Technology Billionaires Trying to Remake America's Schools
nytimes.com
nytimes.com
To extend the analogy, I have this feeling that we're putting a great deal of effort into "taking the designer out of the design" rather than making tools to augment the designer's abilities/speed/workflow. Yes - we can probably create a system that can spit out company logos just as "good" as a mediocre designer, but if our best designers are still making do with knives, glue and paint then the entire field is being held back.
There are purpose-build tools for all kinds of professions, but teachers are largely making do with the same processes and tools that they did fifty years ago.
They don't need a LogosmartAI to replace them; they just need Photoshop.
Today on the HN homepage there was an article about how NJ was the innovation hub of American before the Valley, which generated a nice discussion. Akarnani recounted the story of Fred Terman's failed effort to make NJ more like the Valley. https://news.ycombinator.com/item?id=14497136
This historical anecdotes illustrates how even when many essential pieces are present (money, smart people, etc) it is a nearly Sisyphean task to change a system and a culture. The great management thinker W. Edwards Deming once said "a bad system will beat a good person every time" and that is the history of the past 60 years in education reform. I wrote about that in more detail here. https://medium.com/@m_pettyjohn/on-humans-hard-problems-and-...
This NYTimes article reaches out and touches nearly every broken aspect of education that makes improvement seem intractable. Zuckerberg wasted $100 million in Newark and seemingly came away learning the wrong lessons. Benioff, big hearted and charitable, has his heart in the right place, but long-term will have little impact.
With that in mind it is almost inconsequential what the Silicon Valley billionaires do in education.
There is massive potential in our teachers and students, but to unleash it we need to understand our history, take a systems view and approach, and genuinely understand the discipline of management, which even to the big successes in the Valley seems like an afterthought.
If you're outside (or even inside!) education, reading The Case of Mrs. Oublier, a study of one California teacher during the state's massive 80's reform effort, can be enlightening to the challenges faced.
https://393methods1.wikispaces.com/file/view/Cohen,+D.+K.+(1...
Next let's substitute computers for manipulatives (something you can model math with such as a place value block) for the educational treatment. In the PDF I linked above about Mrs. Oublier, she reports she is using all of the prescribed educational treatments in her classroom. From manipulatives to small groups. She's even excited about them. Calls it a revolution! Then the researchers visit her classroom and while all of the pieces are there, none of the understanding how to use the treatments is present. This happens systemically, and it's what I have seen happen with technology in schools.
In education we talk about how you teach, pedagogy, and what you teach, content as 'pedagogical content knowledge.' Whether it's blocks, small groups, smart-boards, or software it won't change the pedagogical content knowledge the teacher is bringing to the table. And in the great irony of our education system, our schools are not set up to learn (at least not effectively) so it's very hard to meaningfully move the pedagogical content quotient, especially throughout the system.
There is some nice software out there. Desmos comes to mind. https://teacher.desmos.com
But most software that makes its way into schools is not Desmos. It's usually something that merely replicates the content delivery method of teaching. Think Khan Academy. Other software is often just a digital way to practice facts or algorithms (my mind is in the math world for these examples).
I noticed an older comment of yours about expectations your first CS class had including how to navigate a nix system, at least one programming language, and Matlab. Let's pretend this class was instead algebra I where a student is expected to know the fundamentals of arithmetic including properties of operations (so you can use them to manipulate equations), that the equals sign indicates balance, exponentiation, strong proportional reasoning with a solid understanding of multiplication/division, and an understanding of integers. Combine this with a course where the teacher delivers a lesson each day, you do homework, and then have a test each week, never stopping to go back to things that may not have sunk in, let alone pausing to work on any holes in the things mentioned above that are critical to understanding algebra. I can introduce Desmos to help you model linear equations in a wonderful interface, but it isn't going to do a damn thing to increase outcomes if a) you are lacking in any of the requisite starting knowledge and b) don't go to extraordinary lengths to keep up with the pace of the course.
I like tech, but I don't think it's any sort of panacea for what ails our schools because it doesn't (and maybe can't) address the root causes.
Do I have a better plan? I do. Research backed.
Want a hint? Every child should have Music Instruction two times per week for at least one hour per session. That would do more than any stupid spreadsheet or Board Room review. I know, because Research backs it up. Not this..."venture" of sorts.
But it's not hubris. Silicon Valley Utopianism is just a sales pitch for people to insert themselves as privatizers and middle-men using technology.
