies separate assisted performance from learning

A learner using AI can produce better work today without being able to do that work tomorrow. The studies below test three different claims: that AI improves performance while it is available, that it builds capability that remains once it is withdrawn, and that this capability becomes lasting human capital at work.

  • Unrestricted AI harmed exam performance. In a field experiment with nearly 1,000 high school mathematics students in Türkiye, Bastani and colleagues (2025) compared an unrestricted GPT-4 tutor, a version with guardrails, and ordinary textbooks and notes. Unrestricted access raised practice scores by about 14 points on a 100-point scale. On the closed-book exam that followed, the same students scored about 5 points below the control group: the tool did the work, and the students learned less. The guardrailed tutor, which withheld direct answers and used teacher-written hints, largely removed that harm but did not beat ordinary practice. The exam was immediate, so the study cannot say what students retained months later.
  • Well-designed tutoring produced immediate gains. StudentBench, a September 2026 preprint, randomly assigned 2,383 adults to AI tutoring, human tutoring or a control condition for one hour of GRE preparation (Northcutt et al., 2026). AI tutoring raised unaided post-test scores by about 6 percentage points over controls watching unrelated videos, and the authors report statistical equivalence with expert human tutoring. The post-test was immediate, the control did no relevant practice, the human arm was small, the sample was filtered after assignment, and the funder supplied the participants through its platform.
  • Novices gained most at work. Among 5,172 customer-support agents, an AI assistant raised issues resolved per hour by 15% on average and by 30% for less experienced and lower-skilled workers (Brynjolfsson et al., 2025). During outages, treated workers still beat their pre-AI baseline, which the authors read as “durable learning.” Their data “do not allow us to observe changes in wages, overall labor demand, or the skill composition of workers hired.” Top performers followed AI suggestions even when suggestion quality fell.
  • AI helped individual scientists and narrowed science. Hao and colleagues (2026) analysed 41.3 million research papers. Scientists who used AI published 3.02 times more papers, received 4.84 times more citations and became research project leaders 1.37 years earlier than those who did not. Across science, AI adoption shrank the range of topics studied by 4.63% and reduced scientists’ engagement with one another by 22%, as AI-assisted work moved toward the areas richest in data. AI teams also had fewer junior scientists, 1.99 on average against 2.89 in other teams. The authors conclude that AI tools “seem to automate established fields rather than explore new ones.” A university that rewards individual output would see its researchers thrive while the questions it asks, and the junior places on its teams, shrink.

In these studies, the effect of AI depends on its design, on what is measured and on when it is measured. None follows learners from an AI-supported intervention through durable capability to employment and earnings.

MIT concluded that AI has emptied its core teaching tools

In August 2026, MIT’s Ad Hoc Committee on AI Use in Teaching, Learning, and Research Training reported that AI can already “produce credible solutions and provide reasonable responses to almost any written assignment in our undergraduate curriculum, including essays, math and science problems, proofs, and coding assignments.” Problem sets, long written assignments, take-home exams and out-of-class projects have long been how MIT students practise and show what they know. “AI saps these tools of their value for both teaching and assessment” (Massachusetts Institute of Technology, 2026).

The report opens: “This report is a call to action.” Charged with proposing an AI use policy, the committee concluded instead that MIT “must tackle a set of deeper questions about the structure, meaning, and value of an MIT education.” Its carefully worded finding, that “most classes taught at MIT will require careful review and, in many cases, substantial adaptation to survive an encounter with AI without serious disruption”, amounts to a review of almost the entire curriculum, and the report says so elsewhere: “Every subject taught at MIT will likely need to be reexamined and potentially revamped.” It goes further still: “the advent of AI demands a broader, holistic reassessment of the nature, scope and purpose of higher education today.” In the committee’s words, “this is not a moment for patches and duct tape.” Its recommendations reach the foundations of the institution:

  • Review, “campus-wide and as soon as possible”, what students need to learn, and ask whether long-standing goals still hold, such as whether most students “need to be able to write complex programs by hand”.
  • Investigate “the nature of grades and grading”. The committee discussed the thought experiment that “if MIT did not have grades, many of the incentives around AI cheating would disappear.”
  • Change governance so departments can alter curricula “without fixed, multi-committee, year-long review processes – or we will be left behind.”
  • Protect undergraduate research, because faculty “may find it advantageous to replace novice researchers with AI agents” and students would lose “opportunities to become participants in research communities.”

