G.T. School’s Bet on Gifted Ed: Cash Rewards, 2 Hours of AI Tutoring, No Lectures

If you’ve heard of Alpha School, you’ve heard the pitch: two hours of AI tutoring in the morning, life skills in the afternoon, no teachers, top-2 percent standardized test scores. It’s the Austin, Texas, tech-money education project that’s been profiled credulously and picked apart skeptically in roughly equal measure over the past year. The Trump administration’s education secretary called the model “exemplary.” CNN ran a long piece on it in January, asking “Is AI schooling the future of education — or a risky bet?” 

G.T. School is the gifted-and-talented branch of the same network, just a few miles north of Austin’s city limits. Pamela Hobart is its gifted-and-talented-education evangelist. Trained in philosophy and education at Columbia University’s Teachers College, she spent years as a “philosophical life coach” before joining G.T. in spring 2025, after enrolling her own daughter at Alpha. She writes a Substack newsletter called Above Grade Level and is, by her own description, a partisan: She thinks academic acceleration works, that screens are tools rather than poisons, and that most American schools are quietly lying to most American parents.

She is also willing to articulate the case against her own project. The voucher program her school benefits from will probably bid up tuition the way federal aid bid up college costs. Bad ed-tech is worse than no ed-tech. Putting the tuition money in an index fund and handing it to your kid at 18 is, she’ll concede, not a crazy plan.

We met at G.T.’s Georgetown campus, where new rooms are still unfolding into the building, including a science lab, a music room, and a podcast studio for the students who want to make their own content. The conversation that emerged was less about AI than about a much older argument: that the age-based grade system is itself a failed industrial experiment, that “teacher” is a job title carrying three incompatible roles, and that the most expensive thing in American education is the time spent teaching kids what they already know.

Reason: Let’s start with the basics because most readers haven’t actually been inside one of these schools. What does a typical day at G.T. School look like for a student?

Hobart: Students roll in around 8:30—later than regular elementary school, which is great—morning kickoff, sometimes a brain teaser or teamwork thing. Then they basically do apps until lunch, with breaks. The core academic apps run on a proprietary platform called Timeback that was built for Alpha and G.T. Each grade level’s content has been broken into very specific subskills, and the apps feed each student problems at the subskill level they need for mastery.

In the morning, they go outside during one of the breaks. They have lunch at noon, then go outside again. In the afternoon, they do a rotating set of workshops designed to develop life skills in a more rigorous, more academic way than at the main Alpha school. They 3D printed and designed buildings for a city. They did a chemistry workshop where they learned principles of molecular gastronomy and had to make samples on the fly for the parents at the end of the session. They vibe-coded writing feedback apps and demonstrated them for us. The workshops here are rigorous, in the kind of nerdy way you’d expect.

No teachers; there are “guides” instead. What’s the difference?

Guides do no lecturing. That’s the main difference from a classroom teacher. They don’t deliver content. The reasoning is structural: If a guide shores up a bad division lesson by teaching division, the app stays bad. Another school running the same Timeback platform will be dependent on its teacher figuring out the lesson is bad and plugging the gap. That’s basically what teachers in regular schools do all the time: individually reinventing the wheel, fixing bad curricula they’ve been given. If they don’t figure it out until the year ends. No other student ever benefits. There’s nothing scalable about it.

Guides are prohibited from doing instruction. But they have plenty of job left. The job of “teacher” as we currently have it is massively overloaded—guide, lecturer, plus huge amounts of administrative paperwork. We say: What if there’s no paperwork, you don’t prepare lessons, you don’t lecture? Your job is to know the student. They have a meeting maybe once a week about the student’s goals, what’s going well, what isn’t. When a kid has a bad day, who’s keeping an eye on it so it doesn’t become a bad week? Doesn’t become a bad year?

In the regular system, teachers have so many students and responsibilities that you can lose literally years of progress before anyone notices. They just promote you grade to grade. That problem is even worse for gifted students, because when a gifted student loses progress, they may still be at grade level. They should have been ahead. They’re getting good grades, passing the standardized test, and nobody’s around to say, wait, this kid could have been doing better. If they’d been on their early-life trajectory, they’d be three years ahead. Now they’re just on grade level. The guides have an amazing, important job. The students love them. They’re routinely polling the students and the parents about is this guide one of the most motivational people your students ever met? Would your student rather go to school than on vacation? Stuff like that. It’s not a de-skilling of teachers. It’s just a focus on the part many of them love the most.

