Great observations. I've been in the edu (coding school) space for over a decade and it tracks with my experience (though it seems you've tackled a lot more market segments than I have). I'm very bullish on AI + Education. If Karpathy's announcement doesn't track, what do you think might work instead?
That's a good question to which, unfortunately, I don't have a good answer. None of my startups managed to make our edu-focused products long-term successful. Fortunately, we designed our major technology stacks so each could result in products in at least two different markets. While our edu products never did better than break-even for us, our non-edu products usually did much better with several becoming breakout hits. Personally, I'm very passionate about the need for innovation in education, likely due to challenges I had with traditional school. But after spending many years and substantial funds, I eventually stopped trying to do edu-focused spins of my products (despite clear potential of the underlying tech). Although I now understand many of the reasons the edu market is so hard and counter-intuitive, I also learned succeeding there requires vastly different skills and interests than I have (or want to have). K-12 was especially frustrating for me because success there is so untethered from actual product effectiveness, along with being unbelievably bureaucratic, largely opaque, and having a glacially slow sales cycle. As someone who started out so idealistic about making education better, actually trying to build a successful tech startup selling to K-12 can be soul-crushing.
I agree that in terms of increasing actual learning, AI could be revolutionary. I commented here because my experience was specifically being a startup trying to sell products based on revolutionary new disruptive technologies into edu.
I'm B2C and selling ed to students is an eye-opening experience. It's actually very easy to get people to pay, so the problem isn't conversions. The problem for me is that if you keep following the money, it's natural to become predatory because the most vulnerable need the most education. The more vulnerable, the more likely they'll buy hype. You're essentially selling to the most needy, which feels like payday loans and that sort of business. I don't have it in me to deploy the sales/marketing playbook on vulnerable audiences.
So, the advice of selling to businesses seems better, but I also hear you on selling to K-12 or higher ed as having all the negatives of enterprise sales without the lucrative upside.
So then we're left to selling ed to businesses. But I feel despite all the lip service paid to training and upskilling, corporations don't *actually* want to upskill their workforce. Or at least not at the sacrifice of productivity.
Education is very unnatural, but that's what makes it a worthwhile problem to solve, too.
There's a difference between the educational-industrial complex (or the ed landscape in terms of market), and motivated learner-centric models which can be done independently.
As much as degrees and certs are important in the current paradigm, it's long been the case in tech that credentials matter less than output, skills, and soft skills.
As much as I personally believe in a stronger credential system, the shift is toward individual learning. The popularity of coding bootcamps in the tech space I think demonstrates that.
Personal AI tutor? Without the pesky SME credentialing? Sign me up!
IMO, a lot of that leads to sink or swim. Not that elite institutions are perfect (by a long shot). But largely cater to kids (and families) that can do things on their own with help from qualified people and you almost certainly get better results--and lower costs--for them.
I'm the founder of Launch School and I have been thinking about ISAs, coding bootcamps, and software ed for over a decade. I've tried to insert my thoughts here and there, for example:
Thanks for replying! The reviews of Launch School are impressive.
I think you're dead on with that first link. I've heard of some schools that did not apply an effective filter up front, despite instructor pleas, and suffered for it greatly with lowered job placement, ultimately failing with the ISA model.
I feel there's a class of students for whom ISAs are perfect, and, just like you said, a class of students for whom they are ineffective. The ones where the ISA works and standard school is out of reach have incredible success stories.
I found something similar when I taught a nearly free C++ class. I put a really simple test on the front with a refundable $20 fee (if you took the course). Everyone who ultimately took the class was really motivated. $20 was all it took to filter. $5 might have even worked since I speculate the effect is psychological, not economical.
The second link is good food for thought, some stuff I hadn't considered. If the ISAs are sold cheaply enough, you really don't need much student success to get a return.
IIRC, in Lambda's case, the ISA sales were stopped while the company was still young (having gotten another round) [caveat: I didn't have much visibility into this side of things], and anecdotally I think they had some of their biggest successes early on, but I agree with your points on this.