I hate this attitude. Their overall vector is in the right direction; even if it's off slightly, at least they care and are doing something about it. You'll catch a lot more flies with honey than with vinegar... so perhaps you should try reaching out and volunteering your own resources (time & money) rather than just sounding off -- and not even linking to the research!
No, this is not right and the attitude dripping off your post isn't very constructive either.
These individuals have disproportionate leverage as a result of their wealth and available time. The likelihood is that they are going to make a bit of a mess with this leverage (you even say it yourself - "overall vector...off slightly"). So they can have a seriously deleterious effect on really very large numbers of children as a result of the power they wield.
A group of misguided parents in a PTA might might end up getting, say, Intelligent Design onto the science curriculum in their district; an opinionated billionaire could do rather more harm than that to very much larger numbers of students.
As for volunteering, many people do exactly that (and how do you know 6stringmerc doesn't?) by standing for school board elections, chairing the PTA, campaigning locally, giving talks in schools, and so on. The key point is that these are _local_ actions with local fallout. Conducting experiments at scale based on the idiosyncratic preferences of individuals who happen to have got lucky in business is a terrible, terrible thing to do to a city full of children.
I mean look at the proposals! They're classic examples of turning problems into nails when you're holding a hammer. Neflix recommendation algorithms to select lessons! Why? What the hell happened to pedagogy? Frankly, where's the humility? If Reed Hastings had got himself a masters in education and spent a few years teaching, even part time, there'd at least be some practical experience in the problem domain to guide his transformative efforts; otherwise it looks like dilettantism.
I have a hard time seeing how anything they decide to do to change schooling is going to make it worse than it is.
Our current education system is largely still in a factory style, where students are all taught the same material, run on the same idiotic schedule, are taught aimed at tests, and generally have any genuine interest in learn crushed out of them by the second grade.
Classes teach generic material and completely fail to demonstrate useful applications. Useful things like critical thought, philosophy, personal finance, cooking, computing systems/programming - not taught. Students learn to focus on rote memorization and local application of principles - see the articles the problems of getting students to solve conceptual physics problems vs. mathematical ones - and then proceed to forget most of the material as soon as there's no test on the horizon.
And we've known how to do better for a long time. Montessori's analysis of children and how they learn, her experimentation with the classroom and materials - this sort of thing is a hundred years old at this point, but we don't apply it widely. Instead we largely have the same old ineffective crap that we pass for "teaching". There's a reason Khan Academy became a thing.
We already know that schools do a bad job of preparing students for the current economic landscape. There's no reason to look at this and say "but some of these things might not work! Think of the children!" The children will be fine (or at a minimum, no worse off). We've already seen this where school districts have blown millions of dollars on new tech, to no useful effect. Zuckerberg threw a hundred million at some school district on the east coast, to zero benefit. But stuff like that isn't going to make things any worse. And on the flip side, they won't get better without action like this and without experimentation.
Better education is, imo, the best way to put this country on a strong footing for the coming decades. The government isn't going to do it. The private market isn't going to do it at scale. The average PTA isn't going to do it, and really, what the hell do most parents know about education, other than what they experienced? Committees and groups are the kind of thing that get bogged down in details, have no clear direction, and take forever to decide anything. At least a random billionaire can have the money and direction to get things done, even if not all of their ideas pan out.
One last thing - I feel these days schools are much less important than in the past, in so far as learning. The internet provides tremendous access to information; sites like Khan and other education material has improved over traditional textbooks by leaps and bounds; Stack Overflow and similar provide places to ask questions and find answers/explanations to specific problems or concepts; and so on. Fifty years ago, if a billionaire decided to bulldoze some schools and leave a bunch of children with no place to learn, that would have been a dire problem. Today, it's much less of an issue.
Otherwise I found your comment a bit contradictory. You don't like recent investments in classroom technology and say that technology isn't a solution, but you seem to think that technology is an at least partially adequate replacement for the classroom itself. Mark Zuckerberg wasted $100m in New Jersey (was it a waste? How do we know?) but seem happy for that pattern to continue.