The official line in American education is that gifted programs are being closed for equity reasons. Do you think the programs being closed weren’t actually doing much in the first place? Make the case.

A few lucky people in public schools in better cities or districts received separate full-day gifted instruction from kindergarten or second or third grade through high school. They tend to think that’s what gifted ed is everywhere. It’s so far from the truth. When they hear about gifted programs shutting, it has a different meaning to them than when the gifted program was do Sudoku for 20 minutes once a week with a substitute teacher.

There was a movement over the past 20 or 25 years that was equity-driven. They’d look at gifted programs and see different percentages of racial groups or socioeconomic groups. They thought we need to change how we screen for talent to make the groups fair. It’s true that many underprivileged learners with high aptitude do get missed—sometimes the instruction they receive is poor, so they look average. Sometimes they don’t have parents or teachers who’ll go to bat for them.

But the way you actually find very high-ability learners falling through the cracks is you just standardized-test everyone. Give literally everyone the same aptitude test and see where the chips fall. Instead, schools went to a “multiple measures” model, standardized test score on the list, plus teachers and parents looking for creativity, leadership, sometimes kinesthetic aptitude or fine art ability. When you move to multiple measures, the group you get leaves out students with high IQ who haven’t displayed leadership—awkward nerds. It also leaves out kids who aren’t a fan favorite of a parent or teacher who’ll nominate them.

Once you have a group that’s heterogeneous academically, it can’t be a hard math program—that’s not something all those students can even do. Now they’re doing enrichment stuff that isn’t even really related to rigorous academics. It’s the old joke: This restaurant is terrible, and the portions are so small.

You screen with a real aptitude test, the CogAT, with a cutoff around the 90th percentile. Is that a deviant position now in American education?

As far as I can tell, most places using the multiple-measures model, public and private, are run by people who went through standard schools of education, and in many cases they think the problem is that those equity goals still haven’t been met. Or they think the stuff in the gifted program should actually be available to everyone.

That’s the limit case: If everyone has a special intelligence, why not? Most of those things, everyone could benefit from. I was in a gifted program where we dissected a shark. It wasn’t rocket science. We learned stuff from it, but it wasn’t rocket science. We were off there dissecting the shark, the whole school smelled, and most of the kids were stuck in the classroom doing their worksheet or whatever. It’s true that things would have been good for everyone if they could have done it.

You bribe the kids—giving them points they can bank and spend at the school emporium and, in some cases, even cash. Walk me through the philosophy, because that’s going to make a lot of people uncomfortable for different reasons.

We use motivators that are tangible, like bribes, and in some cases cash. A lot of schools won’t give you much more than a sticker. We keep kids working at hard stuff during this block of time that’s really quite challenging here at G.T. School. They’re mostly just ordering stuff off Amazon at this point. There’s a platform that facilitates banking points and then redeeming them at Alpha. For some of the younger kids, the guides get to know the students and what their interests were, and they’d flip the emporium with hand-chosen items that were the right colors and styles to try to get the students interested.

This is controversial. Some people think if you introduce rewards of this kind, students will never work without them. It’s just kind of not what we found to be the case. Other factors—getting higher levels of mastery and seeing your scores go up and feeling your competence increase—are pretty inherently motivating, even if it’s not 100 percent enough to do the full work every day. Especially where motivation is not super high to begin with, you don’t have much to lose introducing the motivator. Some of the kids just kind of ignore it after a while. They’re in their groove. They’ve developed a relationship with the guide and they’re on the sprint kind of modality. You’re not trying to bribe them to do work from sunup to sundown every day.

My daughter, when she started at Alpha, she was just buying all kinds of junk and bringing it home and stuff. She hardly looks at her points anymore. She has tens of thousands of points saved up. Sometimes the rewards shift toward experiences—you can buy lunch with your guide, and they’ll order food to be delivered. That’s blurring the line between: Are you at an AI school with no teachers, or are you working really hard because you have a good relationship with your human guide and it just happens that you answer questions on a computer? It’s much more relationship-based than people assume.

