| Share | Pattern |
|---|---|
| 28% | asks for a finished deliverable: an essay, a calculation |
| 42% | asks, receives the answer, ends the exchange |
| 20% | multi-turn questioning, but the synthesis stays with the AI |
| 4% | genuine dialogue, the student tests their own thinking |
Outsourcing is the default case. Á. Tamási: The Power of Outsourcing (forthcoming).
The task is not to grow the 4%. The 96% can move — and course design is what moves it.
What this tells us
A general model is not enough. The course requirements, the standard for a good answer and the group's sticking points are known by the educator. That is local knowledge — course-specific AI is what brings it into play.
The source decides what an answer is worth. From authorised policies, documents and other knowledge sources, a usable answer can be given that leads to the right administrative service.
An action needs permission. AI can gather information, prepare a task, or connect to an existing system. It performs only the permitted actions, and where required a human approves.
What we build from it
There is no need to choose in advance which one comes first: the task decides, and whichever works can be taken further. In every case we say what is already available and what has to be built.
AI in university teaching
Students ask questions, request explanations and practise from the material the educator provides.
This is where GoSchool began — it runs in real educator and student use today.
Institutional knowledge and services
We design support from the faculty's and the institution's own documents. Actions inside a system always depend on the project and permissions.
A project designed around the sources, the roles and the specific task.
Processes and system connections
A solution connected to existing systems and designed with approval points.
Depending on the task, an integration or custom development.
The technical model, the permissions and the possible system connections
What we do inside the organisation
Independently of any one course, we see that university AI use raises a great many questions, different ones by role: leadership needs one direction, educators and staff need usable practice, IT needs clear boundaries. The shared need is for AI to become an ordinary working tool.
Role-specific workshops and training — for internal AI champions, among others — and AI-aware course and material development. For the AI goals set in the institutional strategy we design measurement and evaluation, and we establish the feedback loops. Independently of the technology chosen.
GoSchool has transformed our university's approach to AI-assisted learning, making it seamless and engaging for both professors and students.
Goschool.ai encompasses numerous pedagogical innovations, enabling students to creatively engage with course literature. Instructors benefit from analyzing student chat logs, allowing them to adapt course plans effectively and address knowledge gaps more precisely.
Together, we turn the goal into the right solution.
Show us the goal, the obstacle, or the initiative already in motion.