To pick on a specific theme you touched on a few times: learning isn't even mostly about information. Ask anyone who bought a textbook in college then slept through lectures and never spoke to their professor. Kahn Academy doesn't represent a visionary, Montessori-compatible approach to learning math; instead it's a scaled up, gamified, alternative to an already stale math education pattern of rote memorization, depersonalized instruction, and practice problems. That's a part of math education - you have to learn how to apply techniques after all - but hardly represents the paradigm shift in math teaching that gets talked about here frequently. Don't get me wrong: Khan is a very good thing, especially for students struggling with bad teachers, but it's not the next step. And that's just talking about math education.
I don't agree at all that schools are less important. Schools are the only place where kids have mandatory exposure to specialists who's only job is to teach them things. To my point above, information is not education. 'Education' is the thing that happens when decent teachers with adequate materials and reasonable tools spend focused time with children. So you need time - which means smaller class sizes; you need tools, which really is a solved problem requiring only situational application of cash; you need materials, same; and you need decent teachers, which is a matter of training and status and autonomy and job satisfaction and vocation, and that is a really thorny mess. It's a generational political problem, not a thing a billionaire can meaningfully affect with a recommendation algorithm for syllabus modules.
So I guess my conclusion is - a really just absolutely epically well-funded lobbying foundation with a world-class policy board would do far more good than a series of expensive experiments playing with the margins of the existing borked system.
I've actually got a hard time pinning down all the work I've done in Education. A Degree in Curriculum Design. Substitute teacher for a couple years. Reviewer for Textbooks for Compliance with NCLB. Writing Tutor. Music Instructor. Public and Private Primary through Secondary Education product. Private and Public University product. Internationally travelled, bi-lingual and can pick up a couple other languages. I can often guess a person's Ethnic lineage based on their face, complexion, and last name. All of these pursuits come from a love of learning and turning around and sharing.
Localized Education is key. What works in San Diego, CA may not work well in Nashua, NH.
If there's one reason for a person like me to be pissy about the state of affairs in the United States, it's that I'm more qualified than Betsy DeVos to be the Secretary of Education. If that doesn't scare the crap out of smart people with the ability to change things, then we're straight up doomed.
Invest in Private Prisons.
I wouldn't draw any conclusions from this.
True, meaningful educational impact happens when you can deliver authentic instruction -- instruction that creates a really tight feedback loop between the learner, the instructor, and the subject. We have tried with technology to mimick that at scale. Khan Academy and online quizzes is not a new idea; B.F. Skinner had a device doing this since at least the 60s. It works to stuff CEOs' pockets with money, feed the ever-larger administrative behemoth that schools are becoming, etc. Results, not so much.
There is no greater evidence than the fact that schools are probably the only sector which has literally gotten less efficient over the past decades; we literally spend more money per pupil for worse outcomes. The incentives in public education are absolutely abysmal, and we've seemingly done everything in the book to strip the voice of students and families in this sector.
As schools, our feedback loop between our own quality is practically nonexistent. No wonder we don't do right by these kids. Lots of money to be made selling software, though.
Do as I say, not as I do.
> it can't scale without an...investment that, so far, society has been unwilling to make
Without making any judgment on what is proper or appropriate education funding, tax policy, or anything political like that, the GP is right. Getting tech out of the classroom 1) is not a 100% positive thing; 2) costs more than society, to date, has been willing to spend.
> Are we trying to make education cheaper, or better?
It's not a binary proposition and pretending it is is dishonest. I want to see the median increase. I want to do it cost-effectively. If the choice is raising the median 5 points by utilizing technology for a marginal change in overall cost, or raising the median 20 points by quintupling education funding, one is possible and one is not.
All humans have finite amount of time during the day, finite resources and finite lifespans. All of which is allocated pretty tightly. If you want to add value to the lives of people, your best bet is something that doesn't come at a great cost to them. Hence scale.
A secondary benefit is that big things which don't scale are only attempted by few players with lots of resources; making things more scalable means more smaller players can compete on the market.
2. How about we reduce the number of teachers (and have the remaining be paid fairly) by making education scale? See also my agriculture example.
But they're consuming money and attention that could be redirected to solving real problems. I know, this is a perpetual complaint with someone in any crowd.
How about we reduce the number of teachers (and have the remaining be paid fairly) by making education scale?
Unless you have an innovation in mind that I haven't seen, I fall back to the copiously documented truism that you can't use technology to solve a people problem.
See also my agriculture example.
I don't concede that children can be educated within the same concepts that are used to harvest cows and cotton.
Thing is, I don't buy that truism. IMO technology is sometimes the best way to solve a people problem, because it sidesteps all the issues of politics and ego.