The committee also reports that, in less than three years, AI has reduced attendance at office hours and participation in online discussions, and, by anecdote, in-person study groups. Students told it that instructors’ guidance “is often confusing and unclear.” Its survey had 1,632 responses at a 12% response rate, so attitudes may not represent even MIT. One of the world’s best-resourced universities is telling itself that its assessments, its grades and its approval processes all need rethinking.

Potts argues that universities sell verification

Jason Potts explores in The University after AI why the disruption MIT describes reaches beyond the classroom (Potts, 2026). He describes the university as “neither a cloister nor a factory” but a multi-sided platform, a system of systems that coordinates “at least fifteen distinct sides”: undergraduates, international students, research students, employers, governments, alumni, donors, high schools, parents and others. Each side “derives value from the presence and quality of the others”, and cross-subsidies bind them. Teaching revenue pays for research, and research builds the reputation that attracts students.

Earlier technologies lowered the cost of content, but “content was never the product that universities sold.” Universities sell “matching and verification”: they bring people together and vouch to employers that a graduate knows what the degree claims.

Potts argues that AI is the first technology to reach both functions. In “signal collapse”, AI lowers the cost of producing assessable work for able and less able students alike, so the credential stops telling them apart.

The Bastani results and MIT’s account of emptied problem sets show this happening in classrooms.

Disintermediation strips away the “informational sides of the platform, which are also the paying sides”.

The university’s strategic asset becomes its ability to act as “a credible verifier”, which requires independence “of the parties whose output it verifies.” These mechanisms remain hypotheses, and their force will depend on how institutions and employers respond.

Because the sides are bound together, a tool that improves one can damage another.

  • A cheaper AI tutor weakens the teaching revenue that pays for research.
  • Easy AI-written assignments weaken the credential employers rely on.

Potts also describes a “missing junior loop”: AI “removes the junior tasks through which learning occurred”, which weakens firms’ incentive to train. He calls this “a market failure with an obvious institutional remedy” and proposes supervised placements, apprenticeship years and simulation-based training. The shrinking number of junior scientists on AI-assisted teams in the Hao study is the same loop breaking inside research.

Siemens tells universities to build their own AI

George Siemens was amongst the first who articulated a theory of learning for the Digital Age. His now-defunct newsletter, SAIL: Sensemaking AI Learning, recorded two years of meetings with university leaders and builders.

  • On buying. “I’m a broken record on this, but universities need to become AI product builders, not simply buyers of what is being built by startups and frontier labs” (Siemens, 2025e). Most institutions, he wrote, “have primarily signed licenses with OpenAI or Anthropic or Google. Buying technology is hardly vision” (Siemens, 2025c). Without a shared vision, he said in a December 2025 interview, “we default to writing checks for solutions that were not built with us in mind”, ending up “renting systems that do not reflect our values and handing over control, especially of our data” (Salisbury, 2025).
  • On access. At one panel, he recalls, “the LLM representative stated that this approach was overkill and simply giving students access to AI tools would be sufficient.” He separates point solutions, which fix a problem inside the existing system, from systems change: “Every university will adopt point solutions. Only a few with forward thinking and visionary leadership will achieve the systems change” (Siemens, 2025a).
  • On leadership. After visiting several university systems in August 2025: “The recognition of impending change is palpable. What’s lacking is vision, direction, a place to move toward” (Siemens, 2025g). “Your CIO will not get you to systems level change with AI. Neither will your provost” (Siemens, 2025h).
  • On assessment. “If a student’s essay can be created with AI, then perhaps it’s not a meaningful assessment tool, outside of it’s ability to scale” (Siemens, 2025f). With colleagues, he has argued in Nature that universities “need to radically rethink exams in the age of AI” (Kovanović et al., 2025), and he holds that “high-stakes decisions should remain in human hands” (Salisbury, 2025).
  • On credentials. Siemens names the lever that would let new models of learning compete: “Once someone solves the credential bottleneck (i.e. requirement that validation of learning comes from an accredited agency), true systems change will begin in learning, especially with learners who are not served by the existing sector” (Siemens, 2025b).
  • On what universities teach. Through “cognitive escalation”, tasks people used to learn “can now be somewhat automated and handed off”, so universities move from “teaching learners certain types of things” to “teaching learners to be certain types of ways” (Siemens, 2025d). He expects “a broad shift from epistemology to ontology”, from what students know to who they become, which “signals a critical need to articulate a new social and economic contract with society”. He calls community and social connection “the meta-durable moat in the age of AI” (Siemens, 2025i).
  • On entry-level work and struggle. The evidence on entry-level jobs is “alarming”, “hitting new grads particularly hard while actually helping existing employees continue to succeed” (Siemens, 2025g). He highlights the warning that students relying on AI may lose “essential opportunities to develop these fundamental skills”, a “weakening of the very process of becoming a mathematician” (Siemens, 2025h). With colleagues at Georgetown, the University of Minnesota and the University of South Australia, he has Gates Foundation funding to study maladaptive use of AI in learning (Siemens, 2025g).

