Unbound Academy hasn’t replaced teachers with AI | Hacker News Reader

Unbound Academy hasn’t replaced teachers with AI

danmeyer.substack.com

There's a lengthy, and quite good, deep-dive into Alpha School by a current parent here, for anyone interested. Spoiler, "AI" isn't that big a portion of what they're doing, but some of their insights and systems around student motivation are actually interesting and very effective.

https://www.astralcodexten.com/p/your-review-alpha-school

I heard one of the founders interviewed on the Hard Fork podcast[1] (which confusingly is primarily concerned with AI, rather than crypto.) I went in with very negative expectations, but came away with a positive impression and optimism that they might be onto something. As you say, AI is not core to the project. Instead, the focus is on using technology to facilitate individualized learning. It is true that teachers are 'replaced', but by humans whose job it is to keep the students focused and motivated, rather than to convey information.

1. https://www.nytimes.com/2025/09/05/podcasts/hardfork-educati...

I had the same reaction to this podcast:

https://joincolossus.com/episode/building-alpha-school-and-t...

I like the vision and believe in the good intentions. I don't know whether they've achieved much so far.

I've had the same opinion since I was a TA. Most of the stuff the students learned was from the textbook. The value the instructors provide should above all be the motivation , the enthusiasm and the instilment of meaning into what the students are learning.

>Instead, the focus is on using technology to facilitate individualized learning. It is true that teachers are 'replaced', but by humans whose job it is to keep the students focused and motivated, rather than to convey information.

I have good news for you. Tell your kids to get into a few fights, or get caught smoking weed in school too many times. They will be sent to an "alternative school" as punishment that uses this same insight - let the kids sit in front of a computer all day "learning" while a teacher nags them to quit falling asleep. In fact, they can do it for around 6 hours a day, three times better than this charter.

If your child took the MAP Growth test in Fall 2024 or Spring 2025, you can compare their RIT scores to the mean score of an Alpha School student in the same grade:

https://go.alpha.school/hubfs/MAP%20Results%20-%2024%2025/20...

Assuming a normal distribution, this will indicate whether your child is above or below the median Alpha School student. This may be impact your view about how well Alpha School is doing vs whatever school your kid goes to.

Not to go against the OP, but that headline has to be one of the dumbest framings I've seen in a while.. why would you want to replace human teachers with machines? There are an estimated 250 million kids going without any kind of schooling in the world [1], if AI could provide even the most basic kind of education to them it would be a net benefit.

[1] https://news.un.org/en/story/2024/10/1156366

Realistically though, how would it ever come to that? This sounds like the least profitable thing anyone could do.

Crazy, right? Doing a good thing that's a net positive for humankind is being punished by the market.. it's almost like markets don't foster beneficial long-term outcomes and we need other entities to enforce those.

This is OLPC levels of naivety, chat bots aren't going to solve the socioeconomic and political problems that cause illiteracy.

If you want to reach more kids with the same teachers, you need fewer teachers per student. Unbound has more teachers per student, so it's in the opposite direction from that.

The tech industry has played a big role in scamming society into changing education to conform to them, instead of the opposite. Nowadays, schools are trying to "educate" children to use technology and in the process they're making the other more important goals become secondary. Critical thinking, reading, writing, math skills, etc., are all going down because schools think it is more important to use the latest gadget and software.

Being a teacher nowadays must be tough, low pay, students that don't respect their teachers, low-attention span of kids due to tech.

Depending on the area you teach, you might just be a daycare worker.

This could be a quote from some Roman around the year 100 :p

Regarding the 'low pay' part of your comment.

Pay varies significantly between different states and cities.

There are elementary school teachers in San Francisco whose total pay and benefits in 2023 were $150k or more:

https://transparentcalifornia.com/salaries/search/?q=Eld%20C...