There’s an argument that if you really wanted to maximize a kid’s outcome, you wouldn’t pay $25,000 a year on private school at all. You’d put the tuition in an index fund and hand them the proceeds at 18. What’s your honest answer to that?

Our motto at Alpha Marketing has been for quite some time that we educate parents. That doesn’t mean all parents must value education or a certain type of education in the same way. What really bothers us is that a lot of schools are more or less lying to parents about what’s happening there and why, and just gross misuse of resources. When they tell you that your kid’s standardized test score—if they send it home in an envelope and never mention it, and no one ever tells you what that score means or how it compares to other schools, and then they send a report card home and your kid has all A’s—they’re telling you what you want to hear. The teacher is some super-nice lady who has a teaching degree from your friendly state university. She gets professional development all the time and is coming back from conferences. The school closes so she can get professional development. It’s all basically designed to make you think these people are experts, that they’re well-placed to tell you what’s working and what’s not, that they’ve made really good principled decisions about what curriculum and tools they’re using. All of that is just basically not really true. That doesn’t mean it’s all evil either.

There are a lot of situations where the costs of an educational choice are diffused over time and over families and over taxpayers, and the benefits are concentrated in someone selling a curriculum or someone selling a training program. They have chosen markers that have become useless. Teachers get raises when they get graduate degrees that do not seem to make them better teachers by any standard we can measure. You can’t really blame them for getting the graduate degree, but it’s just a systematic misallocation of resources.

I hear people say sometimes, why would you put your child in private school instead of just putting all that cash in an index fund and giving it to them when they’re 18? That’s also a pretty good plan. Depending on the child and depending on what the school they would otherwise attend is like, you don’t have to optimize for education. Educational nonconformism; it’s a hobby in a way, or a project you have. Something your family’s into that others may not be. I realize that’s sort of an expressive purpose as well. But it’s the lying and the omission that really bothers me. Some people would make different choices, even if it wasn’t Alpha School, if they had any idea.

On standardized tests, there’s a real and growing opt-out movement, and people have argued the tests narrow curricula and create undue stress. You’re going the other direction.

A lot of people are uncomfortable with standardized tests. They think tests leave out something important, or that test anxiety prevents students from revealing what they know. But when push comes to shove, test anxiety, unless someone truly has an anxiety disorder per se, is usually about simply not being prepared for the test, not having been taught well, or not having received enough reps at that skill where you have confidence when the test comes that you can just do it. The best antidote is to simply prepare the students for the test so they know they can do it.

It’s true that getting good standardized test scores is not everything in life. Some of those softer skills are a little harder to assess, but it’s not a reason to ignore the standardized test scores or to give up on it. Our attitude at Alpha and G.T. School is much closer to: If you can totally crush the test with a few hours a day of good academic work, then just totally crush the test and move on to other stuff. Other people are urging you not to take the test or not care what score you got. It’s much more like why not both?

You once mentioned a 1909 report called Laggards in Our Schools that you think explains a lot about why the system fails kids at every level, not just gifted ones. Walk me through it.

I went looking because I was wondering what things were like when they cooked up grade levels—industrial America. Laggards in Our Schools was written in 1909, when city schools had outgrown the one-room schoolhouse and were getting multiple grades. They didn’t move students up for social reasons. If you didn’t pass the test at the end of the year, they held you back. What they found, just a decade into the experiment, was that 30 or 40 percent of students were already the wrong age for the grade. They hadn’t learned the material they were supposed to learn, and there was never any plan to catch them up. They just kept them in the same grade. Students ended up one, two, three years behind—right out of the gate of the age-based grade system.

Since then, everything about it has only calcified. All the laws about funding students, how long they can be in education, what you have to give them, it’s all based on credit hours and the age-based grade level. When I look at what I know now, this isn’t really about carving out a special set of kids who’d “be fine anyway” and giving them more. It’s the much more basic question: Is the educational treatment we’re spending money to provide actually doing anything for the people we’re administering it to?

If it’s the wrong level in any direction, it’s done nothing. There’s a paper that asks: What’s the financial cost of teaching students things they already know? If you’re in a class and that teacher is paid, their health insurance, their pension, it’s waste. But it’s a very quiet waste that no one pays attention to. If there were government employees all over the country being paid to wash windows that were already clean, you’d say, why are they double window washing? Why are my tax dollars washing clean windows? Every day in every classroom, there are students ahead, and money is being spent on them this way that affects nothing.