In Reclaiming Purpose: The University in an AI World (LeBlanc et al., 2026) and in their work on “precision learning” (Gamby et al., 2025), Siemens, Paul LeBlanc and Tanya Gamby argue that institutions must move from transferring knowledge to developing human capability.

Degrees carried a weak signal before AI

Potts’s signal collapse and Siemens’s credential bottleneck both assume that the degree carries a signal worth protecting. The evidence says that signal was weak long before AI.

Health worker education shows the same pattern. The current Cochrane review of continuing education, which pooled 215 studies of more than 28,000 practitioners, found that educational meetings only slightly improved compliance with desired clinical practice (median 4 percentage points) and only slightly improved patient outcomes. Whether interactive meetings outperform lectures was downgraded to very low certainty (Forsetlund et al., 2021).

Accreditation explains why weak signals persist. Burck Smith showed in 2012 that accreditation rests “primarily on a higher education institution’s inputs rather than its outcomes.” Input-focused accreditation, taxpayer subsidies tied to it, and colleges’ control over which credits they accept form an “iron triangle” that keeps prices high and innovators out (Sadki, 2012). In 2021, I argued that the World Health Organization’s call for a “revolution” in public health education was “unlikely to be answered by institutions that form a protected monopoly”, and asked whether we could invent “a ‘lifelong credential’ that would increase in value over time, as its holder applies what was learned” (Sadki, 2021).

Read with this evidence, AI threatens the monopoly built around issuing degrees more than the information degrees carry. That is why vendor credentials matter. Anthropic’s certification programme, which tests skills in proctored exams, could outperform transcripts that predict little, while certifying the vendor’s own product. Universities could compete by certifying, independently, what a person can actually do.

Assessment can measure how well people work with AI

Most assessment reform after ChatGPT has tried to keep AI out: proctored exams, oral defences, handwritten drafts. MIT recommends oral exams, semester portfolios and in-class conversations, and warns against relying on AI detectors, which can mistake the writing of non-native English speakers or neurodivergent students for machine text (Massachusetts Institute of Technology, 2026). These methods test what a person knows unaided. They do not test what employers now pay for, which is the ability to deliver results with AI.

Lovett calls that ability Distributed Mastery: producing results by coordinating knowledge held across people and tools, including “selecting an artificial intelligence system, writing useful instructions, comparing outputs, arranging a workflow, and combining machine output with material from colleagues or databases” (Sadki, 2026f, discussing Lovett, 2026). These are real forms of professional competence, and assessing them should not be hard, because the work leaves a record. Engineering teams already judge AI agents by whether the task was completed, how, and whether the output held up when checked (Sadki, 2025b). A learner’s work with AI can be judged the same way:

  • Result. Did the learner deliver a working solution to a real problem? MIT now expects capstone students “to use AI coding tools to build near production-quality software artifacts in a single academic term.”
  • Process. What did the learner ask, which outputs did they reject, and why? MIT suggests platforms “that capture a history of versions”, submitted with the work as “process evidence.”
  • Verification. Did the learner catch the model’s errors? Students remain “responsible for all work they submit, including any inaccurate, biased, offensive, or unethical content produced with the assistance of generative AI” (Massachusetts Institute of Technology, 2026).
  • Disclosure. TGLF’s Honor Code “asks for disclosure, not abstinence: declared AI use is context, never a violation” (Sadki, 2026d).

The two kinds of assessment depend on each other. The Validation Tether holds that “people can supervise artificial intelligence only when they possess enough independent knowledge to judge its work” (Sadki, 2026f). An assessment system therefore needs unaided checks of foundations, so that learners can judge machine output, and assisted tasks that test whether they can deliver with it.

Expertise forms in work that AI is removing

My own work on health and humanitarian workforces points to the same pressure on how experts are formed.