And to compare those salaries to other jobs, you have to consider:

- the typical academic achievement of those teachers and the alternative roles available to them, and

- the fact that in another role they would have to work 25% more (50 weeks per year instead of 40 weeks per year)

> There are elementary school teachers in San Francisco whose total pay and benefits in 2023 were $150k or more:

In 2025 San Francisco we have ads on buses advertising low income housing to people making less than $185k/year.

https://x.com/Swizec/status/1972053995050119631

America really does sometimes looks like a first world and third world country in a trench coat.

Because we provide social housing for people? That seems hardly a sin. The limits published here seem reasonable https://www.sf.gov/reports--may-2024--income-and-rent-limits...

So a primary school teacher who is the sole breadwinner of a 4-person household will be eligible for housing support. That doesn't seem that outlandish. If anything it seems very much like a developed nation property. I grew up in the Third World: America is nothing like it. Even the ways it fails are not like the Third World.

Oh no no, just a superficial observation of how you can see a streetcar ad that reads like budget homes… for those under $185k.

My region’s version of that ad is just as ridiculous to 95% of the world.

However I do think the U.S. does have a lot of range, which does look really weird at times. You can cross a county line and suddenly the roads turn into an amusement park ride. I think that’s the main “whaaat?” That comes to mind when I’ve travelled there. Well, that and abandoned cars on the side of the road.

That is an entirely justified opinion. The range is truly wild.

That $185k is 110% of AMI (Area Median Income) for a household of 5 people.[0] If you're a single-income household with 4 dependents then, yes, you may find it difficult to live in San Francisco.

But, again, CoL isn't relevant when we're discussing whether a particular job has 'low pay'. My point was (and is) that teachers in San Francisco are well paid compared with people with similar academic achievement and similar capabilities in the same city. Even when you don't adjust for the number of weeks they work.

[0] Here are the 2025 110% AMI caps by household size from the City’s chart (gross income, before taxes):

1 person: $120,000

2 people: $137,150

3 people: $154,300

4 people: $171,450

5 people: $185,150 ← this is the "$185k*" you saw

6 people: $198,900

https://media.api.sf.gov/documents/2025_AMI-IncomeLimits-HMF...

> But, again, CoL isn't relevant when we're discussing whether a particular job has 'low pay'

Well it’s a little relevant when talking specific numbers. It’s hard to know if $number is high or low without some reference.

Thanks for clarifying the AMI stuff

  It’s hard to know if $number is high or low without some reference.

Yes, and the correct reference is pay for other jobs in the same location that require comparable aptitude, level of skill and the like.

Also kind of important to have teachers that can live within commuting distance of where they are teaching, so CoL ends up pretty important. Even more so if you want teachers that are older than, say, 22 and might have a partner and/or kids so won't want to live in a studio or have multiple roommates.

The median individual income in San Francisco in 2023 was 69,260 USD. The median household income (which may include income from more than one earner) was 141,446 USD.

Many people commute to San Francisco from other places in the Bay Area.

CoL isn't a concern that is unique to teachers. When discussing 'low pay' of a particular job, it's relevant to compare it with other jobs in the same location, which those same people might be able to get.

(Also - there is research (which I don't have time to dig up now) that shows public school teachers who leave teaching tend to earn the same or less in their new career.)

> CoL isn't a concern that is unique to teachers

Of course it's not, which is exactly the point. In many places, low teacher pay is predicated on the fact that there are plenty of people willing to accept the low pay because there are plenty of people willing to accept the low pay. That's not true everywhere.

If you want to have teachers at your school, you have to pay them enough to live within commuting distance, subject not just to tolerance of the commute, but you're also in competition with all the communities also within commuting distance. In my experience, aside from usual teacher attrition, Bay Area teachers don't leave for a different career, they leave for a different location, and usually those places need teachers too.

Are you saying that teachers working for SFUSD have 'low pay'?

$150k is not that much for San Francisco, especially when it includes benefits. Two people earning that salary would not affort the median home in the area. In that sense, it's lower than the salary people generally expect from a middle class job.