One thing that really impressed me about Lambda was how diverse in every respect the student body was, people from just every walk of life. Waaay more so than I'd seen at any university. I credit the ISA for making this possible.
One of my missions in life is to enable people to get the training they want to get. ISAs were wonderful in that regard.
I think you have to be careful about looking at the diverse student body as "giving them an opportunity". For example, predatory for-profit universities specifically target low-income minority populations not to serve them but because they make for easy prey[1]. I'm not saying Lambda or any particular bootcamp is in this category, just that having a diverse student body isn't a feature unless you know for a fact that you're helping 100% of the students.[2]
Education has a couple of unique attributes that makes it difficult to assess:
1. the value cannot be perceived until much later after the service is rendered
2. alternatives are mutually exclusive; eg, people usually just attend one university or coding bootcamp, not all of them (or even two of them)
This is in contrast to, say, a restaurant where diners can immediately determine if they like the food and can compare it with competitors (because they dine at all the restaurants).
Those two attributes make it very easy to lead with hype and marketing and vulnerable people are particularly susceptible to it. imo it's not appropriate to deploy the standard startup playbook in edu, especially if you find yourself attracting vulnerable students.
Ok, now combine that with ISAs, which has some positive qualities but are not as incentive aligning as marketed. And then if you sell the ISA in bundles, then it becomes even less incentive aligning. Well, I guess it's now aligning with investors and loan brokers. But it's certainly not aligning more with student outcomes.
It all makes for a very delicate situation where you have vulnerable students biting on the ISA bait.
I know it seems like I'm just complaining but I've been thinking about these problems for a long time and I come with solutions. Or, particularly, a solution: imo the best thing an edu institution can do is allow students to leave easily.
Why do we never think of restaurants as predatory? The idea is ludicrous to even consider. There are of course terrible restaurants but we can just not dine there again and eat elsewhere. There are lots of restaurants around.
And here's the issue: there are also lots of edu institutions around but every single one of them deploy the marketing->entrapment playbook.[3]
How does marketing work when you have a bad school? Because of the time-lapse between value received and service rendered. The ISA is easily abused in this environment because it's both the marketing and the trap.
Anyway, I'm just riffing here... btw, huge fan of your books and work! Despite what I wrote above, I knew Lambda was making an honest attempt at their curriculum when they hired you.
[1] https://failstatemovie.com/
[2] imo, edu institutions need to be judged on how they treat their most vulnerable students, unlike startups who are judged on their best wins (https://twitter.com/cglee/status/1781129096250179640)
[3] https://medium.com/launch-school/educational-entrapment-f5cc0472051e
I agree with you on the dangers of targeting minorities. My thinking is still, though, that there is a category of people well-served by ISAs, and those people are underrepresented in schools with traditional payment models.
I've often thought that anyone can be a dev--if they want it. Meaning, it has to be someone who likes the material enough to put in the effort. It doesn't matter if you're smart enough; it only matters if you're going to put in the time. There's a reason I'm not a CPA. I'm absolutely smart enough, but eff that!
And lots of times schools do advertise "we guarantee you a position in a high paying job if you just put your ass in that seat for 6 months". Who wouldn't want that? But they leave off the "you gotta want it" part. And then people get trapped.
Related to "easy to leave", we fought to filter the front-end heavily for people who "wanted it". And we fought to allow students to attend for as long as possible with no obligation. The goal was to allow them to discover if they wanted it. However, this was not realized while I was there. So much more I could say here about how that didn't happen, but I'd wager you have a pretty good guess.
Hand-in-hand with wanting it (IMHO) is having a comprehensive, heavy-hitting curriculum. The guy who hired me left Lambda ages ago, but he and I came up with a list of things people should learn to be decent devs... and that was one helluva list. Needless to say, not all of it got covered, and as time went on, less and less of it did.
I love that California community colleges are now free. Easy to leave!
I really appreciate the conversation, btw. I also like geeking out about this topic, and it sounds like I could learn a lot from the ground you've already covered. I'll check out the reading after I finish prepping for class next week. (0-1 Knapsack and dynamic programming. Phew.)
