Guide
What does college become in the age of AI, according to Howard Gardner?
Gardner says campuses turn into lifelong learning centers: a small floor of reading, writing, math and coding, then coaching toward what you care about.
The short answer
Howard Gardner's answer is that college stops being a four-year sorting machine and becomes a place you keep coming back to. He points out that the country has six or seven thousand colleges and universities, most of them empty 90 percent of the time, and says we should convert those campuses into lifelong learning places open to people of any age. The floor stays small: reading, writing, simple calculation, and understanding what coding is. Everything above that floor looks more like coaching, where someone helps you find what you are actually interested in and you take it as far as you want. His reason is blunt. In an age where AI can do most things better than people, a curriculum built around what the SAT and the IQ test measure no longer makes sense, and the two things Gardner refuses to make optional are the respectful mind and the ethical mind.
Step by step
Keep the required floor small
Gardner would require four things of everyone: reading, writing, simple calculation, and understanding what coding is. That is it. No mandatory history, chemistry, geology, or foreign language. If a student wants any of those, the door stays open for the rest of their life, but nobody is forced through it.
Treat exposure as the real work
His model is the children's museum, where nobody ever flunks. Take a kid regularly, then watch what they return to and where they start trying something new instead of repeating themselves. Gardner's own life turned on one visit to a friend's house at age five, where he found a piano he had never seen before and started picking out tunes. His parents, who had arrived in the country with five dollars, bought him a thirty dollar piano. He still plays every day.
Use the campuses that are already sitting empty
Gardner says Harvard is a ghost town most of the year, and the other thousands of campuses are no better used. He wants them reopened as places where people of any age show up to get better at something, meet others who care about it, and work with both human teachers and large language models. You do not need a degree program to start a study group in a room nobody is using.
Learn to ask the questions, not collect the answers
Gardner credits the Australian scholar Anthea Roberts for the formula. Until 1870, law students memorized cases. From 1870 to about 2020, they answered the professor's questions. After 2020, the skill is asking questions of the machine and holding a conversation with it. The best educated people now will know which questions to ask and how to judge what comes back.
Make the models argue with each other
Roberts runs eight or ten different large language models with different kinds of expertise and has them debate. Gardner thinks that is a good use of the tools, because you learn from the argument rather than from any one answer. Ask for ten solutions, then set them against each other.
Never hand over the decision
Gardner is direct about the limit. Use every form of AI you can, take the advice, then make the judgment yourself. You cannot say a particular chatbot said this, so that is what I am doing. Humans are mortal and have a stake in the outcome, and he does not believe the machine does.
Check what you read against other sources
Gardner spends almost no time online but reads twelve newspapers a day from around the world. When the Guardian ran a story about his own school that no other paper had, his only way to test it was to check it against what he already knew and against other papers. A news feed is not a source. Ask how you know the thing in front of you is real.
Keep being a good person off the optional list
Of Gardner's five minds, three are cognitive: the disciplined mind, the synthesizing mind, and the creative mind. He is willing to make all three optional. The respectful mind and the ethical mind are not up for negotiation, because respecting other people and being a good worker and good citizen is the part we cannot consign to the machines.
“I think we need to convert those entities into lifelong learning places where people of any age who want to get better in something can come.”
Howard Gardner, in the episode
Key moments
- ▶5:04Gardner explains why the theory of Multiple Intelligences was named deliberately to challenge IQ
- ▶12:15Gardner recounts an unpublished flow moment Csikszentmihalyi witnessed on a Colorado hike
- ▶19:25Gardner describes receiving a deepfake video of himself made by New Jersey students
- ▶26:33Gardner envisions converting universities into lifelong learning centers open to all ages
- ▶27:28Gardner insists respectful and ethical minds are the only parts of education that must not be optional
- ▶33:10Gardner argues AI entities surpass humans on every dimension except consciousness
Questions people ask
what does howard gardner think college will look like in the age of AI
He thinks the four-year degree gives way to lifelong learning centers. There are six or seven thousand colleges and universities in the country, empty most of the year, and Gardner wants them converted into places anyone of any age can return to when they want to get better at something. He even says it is not at all clear whether we will have colleges at all in the current form.
does AI make the SAT and IQ test obsolete
Gardner says the curriculum those tests measure is on its way out. His comparison is that 150 years ago you needed Latin, Greek, and Hebrew to get into Harvard and nobody cared whether you could do math. Tests reward whatever a particular era values, and in an era where AI does most of those tasks better than people, building school around SAT content stops making sense.
is AI smarter than humans according to howard gardner
He says AI entities surpass human beings on every conceivable dimension except consciousness. He is careful about what he means: consciousness is not just thinking about what you are doing, because any apparatus can claim that. It is a feeling, and he does not think there is persuasive evidence machines have it.
what should schools still teach when AI can do the work
Four basics for everyone: reading, writing, simple calculation, and understanding what coding is. After that Gardner would replace requirements with coaching that helps students find what genuinely interests them. The one thing he will not make optional is the ethical and respectful part, because that is the piece he says we cannot consign to the machines.
what is the children's museum model of education
It is Gardner's picture of what school should feel like. Nobody flunks a children's museum. You bring a child back regularly and pay attention to what they return to and where they push into something new, and that tells you what to help them pursue. He would run high school and university campuses more like Aspen or Jackson Hole, with museums, concerts, festivals, and study groups people choose to join.
will AI help personalize education
Nick asked Gardner that directly and his first word was no. He then softened it: AI is not necessarily a hindrance, and we should make use of every form of it we can. But the species and the planet have a stake in the outcome that a large language model does not, so the final judgment has to stay with people.
what are howard gardner's five minds
The disciplined mind, the synthesizing mind, the creative mind, the respectful mind, and the ethical mind. The first three are cognitive and Gardner is willing to call them optional in a world with AI. The last two are not optional, because respect for other people and the habits of a good worker and good citizen are what he thinks education exists to build.
howard gardner deepfake video students
About six months before this July 2025 conversation, students in a New Jersey psychology class sent Gardner a video of himself giving a talk. It was not him. It looked like him and moved his lips, and it said some things that were accurate, some things the students thought were accurate but were not, and some things that were nonsense. He was freaked out by it, and he points out that if he died tomorrow, anyone could keep publishing in his name.
how should students use chatgpt for schoolwork
Gardner frames it as a question about the rule you set. If the rule is get to the answer as fast as possible, that is very bad for education. If the rule is get an answer, then challenge it, bring another model in to argue with it, and talk it through with a group, that is what education should be.
how do you teach ethics when you cannot trust anything online
Gardner says this is the hard question, and he raises it after a leading AI-and-kids researcher told him over lunch that she no longer believes anything she reads online or any photograph she sees. His own practice is cross-checking: he reads twelve newspapers a day rather than a feed. He also argues ethics cannot start at fifty, which is why his team now works with students down to preschool age.
is it too late to start learning something new
Gardner says you can decide late in life. His example is his wife, who is almost as old as he is and has never studied painting but wants to. His condition is that nobody should be forced into it, because the point is following what a person is actually drawn to.
why did gardner call it multiple intelligences instead of multiple talents
Because the word was the argument. Critics tell him we already knew people have different talents, and he agrees, but says if he had written a book called the theory of multiple talents nobody would know his name. Using the word intelligence was how he took on the standard IQ test and the idea that you are smart, average, or impaired across the board.
This guide is drawn from a full conversation on the show.