$150k is over 2x the median individual income for San Francisco.

The listed salaries are in the $110k to $120k range. To reach $150k, you need to add benefits such as healthcare and employer matches to retirement contributions.

Those numbers are marginally higher than the median individual income for someone with a bachelor's degree in San Francisco ($106.5k in 2023, according to some estimates). Even the best-paid teachers only get an average salary, while most are paid much less. And because it's San Francisco, that average salary is worth less than the average salary in an average location.

There are many conflicting estimates of the annual working hours of teachers. But according to most of them, teachers work more than the average full-time job.

  the median individual income for someone with a bachelor's degree in San Francisco

Do you believe that teachers working at SFUSD schools are comparable to the median person with a bachelor's degree in San Francisco?

  you need to add benefits such as healthcare and employer matches to retirement contributions

Why would you not include those in the comparison? For most private employers, these amounts are single digits or zero. But for SFUSD, being able to retire and still be paid 90% of your salary indefinitely is a major part of the total compensation story.

If you compare base salary only, you're not comparing like with like. You get closer to comparing like with like if you compare SFUSD total comp with private sector current comp (base salary and bonus).

The point was that teachers earn less than what they could get in other jobs in the same area.

Comparisons should be based on the lifestyle the compensation enables rather than some abstract numbers. People generally have the greatest need for money when they are buying a home, starting a family, and raising kids. Money in hand today is much more valuable than mandatory retirement benefits you can only enjoy decades later.

  The point was that teachers earn less than what they could get in other jobs in the same area.

I don't believe that's true.

  People generally have the greatest need for money when they are buying a home...

Maybe, but the fact that you don't have to save for retirement frees up money to spend on a home!

Most pensions are not purely free money. Employees are usually required to contribute a nontrivial fraction of their nominal salary towards the pension. For SFUSD, the typical rate seems to be 9%.

Yeah, I think they pay about a third of the total contributions.

But the state also steps in to pay for any shortfalls. So I think effectively pay less than a third of total contributions.

You appear to be very strongly invested in this completely bizarre narrative that teachers in SFUSD are not low paid, and basically every number you’ve used is wrong or misleading. Nobody is getting 90% of their salary as a pension. If you retire at 62 after 30 years teaching you will get 66% of your salary as a pension.

Personally, in my private sector job, my healthcare and retirement matching is more than the benefits amount for SFUSD, so excluding that is just ridiculous.

  Nobody is getting 90% of their salary as a pension

If you work from 27 to 65 (38 years) as a teacher at SFUSD, your pension would be calculated as:

Years of service * 2.4 % * final salary

In retirement, you would earn more than 90% of your final salary.

  Personally, in my private sector job, my healthcare and retirement matching is more than the benefits amount for SFUSD, so excluding that is just ridiculous.

That may be true in your private sector job.

oh man, you're right - there are some people in California who get that! I should have said "90% permanent salary replacement is so far from normal that nobody can in good faith call it a major part of the compensation story. Anyone who used that as advertising for the job would be sued for lying. The average length of service at time of retirement across the entire California public education system is only 25 years. Only around 11% of all CA education employees even hit 35 years at retirement, and less than a quarter of those make it to 40, so a maximum of 5% of all education staff end up with a 90% salary replacement - and this group is more likely to be administrative staff than elementary school teachers, and more likely to be out of SF than in it."

BUT for the arbitrary reason that it supports my argument, we are going to ignore any teachers that are getting 90% salary replacement, as a fair way to compensate for arbitrarily discarding all private sector benefits from the comparison. So I was right after all - nobody in the comparison group is getting 90% salary replacement.

> - the fact that in another role they would have to work 25% more (50 weeks per year instead of 40 weeks per year)

Heh. I left teaching for three years to work in the private sector and the "25% more work" was honestly the best work life balance I've had in 20 years.

OK but I'm talking specifically about SFUSD, of which I have direct/indirect knowledge (from publicly-available information and talking with teachers employed there).

