Online: Trending Now

Artificial Intelligence continues to develop at an unprecedented rate and scale. What will be the changes that we will see in higher education by the end of this year?

Let’s begin with the advent of generative AI (GenAI) in 2022. It was on November 30^th of that year that OpenAI released ChatGPT. The release resulted in the fastest adoption of a technology at scale. Just two months later, by the end of January, 2023, Chat GPT had more than 100 million users. The developments have continued at breakneck speed with an increasing number of corporations, start-ups, and even governmental programs dedicated to maintaining accelerating progress. We have seen this initial level provide impressive improvements such as Khanmigo, the 24/7 learning tutor developed jointly by Khan Academy and OpenAI. Another early AI product used widely has been the Grammarly AI writing assistant.

The version of AI in those days was GenAI running chatbots. That is the threshold of a five-step development of AI as desribed by the CEO of OpenAI, Sam Altman:

• Level 1: Chatbots, AI with conversational language
• Level 2: Reasoners, human-level problem solving
• Level 3: Agents, systems that can take actions
• Level 4: Innovators, AI that can aid in invention [AGI]
• Level 5: Organizations, AI that can do the work of an organization [ASI]

Just last year, we saw the development of level 2, reasoning AI, in the form of OpenAI o1. A preview of o1 was released by OpenAI on September 12, 2024. It spends time “thinking” before it answers, making it better at complex reasoning tasks, science and programming than chatbots. The full version of o1 was released on December 5, 2024. This brought to us a series of versions across multiple providers that have been tested against standards of knowledge and reasoning of humans. On Ph.D. level questions in science, o1 scored 78 compared to human experts with discipline relevant Ph.D.s on the GPQA Diamond test who scored, on average, 69.7.

On January 23, 2025, OpenAI released a research preview of an agent called Operator, level 3, that can use its own browser to perform tasks for users. The tool is still in preview. It will require further development and refinement. Yet, this early version of a Computer Using Agent (CUA) shows the enormous potential of the tool to enhance and enable efficiency and effectiveness in daily use in higher education teaching, learning and administration. Still to come this year is likely to be the level 4 “Innovator” that will mark Artificial General Intelligence (AGI). The AGI definition varies, but centers on an AI tool that encompasses broadly the collective knowledge and intelligence of a human. There is speculation that AGI does already exist in developmental models at the frontier AI enterprises such as OpenAI, Microsoft, Google, Anthropic, Meta and others. It may be two more years (2027) before the awe-inspiring Artificial Super Intelligent (ASI) tools are released.

In an interview with Reddit co-founder, Alexis Ohanian, Forbes reports that OpenAI CEO Sam Altman said: “In my little groupchat with my tech CEO friends there’s this betting pool for the first year that there is a one-person billion-dollar company,” Altman told Ohanian. “Which would have been unimaginable without AI and now will happen.”  Speculation ranges from 2026 to 2027 for us to see in real life, AI capability to run an entire billion-dollar company, with only one person, presumably the CEO, to be engaged as an employee and an autonomous intelligent AI agent.

So, what can we expect by the end of the 2025 fall semester? How might a robust Agentic AI begin to contribute in higher education? First, in teaching in learning, we will see a far more robust, creative, and consistently contributing AI agent throughout the development and delivery of the course.  It may be helpful to consider the agent to be the virtual equivalent of a Ph.D.-holding research and teaching assistant working 24/7 on the project without break, every day of the year.  I envision by the end of the fall term agentic AI will be used by faculty, instructional design staff, and students on a daily basis:

• For faculty members, agentic AI will provide services daily, or more often:
  • Review student submissions into online discussion boards, posting follow-up questions and stimulating discussions (subject to faculty approval)
  • Scan newly released research reports and related literature relevant to the course topic, posting this to the LMS for student review and response (subject to faculty approval)
  • Based on the continuing scan of relevant research, the agent will recommend to the faculty member additional material to be inserted into the schedule to accommodate the updates, including lecture notes and even quizzes as the agent sees appropriate.
  • Review student work and automatically inform the faculty member of assignments that are missing, tardy or outside the norm of responses in terms of content or length of postings that are submitted (subject to faculty approval with copies to the students involved)
  • In reviewing student work, the agent will create, or identify, relevant tutorials to assist students in correcting misconceptions or incomplete understanding of concepts offered in the class. 
  • Automatic, instant responses to student questions regarding course management, such as deadlines and other expectations.  Reminders to students based on their recorded time in LMS app, postings, etc. Parameters will be set by the faculty member. 
  • Perhaps more controversial, the agent may review research submissions by students, providing suggestions based on the actual sources cited, key missing sources, nuanced interpretation of source materials, etc. (subject to faculty approval)
  • At the end of the term, the agent will assess results and make recommendations of improvements in course design and content to be implemented in the next offering of the course. Recommendations may also be made to other courses, particularly pre-requisites, in the relevant curriculum to ensure a seamless, consistent flow of learning.
• For instructional designers the engagement will likely be most intense prior to the course offering.
  • With each course syllabus submitted, the agent will test and review learning outcomes, relevancy to the current and anticipated career fields.
  • Suggestions will be made regarding emphases, topical areas, and pedagogies included in the course design.
  • A comprehensive list of relevant resources will be compiled with annotations for the subject matter expert (SME) and instructional designer (ID) to consider. This could be updated as they are released throughout the ensuing semester.
  • Graphics, quizzes, and other such learning tools will be custom-created to meet the expectations of the SME and ID.

Important contributions will be made in the administration of departments, colleges and the university at large. There are too many to enumerate in this posting. Here are some examples:

• The doctoral-level research, writing, and 24/7 nature of the contributions will be the same.
• Agents will tackle the mountains of paperwork required of universities, collecting data, drafting reports, requests and recommendations that, at first, will be subject to administrative review and approval
• Agents will scrutinize enrollment campaigns; it will advise even micro-changes in copy, dissemination, and graphics to optimize results, later assessing outcomes and making further revisions
• Marketing plans will be built upon deep data analytics responding to the prospective students and identify new populations to recruit
• Efficiencies and advances in the daily operation of the physical plant and grounds will be conducted
• Events scheduling and marketing will be coordinated and enhanced
• Monitoring of formal actions, public discussions, and recommendations of governing bodies and related entities will be reported on a minute-to-minute basis to anticipate and identify relevant changes in policy and practice
• Of course, accounting work will be conducted by the agent, including budget projections, identifying opportunities for cost-saving efficiencies, and adjustments to meet changing priorities
• Optimum scheduling of staff will be achieved by the agent in order to save expenditures and enhance work outcomes.
• Job applications will be reviewed and validated by the agent which will then sort them in preference order, detailing the reasons for those recommendations
• The agent will be aware of the latest products and strategies that are being used by other universities; it will project and calculate enhancements and efficiencies to accompany recommendations as opportunities emerge

Even as I am drafting this posting, Chinese company DeepSeek is shaking the AI world with an open source version of reasoning level AI, called R1. This promises even greater competition, lower consumer costs and possibly an international race to advance AI.  As we move into 2026, higher education will have the tools to be much more effective and responsive. Agents have begun arriving. We need only to insert our commands and the smart, swift tools will instantly respond. Will your university be prepared to install and take advantage of Agentic AI to improve the institution at all levels by the end of this year?