We have a saying at Launch School: habits over enthusiasm. I've noticed that wanting it isn't enough. Desire is fine but what the real key is a commitment to studying. I think we're saying the same thing with different phrases.
Agree that ISAs can be useful. But it's a tool that can be used to harm or help. Elevating it beyond a financial tool to some sort of educational breakthrough was a disservice, imo.
We agree on the value of community colleges. I wish they were better funded and more people worked to drive CC graduates to six-figure jobs.
I am the founder of BlueWave Labs - we accept junior developers and UI/UX designers to give them Canadian experience by building open source products, to help them land their full time IT jobs and would love to talk to anyone who has put some thought on bootcamps. BWL is a bit different from bootcamps but the main idea is the same: to help people land full time jobs.
I've spent a lot of time thinking about this topic and I've come to the conclusion that the best generic advice for folks in your situation is to be under-employed, ie a job where you're not maxed out mentally/emotionally/physically, and save your resources to learn on your own.
Entry-level technical jobs are some of the worst jobs to help you ramp up as a programmer. It's chaotic, demanding, emotionally taxing, and a lot of blind leading the blind. And most importantly, employers expect you to perform "just in time" learning, which is detrimental long-term if you don't yet have strong foundations.
I agree. I don't think anyone can get a job not knowing anything.
It's best to keep up your energy and learn on your own time and interest.
Pick projects that you're interested in. A website, a small little todo app, a small game. These introduce fundamentals in a setting you enjoy. You'll do a ton of things wrong, know you did them wrong, and the recognition of that provides all the next steps you need.
Five to ten projects later, you'll be ready to start taking some small jobs. Open source tickets. Charity work.
Do this a little bit longer and you'll be set.
It's fundamentally about doing projects you enjoy though. You need to take interest and defend that interest long enough to overcome the stresses and the road bumps.
I'm embarrassed to say that even after spending nearly a decade operating a mastery-based software engineering school, I've only just heard about this paper now. And having read it just now, I'm feeling better than ever about mastery-based education.
Interesting observation and aligns with my experience of really enjoying small focused tools and apps. This website is a good example.
Further, it feels like there's a corollary here to companies, where financially constrained companies who are smaller and more focused provide better customer experience than cash-flush competitors.
For me it's not shocking that any one person can be this cruel. What's difficult to process is how entire social structures and institutions can be this cruel.
But how was this perpetrated broadly by multiple societies?
I try to remind myself that the part of humans that makes this possible is in all of us. I don’t think it’s anything special – there needs to be constant vigilance to ensure we don’t allow things like this over and over.
It could be illogical, but it seems true to me if this can occur for generations at such broad scale. It’s a feature of humanity, and to an unknown degree even me as a result.
Societies are commonly cruel without knowing it. Even modern Western society which ostensibly tries to minimise cruelty depends heavily on animal cruelty and horrific work conditions in the third world. Individual people being truly horrible is rarer
This is a smart comment because this is exactly what we run into (I operate a coding school that uses ISAs). Employers love our graduates, but if we charged employers, we'd have to play a cat and mouse game of "please don't contact our students directly" or "please don't apply to their website and go through me". By charging students, we can be at ease with making introductions or promoting employer/student activities without having to police their communication.
That said, I believe there's a lot more we can do to help ease the financial burden from students and I love the direction of trying to move the cost of education from students to employers. I'm working on that.
> This is a smart comment because this is exactly what we run into (I operate a coding school that uses ISAs). Employers love our graduates, but if we charged employers, we'd have to play a cat and mouse game of "please don't contact our students directly" or "please don't apply to their website and go through me". By charging students, we can be at ease with making introductions or promoting employer/student activities without having to police their communication.
Have you actually tried it? I'd be pretty surprised if the employers attempted to cut you out like that. They're already accustomed to paying recruiters, and the access / promotion that you can provide them is worth the money. You could also sue them, whereas if you try to sue your students for repayment of their ISAs you run a high risk of bad PR.
