Our Learning Philosophy
TutorThings is built around one idea that turns out to be well-supported by research: learners grow more when they explain their thinking out loud than when they passively receive explanations.
That might sound obvious, but most educational tools are designed the opposite way. They present information, check if you got the right answer, and move on. TutorThings does something different - it asks you to explain what you think, then follows up on the reasoning.
Passive learning feels like learning, but often isn't
Reading through notes, watching an explanation, getting an answer and moving on - these feel productive, and they're not useless. But the memory formed this way is fragile. It tends to fade quickly and doesn't transfer well to new problems.
The OECD's 2026 Digital Education Outlook report put numbers to this: students using AI tools that provided answers performed up to 48% better on tasks - but when the AI was removed, they performed worse than students who had never used it at all. The output looked better. The understanding was shallower.
The reason is that passive learning does not require you to reconstruct anything. You are recognizing information as it is presented to you, not retrieving it from your own understanding. Recognition and recall are very different abilities, and only one of them builds the kind of knowledge that is useful when you face something new.
Explaining forces understanding
When you have to explain something - even imperfectly, even out loud to no one in particular - you have to construct the idea from what you know. You cannot fake it the way you can when you are reading and nodding along. Gaps become obvious, both to you and to the tutor.
This is why TutorThings asks you to explain before it offers help. Not to be difficult, not as a test - but because the act of reaching for an explanation, even an incomplete one, is where understanding starts to form.
Voice makes this easier
Speaking is faster than writing and more natural than typing. More importantly, spoken explanation tends to be less polished - you catch yourself mid-sentence, backtrack, rephrase. That kind of stumbling is useful. It shows you where your understanding is fuzzy in ways that a neatly typed sentence might hide.
TutorThings keeps learner sessions voice-led and off the keyboard because voice is the most natural medium for this kind of thinking out loud.
The tutor follows your reasoning, not a script
There is no fixed sequence of steps or predetermined lesson plan. The conversation follows what you are saying. If you explain something clearly, the tutor moves forward. If your explanation reveals a gap, the tutor focuses there - not with a lecture, but with a question that points you toward what you are missing.
Support adjusts in real time. When you're moving well, the tutor backs off. When you're stuck, hints get more specific. When an explanation isn't landing, a different angle or example is offered. The goal is to give just enough support to keep you moving, then pull back as soon as you can lead again.
Getting it wrong is part of it
Mistakes aren't failures in this model - they're information. They tell the tutor exactly where understanding breaks down, which is exactly where a session needs to focus. A learner who gets something wrong and has to figure out why will remember the correct understanding far better than one who got it right on the first try without much thought.
This shapes how the tutor responds. There are no buzzers, no shame, no "wrong, try again" loops. When something goes sideways, the response is curiosity: "What were you thinking when you said that?" The goal is to understand the reasoning, not just correct the answer.
What real progress looks like
In a system built around right answers, progress is easy to measure: you either got it or you didn't. In a system built around understanding, progress shows up differently.
You start asking better questions. Your explanations get clearer. You're willing to try hard problems instead of avoiding them. You say "I'm not sure yet" instead of guessing. You start catching your own mistakes before the tutor asks about them.
These are the signals TutorThings is designed to produce - and they are the signals that tend to predict long-term success, not just performance on the next quiz.