  • Junior roles. Nolan Lovett argues that organisations “hired juniors because juniors did the junior work, and expertise regenerated as a by-product. AI severs the two” (Sadki, 2026e, discussing Lovett, 2026). An organisation that cuts entry-level roles keeps the savings and spreads the cost across everyone who will later need to hire an expert.
  • Learning. At the OECD Digital Education Outlook 2026 conference, researchers reported “performance gains alongside learning losses” and what Dragan Gašević called “metacognitive laziness” (Sadki, 2026a). At the Empowering Learners for the Age of AI conference, which Siemens co-organises, Stéphan Vincent-Lancrin of the OECD said: “performance in education is the learning” (Sadki, 2025c). Beatriz Moutinho warned that she was “even more worried about young people preemptively replacing themselves” than about AI replacing them (Sadki, 2026b).
  • Credentials and integrity. In Are courses obsolete?, I argued for separating “what a course delivers, and what a course is.” AI can now generate what a course delivers, but the course remains “the only institutional form the world has agreed to recognise, measure, and pay for” (Sadki, 2026g). When we screened every answer and every peer review in a malaria certification course, and when we built an integrity layer to scaffold peer learning at scale, we had to tell real effort apart from empty or generated work, at scale (Sadki, 2026d, 2026j). TGLF certifies only what its platform and team verify: “that a health worker completed a disciplined cycle of analysis, peer exchange, and planning about their own practice, documented in their own words.” What that cycle changed is tracked separately. In the pilot, 45.9% of analysed projects committed to a specific, dated practice change, and each claim is published as self-reported and not independently verified (Sadki, 2026d). Given how little degrees and continuing education meetings are shown to predict, a certificate that states exactly what it verified gives up nothing in credibility.
  • Sovereignty. My advice to ministries of health in You cannot download a data centre applies to education: “invest in the layer nobody can sell you, which is your people, your workforce, and the local knowledge you have”, and keep “the people who can check the validity of what these systems generate” (Sadki, 2026i). The World Bank’s jobs strategy, as I noted, assumes that the people meant to fill new roles “can perform them, adapt to change, and keep delivering results once the project closes” (Sadki, 2026c).

Six gaps in the research agenda for education and artificial intelligence

Anthropic’s S-1 assumes that the chain from tutoring to productivity already holds. The studies so far stop at the post-test. Here are six gaps that we need to close as we chart the future of education.

  1. Measure delayed retention and transfer to unfamiliar problems, not only immediate post-tests.
  2. Test which designs, such as answer restrictions and teacher-written hints, preserve independent capability.
  3. Assess both foundations without AI and delivery with AI, using results, process records, verification and disclosure.
  4. Rebuild supervised practice where junior tasks have been automated, in workplaces and in research teams, and evaluate which replacements work.
  5. Replace input-based accreditation with credentials that state what was verified, by whom, and with what limits.
  6. Follow learners from AI-supported learning to work and earnings, and record who loses access to the practice that builds expertise.