Texas has a new school choice program, Texas Education Freedom Accounts (TEFA). The libertarian temptation is to be unambiguously for vouchers. You’re more careful than that. What’s the worry?

It’s been a challenging one for homeschoolers of a libertarian bent, in particular. You can accept TEFA funds as a homeschooler, but the funding amount is only $2,000—enough for materials. Things like our new remote program, G.T. Anywhere, bend the categories: It’s virtual. Some of the students were being homeschooled, but to receive the $10,000 they have to self-categorize as private school students. Some people, in a principled way, don’t want to be thought of as private school students. They consider themselves homeschoolers—not under the thumb of government education. But you get $8,000 more if you do it.

The bigger structural worry is the one our economist friends always raise: When you give someone a subsidy, it’s sometimes like college. Part of why college tuitions have skyrocketed is due to subsidized tuition—they keep pace. You never really get ahead in some cases, even if you’re subsidized. We’re hoping there’s enough entrance into the microschool market that more resources get marshaled into alternative education uses, and it will not be just the TEFA dollars chasing a fixed pie of private schools.

[With] G.T. Anywhere, we fixed at exactly the TEFA amount, so it’s free for people who got it this year. In-person G.T. School is more like $25,000 a year, so people are topping up. Whether TEFA mostly pays families who could already afford private school to stay in private school, or whether it actually moves kids out of homeschool or public school into private school—that’s what everyone is watching to see.

You’re working at an AI school. But you’re also clearly skeptical of ed-tech writ large. How do you hold those positions together?

The thing I find hard to convey in discourse about ed-tech is that poor educational technology—poor-quality apps—may well be worse than nothing. It may be the case that for all kinds of political reasons, public schools can’t really do this well. They’re going to get vendors that are just as bad as the other curriculum vendors they’ve used. If I weren’t putting my kid in an AI school with someone I could trust, I’d rather they went to a classical school. It looks as though we may get the compromise that’s the worst of all worlds—techified, but still expensive, with a ton of people involved who aren’t using their comparative advantage.

There’s a backlash right now against Chromebooks in schools and screen time. It’s very close to being just knee-jerk in the other direction. In general, it’s better for there to be some heterogeneity, but I wouldn’t mistake that position for thinking that literally all schools are equally good for the right students. People also don’t agree on what they’re even trying to do. Education just touches on a lot of different areas.

Why does gifted and talented education matter? What do you say to people who want to abolish it because it promotes inequality?

The educational ideal we already have isn’t what people think it is. Grade levels weren’t pulled out of a hat. They chose levels of materials that were supposed to be appropriate for each group of students, because that’s what’s appropriate for them. But when you look under the hood, what you find is that’s just not true. The idea of what’s appropriate for a child is best determined at the individual level. We never used to have any way at all of accomplishing that other than being born an aristocrat and having a private tutor.

I went digging for this. It’s hard to find because they often group at-or-above grade level into the same bucket. But if you tried to divide students into below grade level, at grade level, and above grade level, there would actually be more students off grade level in America today than on. It’s not that there’s a thick middle of a bell curve being well-served and a few oddballs who have to deal with it. The baseline expectation is that what your child receives in school is, in some important sense, not appropriate for them. That tells me the age-based grade system is overdue for an overhaul. It’s not a question of carving out the gifted kids. It’s that the system is administered roughly to almost everyone, and the gifted kids are just the most legible case of the failure.

There are still very thorny questions about how to get students to use their AI tutor. That’s where all the bribes and stuff come in. Because it is definitely not expected or evenly distributed that kids will even do it. We’ve seen many experiments—MOOCs [massive open online courses], Khan Academy—where very few people actually use it. But the question is: Is that more or less solvable than overhauling age-based grade-level education across the entire workforce? You would have to retrain teachers pretty much everywhere in actual learning science, which isn’t a thing they received during their real teacher training. But a lot of them, if they wanted to, could be pretty decent guides. If you resourced them properly and put them in a room with a smaller number of students, the students could work at different levels. The teachers would be there to interface with them and get to know them and have relationships with them. It’s the teachers that are choosing to see our vision as dehumanizing in some way.

This conversation has been edited for length and clarity.