Since they're on the market looking for jobs, our graduates are pretty easy to find so any company can just contact them directly without partnering with us. There's no real way to prevent that. And, I don't want to prevent companies from reaching out to our graduates; I actually want the opposite.
Further, grads can (and should) apply to companies that aren't partnered with us. I don't want them to only apply to a limited set of partnered employers.
We've been operating for about a decade, so we've tried basically everything once :) Our current corporate partnership program is here if you're interested: https://launchschool.com/employers/placement
For the past 6 years, I've operated a software engineering school[1] that uses ISAs and I have some thoughts about them. There are two main stated benefits to using ISAs (there are others, but those seem overblown):
a) commission based pricing (aka, incentive alignment pricing)
b) deferred payment
Of the two, imo the second is by far the most important thing for students. To the first bullet, I don't personally find commission based pricing to be all that incentive aligning. For example, it's not uncommon for me to advise someone to take a much lower offer because it seemed like a better long-term opportunity. This is in line with Sean's observation that quality education outcomes is difficult to reduce to salary numbers alone.
To the second bullet, the major problem of deferring all payments, however, is that you attract a lot of people looking for a shortcut. This is exactly the opposite attribute top employers are looking for. This is the "adverse selection problem" Sean mentioned.
Ultimately, the solution here is in selecting for the right type of students into the ISA-based program. Sean mentions that credit scores track with the type of students they're looking for. Other ISA-based programs have stated that they've found a secret sauce other than credit scores for detecting the right students.
We've found a different selection criteria:
We ask students to do a lot of work before we engage them with an ISA. I'm calling this model the ISA-later model, just so we can contrast this with an ISA-first model, which is what Sean and everyone else is doing.
An ISA-later program solves nearly all the problems associated with an ISA-first approach:
- adverse selection is mitigated since you have a long track record of student behavior and performance
- can still be egalitarian, without relying on credit scores or degrees or any socioeconomic markers
- still possible to defer all payments, without the lock-in of an ISA-frst approach
There are many other student-friendly benefits of an ISA-later model, but I'll stop here as this comment is getting long.
> Once students graduate from Capstone, they are expected to spend 40 hours a week Monday through Friday searching for a software engineering career. Weekly check ins are mandatory.
I don't think I could handle a 9-5 job, let alone 8 hours a day applying to jobs under penalty of unspecified legal consequences.
It's rarely 8 hours a day "applying to jobs". This is similar to an employer saying "you should work 40 hours/week". In that 40 hours are lots of varied activity, including breaks and lunch and any other activity that will allow you to succeed.
The idea is that we want to set expectation that people should focus on searching for a job and not, say, go on vacation. The wording here also is far more severe than reality only because we want serious participants.
Huh—I did this program (I didn’t know the founder posted here, just browsing) and I had no idea weekly check-ins were mandatory. For serious (but not fanatic) participants it is, of course, a no-brainer to check in at least that often. Just found that line interesting and wanted to share.
Hey Adam! This is the weekly check-ins during your job hunt we did. I guess your job hunt duration was too short, so maybe you didn't do too much of this :)
This is something that mastery-based learning solves pretty cleanly. Or more accurately, in a MBL pedagogy, there's no need for faster/gifted or slower/remedial paths.
All the debate for/against gifted programs fall into the same trap of not thinking critically enough about the pedagogical model that powers pretty much all of our education from childhood through higher ed. IMO, we're only debating about the symptoms of a outdated pedagogical model, and not talking enough about the underlying problem -- the factory model of education.
I'm not sure what you mean when you say it's a trap. The debate is "current model + gifted programs" vs. "current model with no gifted programs", not abstract pedagogy - NYC is currently executing on that second option, and multiple other school systems have announced similar plans. I completely agree that gifted programs are a band-aid for a poorly designed school system, and in an ideal world special interventions wouldn't be required to let students learn at a faster pace, but that doesn't mean the debate on whether we should or shouldn't have the band-aid is unimportant.