References

  1. Anthropic, PBC. (2026). Form S-1 registration statement under the Securities Act of 1933. U.S. Securities and Exchange Commission. https://www.sec.gov/Archives/edgar/data/2130357/000162828026067168/anthropicpbc-sx1.htm
  2. Bastani, H., Bastani, O., Sungu, A., Ge, H., Kabakcı, Ö., & Mariman, R. (2025). Generative AI without guardrails can harm learning: Evidence from high school mathematics. Proceedings of the National Academy of Sciences, 122(26), e2422633122. https://doi.org/10.1073/pnas.2422633122
  3. Brynjolfsson, E., Li, D., & Raymond, L. (2025). Generative AI at work. The Quarterly Journal of Economics, 140(2), 889–942. https://doi.org/10.1093/qje/qjae044
  4. Ferrer, A. M., & Riddell, W. C. (2002). The role of credentials in the Canadian labour market. Canadian Journal of Economics, 35(4), 879–905. https://econ.queensu.ca/pub/jdi/deutsch/edu_conf/Ferrer.pdf
  5. Finley, A. (2023). The career-ready graduate: What employers say about the difference college makes. American Association of Colleges and Universities. https://www.aacu.org/research/the-career-ready-graduate-what-employers-say-about-the-difference-college-makes
  6. Forsetlund, L., O’Brien, M. A., Forsén, L., Mwai, L., Reinar, L. M., Okwen, M. P., Horsley, T., & Rose, C. J. (2021). Continuing education meetings and workshops: Effects on professional practice and healthcare outcomes. Cochrane Database of Systematic Reviews, (9), CD003030. https://doi.org/10.1002/14651858.CD003030.pub3
  7. Fuller, J. B., & Raman, M. (2017). Dismissed by degrees: How degree inflation is undermining U.S. competitiveness and hurting America’s middle class. Harvard Business School. https://www.hbs.edu/ris/Publication%20Files/dismissed-by-degrees_707b3f0e-a772-40b7-8f77-aed4a16016cc.pdf
  8. Gamby, T., Kil, D., Koblic, R., LeBlanc, P., Moldoveanu, M., & Siemens, G. (2025). From personalized to precision learning: Unlocking the next transformation in higher education. EDUCAUSE Review. https://er.educause.edu/articles/2025/11/from-personalized-to-precision-learning-unlocking-the-next-transformation-in-higher-education
  9. Hao, Q., Xu, F., Li, Y., & Evans, J. (2026). Artificial intelligence tools expand scientists’ impact but contract science’s focus. Nature, 649, 1237–1243. https://doi.org/10.1038/s41586-025-09922-y
  10. Kovanović, V., Barthakur, A., Joksimović, S., & Siemens, G. (2025). Why universities need to radically rethink exams in the age of AI. Nature, 648(8092), 35–37. https://doi.org/10.1038/d41586-025-03915-7
  11. LeBlanc, P., Gamby, T., & Siemens, G. (2026). Reclaiming purpose: The university in an AI world. John Wiley & Sons. https://bpl.bibliocommons.com/v2/record/S75C9736007
  12. Lovett, N. (2026). The tragedy of the cognitive commons: How AI could disrupt the regeneration of professional expertise. Human Resource Development Review. https://doi.org/10.1177/15344843261470602
  13. Massachusetts Institute of Technology. (2026). Report of MIT’s Ad Hoc Committee on AI Use in Teaching, Learning, and Research Training. https://aiandeducation.mit.edu/report/
  14. Northcutt, C., Hasmani, I., Feng, K., Khangi, T., Plesner, A., & Mueller, J. (2026). StudentBench: AI and human tutoring yield equivalent GRE learning gains (arXiv:2609.28470) [Preprint]. arXiv. https://doi.org/10.48550/arXiv.2609.28470
  15. Potts, J. (2026). The university after AI. Social Science Research Network. https://ssrn.com/abstract=7453440
  16. Roth, P. L., BeVier, C. A., Switzer, F. S., III, & Schippmann, J. S. (1996). Meta-analyzing the relationship between grades and job performance. Journal of Applied Psychology, 81(5), 548–556. https://doi.org/10.1037/0021-9010.81.5.548
  17. Sadki, R. (2012). Accreditation in higher education is based primarily on inputs rather than outcomes. The Geneva Learning Foundation. https://www.learning.foundation/2012/09/22/accreditation-in-higher-education-is-based-primarily-on-inputs-rather-than-outcomes/
  18. Sadki, R. (2021). Can the transformation of global health education for impact rely on input-based accreditation?. The Geneva Learning Foundation. https://www.learning.foundation/2021/04/01/should-the-transformation-of-global-health-education-for-impact-rely-on-input-based-accreditation/
  19. Sadki, R. (2025a). A global health framework for Artificial Intelligence as co-worker to support networked learning and local action. The Geneva Learning Foundation. https://doi.org/10.59350/gr56c-cdd51
  20. Sadki, R. (2025b). The agentic AI revolution: What does it mean for workforce development?. The Geneva Learning Foundation. https://doi.org/10.59350/redasadki.21092
  21. Sadki, R. (2025c). The great unlearning: Notes on the Empower Learners for the Age of AI conference. The Geneva Learning Foundation. https://www.learning.foundation/2025/10/14/the-great-unlearning-notes-on-the-empower-learners-for-the-age-of-ai-conference/