I know nothing about how your experience somewhere in Australia is similar or dissimilar. There may be differences in expectations, hours, compensation, job protections etc.

The social structure of a typical American public school is prisoners and prison guards. Why would any of the prisoners have respect for the prison guards? Most of the prisoners aspire to be something higher paid and less degrading than a prison guard. They aren't role models, they are just captors.

I suppose I was one of the lucky ones where my science teacher supported me in my interests/made learning fun

Is the fundamental model of a school from one of the post-historical paradises of education, like Japan, Singapore, or Finland much different? If anything, schools from those countries are stricter and instill more discipline than do schools in the USA. I know people who work in education; if they fail to be more than a prison guard, it's not through lack of trying. They're just hamstrung by administration every time they try to actually teach.

Perhaps a better analogy would be lab techs and lab rats. The students are the rats. They have a battery of experiments performed on them by lab techs (teachers) overseen by scientists (administration) who also determine which experiments to run. The problem is, what's being practiced is not really a science but a form of alchemy: how to determine a repeatable process to refine the lead of incoming children into the gold of model citizens who make all the right decisions according to the latest knowledge of what "right" is? (This was called "Outcome Based Education" in the 90s and "Social Emotional Learning" today.) The missing bit is that learning is an active process which requires involvement of the child.

But anyway, woe betide the lab tech who arrogates to perform the experiment in a manner not prescribed by their betters with Ph.D.s (say, in a way that's been shown to get results)!

> students that don't respect their teachers

Depends on where you teach. Some communities are much better than others.

I believe the majority of those reading this message have kids who respect their teachers (or if you had kids they would). You also live in an area where other kids respect their teachers if you have a choice (some of you don't, but if you have the choice and you will make that choice).

Near where I live there is an "inner city" school where the average teacher has been teaching for about 9 months - despite many teachers who have been there are 30 years. The typically teacher works just long enough to get some experience and then gets a job at a "suburban" school that pays less(!) but the students respect the teachers more.

> low-attention span of kids due to tech.

Kids have always been low attention span. Things are probably better than before because we have ADHD treatments that work that we just ignored in the past. Blaming tech is just the latest thing, but you can always find parents blaming low attention span on whatever the latest fad to blame it on is. Truth is kids are not really "designed" to sit in a classroom for hours every day - but it is still the best way we have to set them up for a modern life so we force it anyway.

I think it's a thing with regard to tik tok videos for example where they use split-screening (main content plays on top, and some car doing flips plays on the bottom)

Also when I was in school we didn't have smart phones, at least a slide phone with a full qwerty keyboard was the latest thing, can do a lot less than a full computer in your pocket/social media

Most of my kid's friends can't even get through a feature-length movie--even one made for kids. Every time there's a lull in the action for 10 seconds or more, the phones come out (or they start talking with each other, if no phones are present). If that happens more than 3-5 times, they're pretty much checked out and usually just leave the room. It's got to be constant stimulation and/or constant dialog to keep them interested. A slow, panning shot across some scenery? Forget it.

Yeah and a lot of the popular videos on YouTube it's a new scene/jump cut like every second or two

Tangent hard work is not appreciated too because a project that took months to work on is condensed to a 3 min video as if it happened overnight, not worth attempting when seeing how hard something is to do/discouarging

I feel like it'd be perfect if all these school replacements (home-school, charter school, online only schools) went after replacing the dearth of quality after school options.

Really the only options for after school activities after graduating from elementary school are competitive sports, competitive math, competitive music, competitive chess, etc. which are pretty much all zero sum in nature.

I'd love options for kids that let them gradually explore their interests to help them discover future vocational interests in a way that was beneficial to society such that they don't have an existential crisis when they hit senior year in high school and have to pick a college major.

  pretty much all zero sum in nature.

Doesn't every participant gain something from the practice and the competition, even if they wind up in last place?

