Where the name comes from
McMinsky comes from Marvin Minsky. The pedagogy comes from Seymour Papert. They worked together at the MIT AI Lab and co-authored "Perceptrons" (1969), the book that dismantled a trend and forced an entire field to think better. That, on its own, already says something about this house: no nonsense before it becomes a slogan.
We are not affiliated with MIT nor with either figure. We use their ideas, tested for decades, and cross them with what we see every afternoon with students, families and now also with institutions. What follows is that.
Twelve ideas, and how they show up in the house
1. Society of mind
Minsky, "Society of Mind", 1986. Intelligence does not come from a single principle, it comes from a society of simple agents, none of them intelligent on its own.
Application: we do not sell "the method". We assemble the right set of pieces, class, training, measurement, habit, family, technology, for each person and each population. It is also the explanation of our name.
2. Learning more than one way
Minsky, "Society of Mind". You only truly understand something once you have learned it through more than one path.
Application: every subject through two or three paths, symbol, image, body, simulation. It applies to a 9th grade student and to a municipal worker learning a new tool.
3. Reorganising what you already know
Minsky, citing Papert's principle in "Society of Mind". Some of the most important steps in mental growth are not about acquiring new skills, they are about acquiring new ways to use what one already knows.
Application: in adults and seniors, the biggest gain is rarely "more content". It is reorganising what the person already has. It is the underlying argument for worker training and active ageing.
4. Emotions as ways of thinking
Minsky, "The Emotion Machine", 2006. Emotions are not the opposite of reason, they are different ways for the brain to select resources.
Application: motivation, frustration and fear of failure are part of the design of any program, they are not noise to eliminate. It is the basis of understanding people before solving anything at all.
5. Negative expertise
Minsky, "Negative Expertise", 1994. A large part of knowing something is knowing what not to do.
Application: our studies and training spend as much time mapping what usually goes wrong, at school, at the municipality, at home, as proposing what to do.
6. Constructionism
Papert, "Mindstorms", 1980. You learn better by building something you can show to others than by listening to someone explain it.
Application: you leave with something made, an agent, a video, a plan, a simulator, never just slides.
7. Objects to think with
Papert, foreword to "Mindstorms". A concrete, loved object can carry an abstract idea into the mind.
Application: the skeleton, the microscope, the electrocardiograph, the pianos and the simulators in our space are not decoration, they are the pedagogy.
8. Microworlds
Papert. Small, safe environments where the rules are visible and mistakes are cheap.
Application: exam simulations, financial simulations, a study room with gamification. It is also the right format for an institutional session: a morning where you see the thing working, instead of hearing about it.
9. Low floor, high ceiling, wide walls
Papert (LOGO) and Mitchel Resnick, his student, who added "wide walls". Easy to start, no ceiling on progress, with many possible paths.
Application: the same program serves the child with difficulties and the gifted one, the senior afraid of technology and the retired professional. It is the practical answer to the word "inclusion", which everyone asks for and few know how to deliver.
10. Hard fun
Papert, column in the Bangor Daily News, 2002. Children do not want it to be easy, they want it to be hard and fun at the same time.
Application: that is exactly "Deep work. Deep fun." Now you know where it comes from.
11. The error as data
Papert, "Mindstorms". The right question is not "did you get it right?", it is "what does the mistake say about how you are thinking?".
Application: the error log in exams, the diagnosis of difficulties in the study room, and the evaluation of an institutional program all treat the mistake as information, not as shame to hide from the report.
12. Technocentrism
Papert, 1987. The mistake of asking "what does the computer do to children", instead of "what do children do with the computer".
Application: it is the line that separates us from those who sell platforms and screens to a municipality or a school. A thousand phones without a learning layer is not digital literacy. It is also our position on artificial intelligence: a tool in the hands of someone who thinks, never a substitute for someone who thinks.
Six principles with strong evidence
Active recall
Testing yourself on the material, instead of rereading it, is one of the most studied and most underused techniques in the whole of learning science.
Application: our programs close every session by testing what stuck, never just reviewing what was read.
Spacing
Spreading study out over time, instead of cramming it the night before, fixes more and forgets less afterwards. It is one of the most replicated regularities in cognitive psychology.
Application: subjects come back on purpose, weeks after they were first taught, instead of disappearing from the calendar as soon as the test is over.
Desirable difficulties
Learning costs more when the path is harder, and it is that extra effort that makes the difference stick. Making the path too easy also makes forgetting easier.
Application: we do not give the answer first. We let the person try, get it wrong, and only then correct it, with the mistake in view.
Fast feedback
The sooner someone knows they got it wrong and why, the faster they correct it. Feedback that takes weeks no longer teaches much of anything.
Application: the dashboard shared with parents, or the report from an institutional training session, does not wait for the end of term or the end of the contract.
High-dosage tutoring
A recent meta-analysis of small-scale tutoring, with 270 thousand students across 213 programs (Nickow, Oreopoulos and Quan, 2024), found an average effect of 0.29 standard deviations, one of the most effective educational interventions that exist, especially when led by teachers or paraprofessionals rather than volunteers alone.
Application: it is what we do every afternoon, in small groups, with the same person following the same student over time.
Socioemotional skills
A meta-analysis of 213 universal programs and 270 thousand students (Durlak et al., 2011) found an average gain of 11 percentile points in academic performance when the socioemotional component is taught seriously, not as an add-on.
Application: time management, communication and emotional regulation enter the program alongside the school subject, not apart from it.
Other figures we use on other pages: Portugal at the OECD average in mathematics, reading and science, and declining since 2018 (OECD, PISA 2022); one in seven adolescents aged 10 to 19 lives with a mental disorder (WHO, 2025); Portugal with 24.3% of its population aged 65 or over in 2025 (Eurostat); 34.1% of Portuguese adults with low combined literacy and numeracy performance (OECD, PIAAC 2023); basic digital skills still below the European average (European Commission, Digital Decade 2025); 14 modifiable risk factors account for around 45% of dementia cases (Lancet Commission on Dementia, 2024).
What we do not do
This is not a medical, psychological or psychiatric consultation. We do not replace those professionals, we complement what is missing between what school does not solve and what nobody has time to explain to families or teams.
We do not replace the school nor the school group. We work with whoever is already there, never above them.
We do not promise grades or guaranteed results. We promise method, measurement, and the honesty to say beforehand what is not going to work.
We build our own tools and only use them where they help the specific person in front of us.