  22. Sadki, R. (2026a). OECD Digital Education Outlook 2026: How can AI help human beings learn and grow?. The Geneva Learning Foundation. https://doi.org/10.59350/1bqm0-1d126
  23. Sadki, R. (2026b). AI self-replacement: What happens when we delegate our thoughts to artificial intelligence?. The Geneva Learning Foundation. https://www.learning.foundation/2026/04/03/ai-self-replacement-what-happens-when-we-delegate-our-thoughts-to-artificial-intelligence/
  24. Sadki, R. (2026c). Future of the workforce: The 780 million job gap, the AI reckoning, and the missing link in the World Bank’s biggest bet. The Geneva Learning Foundation. https://www.learning.foundation/2026/04/07/future-of-the-workforce-the-780-million-job-gap-the-ai-reckoning-and-the-missing-link-in-the-world-banks-biggest-bet/
  25. Sadki, R. (2026d). The integrity layer: Using assessment to scaffold peer learning at scale. The Geneva Learning Foundation. https://doi.org/10.59350/ympfp-gzg71
  26. Sadki, R. (2026e). Health workforce development in the Age of Intelligence: A tragedy of the cognitive commons?. The Geneva Learning Foundation. https://doi.org/10.59350/t0630-9wt12
  27. Sadki, R. (2026f). What is the Validation Tether?. The Geneva Learning Foundation. https://www.learning.foundation/2026/08/17/what-is-the-validation-tether/
  28. Sadki, R. (2026g). Are courses obsolete?. The Geneva Learning Foundation. https://www.learning.foundation/2026/08/19/are-courses-obsolete/
  29. Sadki, R. (2026h). What does the Jevons Paradox mean for artificial intelligence in global health?. The Geneva Learning Foundation. https://www.learning.foundation/2026/09/30/what-does-the-jevons-paradox-mean-for-artificial-intelligence-in-global-health/
  30. Sadki, R. (2026i). You cannot download a data centre: What does AI sovereignty mean for ministries of health?. The Geneva Learning Foundation. https://www.learning.foundation/2026/10/04/you-cannot-download-a-data-centre-what-does-ai-sovereignty-mean-for-ministries-of-health/
  31. Sadki, R. (2026j). Assessment in the Age of Artificial Intelligence: What a malaria course taught us about protecting peer learning. The Geneva Learning Foundation. https://www.learning.foundation/2026/10/07/assessment-in-the-age-of-artificial-intelligence-what-a-malaria-course-taught-us-about-protecting-peer-learning/
  32. Salisbury, A. D. (2025). George Siemens on why universities must stop outsourcing AI, and start building it. The Humanist. https://humanistxyz.substack.com/p/george-siemens-on-why-universities
  33. Schmidt, F. L., & Oh, I.-S. (2016). The validity and utility of selection methods in personnel psychology: Practical and theoretical implications of 100 years of research findings [Working paper]. https://home.ubalt.edu/tmitch/645/session%204/Schmidt%20%26%20Oh%20validity%20and%20util%20100%20yrs%20of%20research%20Wk%20PPR%202016.pdf
  34. Siemens, G. (2025a). SAIL: Systems, but still about those agents [Newsletter]. SAIL: Sensemaking AI Learning. https://buttondown.com/SAIL/archive/
  35. Siemens, G. (2025b). SAIL: Critical thinking, more DeepSeek [Newsletter]. SAIL: Sensemaking AI Learning. https://buttondown.com/SAIL/archive/
  36. Siemens, G. (2025c). SAIL: AI & science, higher education landscape [Newsletter]. SAIL: Sensemaking AI Learning. https://buttondown.com/SAIL/archive/
  37. Siemens, G. (2025d). SAIL: Students, life coach, agents [Newsletter]. SAIL: Sensemaking AI Learning. https://buttondown.com/SAIL/archive/
  38. Siemens, G. (2025e). SAIL: Meeker AI trends, fluency, model use, data centers [Newsletter]. SAIL: Sensemaking AI Learning. https://buttondown.com/SAIL/archive/
  39. Siemens, G. (2025f). SAIL: Cheating, agents, context, Meta [Newsletter]. SAIL: Sensemaking AI Learning. https://buttondown.com/SAIL/archive/
  40. Siemens, G. (2025g). SAIL: Labor [Newsletter]. SAIL: Sensemaking AI Learning. https://buttondown.com/SAIL/archive/
  41. Siemens, G. (2025h). SAIL: Systems change, head of AI, robots, hype [Newsletter]. SAIL: Sensemaking AI Learning. https://buttondown.com/SAIL/archive/
  42. Siemens, G. (2025i). SAIL: Conference, Comet, agents [Newsletter]. SAIL: Sensemaking AI Learning. https://buttondown.com/SAIL/archive/
  43. Sigelman, M., Fuller, J. B., & Martin, A. (2024). Skills-based hiring: The long road from pronouncements to practice. The Burning Glass Institute and Harvard Business School. https://www.burningglassinstitute.org/research/skills-based-hiring-2024

1 comment

Comments are closed.

Get access to the latest opportunities

Subscribe now to receive the latest opportunities from The Geneva Learning Foundation (TGLF).

We first request your permission to send you email. Please look for this secure e-mail in your inbox and click on the safe link it contains.