I mean you just can't take a comment seriously when it lists music class as a zero sum competition. About the only time that's true is if you're dead set on first chair in a specific orchestra, or think a career any less successful than Mick Jagger's would be a failure.

Frankly I think that indicates the grandparent just didn't think very hard before leaving their comment.

Article from January-March.... how's progress been since?

Didn't go so well in Pennsylvaniaa

State rejects application for cyber charter school with AI teacher and two hours of daily class

https://penncapital-star.com/education/state-rejects-applica...

This is a good article. Dan Meyer, like his PhD supervisor Jo Boaler, is very much in the camp that traditional classroom education is good, and that improvements come by working through school districts, administrators and classroom teachers.

I'm not saying this to cast doubt on any of the facts in the article. Just pointing out that Dan, in general, has a less optimistic view of AI in education, than I'd expect of the median HN commenter.

That said, I'll share my thoughts on Alpha School, based on everything I've read (both things published by the school, and things I've read from parents online and in private forums):

- the '2x growth' in their marketing is way oversold; their typical 4th grader isn't doing math at the level of a typical 8th grader.[0]

- the '2 hours/day' in their marketing is oversold; students often work longer than that.

- only 25% of their students use Math Academy. The rest use IXL or ALEKS.

- in their charter school application, the amount they proposed charging for their software platform was unreasonable, given the minor role it plays in outcomes (10% according to Matt Bateman, who works there) [1]

- the core idea of their 'timeback' platform (that monitors student activity in realtime via video camera and screen recording) is good, but I have not seen it and have no idea whether it's real or how good it is

More of my thoughts from back in April: https://x.com/RahimNathwani/status/1912571014107787730

[0] https://x.com/RahimNathwani/status/1971804784475996469

https://x.com/RahimNathwani/status/1971817857286803873

[1] https://x.com/RahimNathwani/status/1912586493086036148

The intellectual pedigree actually explains everything. Jo Boaler is the one responsible for charging $5k/hr to advise schools to end middle school algebra. In an amusing confluence of concepts, Garry Tan (CEO of YC - whose site we're on) describes her as "infamous and disgraced"[0].

I don't know about that, but when I discovered that San Francisco schools weren't teaching algebra I was at first impressed that American children were doing Group Theory in 8th grade (something we only learn in the 12th standard in Tamil Nadu in India where I'm from) and figured moving that to 9th isn't a big deal only to find that they meant the basic stuff (linear equations and the like, what we learn in the 7th grade).

Honestly, I can't take anyone seriously who would try so hard to set back children from learning what is fairly basic Mathematics at that age. Children are capable of learning this. Or at least a sufficiently large amount are that we should be teaching them to a high standard.

For Alpha School, I think the Slate Star Codex review is likely more informative than this clearly polemic article.

0: https://x.com/garrytan/status/1953654484997169443

tl;dr This is from the people who want to delay Mathematics education to later in a child's life (algebra to 9th grade onwards)

Ironically, by taking the author of this article seriously, the preceding comment makes a more much persuasive critique.

I suspect most people will land on this where their predilections already lead them. To me, the effort to delay Mathematics education is sufficiently bad that I can dismiss the rest of that school of education without much concern. If they are right, they are right by accident and there's not much to learn there. I know there are others like me out there, and for them a quick reminder of who this person is will probably be sufficient for them to escape reading (what they will believe to be) a low-quality post. Time isn't infinite after all.

If you feel less convinced by this, it's simply that you're not in my audience. But I think it's probably worth sticking a tl;dr on the original. Let me do that.

I usually regret reading Dan Meyer's articles, but this one I really enjoyed. It pointed at a couple of sources that led me down a rabbit hole back when I encountered it.

I don't know whether Dan Meyer is in favour of delaying math education. I do know he favours delivering education in a school setting with in-person human teachers, but the latter doesn't imply the former.

And he works on making really nice tools for exploring math: https://www.desmos.com/

Desmos is really neat.

Reply on news.ycombinator.com