Sub-shift
“LEARNING WITHOUT WAITING”
Consumers will expect useful teaching in the middle of doing.
A library of courses puts a great deal of teaching within reach. Finding the explanation for a particular mistake can still take considerable work. A teacher can respond to that difficulty, but their time has to be shared and scheduled.
Responsive AI tutoring could change those practical limits. People can ask again, approach an idea from another angle and seek help while they are working on something. For adult consumers, the next step is learning that follows an intention: the conversation they want to have, the project they want to finish, the subject they finally want to understand.
This builds on years of on-demand lessons and tutorials. The new expectation is that an explanation should respond to the learner, including when the first explanation fails. A course or teacher can provide continuity; the help between sessions can become far more personal.
Availability is only the beginning
A useful tutor has to do more than reply quickly. It needs to recognize the misunderstanding, offer an appropriate next step and leave enough work for the learner to do. Someone who has asked to learn may be poorly served by a system that simply completes the task.
Research published on 29 October 2026 by Google and mathematics-learning platform Eedi makes the distinction tangible. In a 10-week trial involving 1,179 UK secondary-school students, supervised AI tutoring improved narrow learning measures relative to the platform’s standard support. Comparable benefits did not appear on the broader standardized test. Human experts reviewed every AI message before delivery. [107]
The findings support a carefully designed teaching approach in that setting, not a promise that an unattended chatbot can teach anything. They also give providers a more useful question than how many answers their tool can generate: what can a person understand and do after using it?
The new expectation is that an explanation should respond to the learner, including when the first explanation fails.
Build around what people want to achieve
For businesses beyond formal education, there is room to connect instruction with the activity that brought the customer to you. A hobby retailer might support people between workshops. A travel business could help them rehearse a conversation before a trip. Whether rehearsal helps in the real situation needs to be tested.
The commercial relationship could extend well beyond a transaction. People may return because your service helps them continue, correct a mistake or attempt something more demanding. That requires attention to progress over time, including whether the help is producing understanding or merely a succession of completed exercises.
Human teaching also gains something when learners arrive with specific questions and attempts to discuss. Build a way to move between automated help, a class and a qualified person, with the learner’s permission. The information carried forward should help the next person teach, rather than make the customer explain everything again.
When a customer gets stuck with what you sell, how much learning could happen before the next appointment?
Early signs of change.
by TrendWatching
1 closely related example from the Innovation Database. Each helps explain an aspect of the shift. None establishes that the whole forecast will happen.
English practice becomes a conversation you can start now
Speak is an AI language-learning platform.
Feedback arrives while the learner is practicing a conversation, with review material built from that practice.
Why this example is here
A paid-plan feature with usage limits. AI responses may be wrong; availability of practice does not prove lasting fluency gains.
Original sourceEditorial selection · reviewed 1 Nov 2026
Dates distinguish launches, announcements and substantial updates. Examples with a more indirect connection are labelled.
See the full selection for “CAPABILITY UNBOUND”THE SOURCE NOTESFollow the thinking 1 source · what they show and where their limits lie
Sources and further reading
These sources inform the article. A product announcement describes an offer, not how widely people use it. Arguments and imagined futures are kept separate from observed results.
The study involved 1,179 students at 10 UK schools over 10 weeks. A human reviewed every AI message before a student received it. Gains appeared on narrower and intermediate measures, without a comparable overall benefit on Star Maths. We read the abstract, study design and results. This supplier-conducted study does not establish that unsupervised AI improves consumer learning in general.
About this research edition
This article draws on selected sources and our interpretation of their implications for consumers. The wider collection contains 50 articles and original illustrations. We have not read everything by everyone on the reading list, calculated the probability of each future or established worldwide adoption. The collection is not exhaustive.
The illustrations are AI-generated, with a visual approach informed by TrendWatching’s 2024 and 2025 reports. They depict ideas rather than observed events. If an article’s central idea changes, its image needs another look. Small edits need not mean a new illustration; automatic replacement is not active.
Each innovation has a source, a review date and an explanation of its relevance and limits. Some are early signs of the change; others offer a useful comparison. None proves the forecast. Monthly monitoring and automatic selection of new examples are not running.