> I don't concede that children can be educated within the same concepts that are used to harvest cows and cotton.
Why not? Just like modern agriculture makes a farmer more productive per unit of land, we should strive to make teachers be able to teach more children and teach them better.
AKA the reasons why it's a people problem.
we should strive to make teachers be able to teach more children and teach them better
You're comparing apples to oranges. Agriculture does not "strive."
All in all this sounds like pie-in-the-sky thinking. The science I've seen says larger class sizes do not result in better outcomes, do you have anything that contradicts that? Anything demonstrating even a clear benefit from technology in the classroom? Anything about scaling? That is, is what you're proposing anything more than imaginary?
I mean, call me an out-of-touch plutocrat.
LAUSD's own numbers for 2016-17 indicate a maximum of $88K, and that's with a doctorate and maximum seniority.
http://achieve.lausd.net/cms/lib08/CA01000043/Centricity/Dom...
*Edit: Actually it was 2015 data. Also note that less than 1/3 of LAUSD's employees are teachers.
Also, it's about cost effectiveness, not just cost. The more parts of education that are scaled, the more money can be spent on other programs that can enrich kids' lives, which we currently don't have the money for in some cases (e.g. music programs, special ed, special programs for the gifted, actual educational research).
Maybe because, specifically in the U.S., there are 73.6 million [0] children aged 0-17. 45.7 million (or 50.4 million [1]) of which are school age. If that isn't a problem of scale I have absolutely no clue what is. You can't give every single one of them the kind of attention and resources a super rich kid gets. It's just not feasible.
We as a country made the decision a long time ago that every child deserves an education. I'm sure there's a few libertarians out there who'd disagree. But, I don't think that's a controversial opinion these days. I'd go as far as saying that the opposite opinion, that not every child deserves an education, is borderline taboo. It's not something I would ever suggest in polite company unless I'm intentionally trying to make people think I'm a huge jerk.
Given that we want all children to be educated, that there are limited resources to do so, and the number of school age children is enormous we have no choice but to think of education as a problem of scale. Currently we spend $620 billion a year on public education [2]. That number is sure to increase, but never to the point where every kid has all of the hands on teaching that would be preferred. This means system-wide improvements are likely going to have to come from somewhere else; namely better methods of teaching which can be applied nearly across the board. Is this the optimal situation? No. But, when you consider that the natural state of things is no education at all, perhaps we shouldn't reflexively sneer at attempts to make an imperfect system marginally better for everyone in the long run.
0: https://www.childstats.gov/americaschildren/tables/pop1.asp
Possibly, if we could actually pay all public school teachers decent salaries.
Which students deserve to be taught by the bottom 25% of teachers? Also if we wanted to move the mean skill of these teachers higher what would it take? 4 years? 10?
I think software provides a faster benefit. And one that can hopefully have an outsized benefit to students who have the bottom 25% of instructors. Plus faster software iteration hopefully lets us find more optimal solutions faster for groups of students.
This gives us some breathing room for fixing the long tail instructor issues.
Yes, there is a wide spectrum of competency/skill amongst all the practitioners of a profession; that's not a remotely novel insight. That's not an argument for underpaying them.
> Which students deserve to be taught by the bottom 25% of teachers? Also if we wanted to move the mean skill of these teachers higher what would it take? 4 years? 10? I think software provides a faster benefit.
Which students deserve to get taught by the bottom 25% of shitty, barely-tested software chugging algorithms designed on specious theories? Not your kids, I'd bet.
I have no way of knowing if this i-Ready is the bottom 25% or the top 5% of barely tested software chugging algorithms. It's the only choice I have anyways, so it's hardly worth educating myself on the matter. Luckily my daughter is in the gifted program, so I at least have some assurance that her teacher isn't total crap. And beyond that, being gifted gives her a decent chance of overcoming a potentially shit education.
Well the politics are not there currently to support teacher pay raises in many states. Its shitty but that is how it is. If at some point we get a political culture able to reverse that, great!
>Which students deserve to get taught by the bottom 25% of shitty, barely-tested software chugging algorithms designed on specious theories? Not your kids, I'd bet.
>My daughter is already a guinea pig for this stuff. Her public school uses i-Ready. http://www.curriculumassociates.com/products/iready/diagnost....
Why do we need to entrust these technology solutions to schools? They are pretty ill equipped to evaluate them except for large districts. The school should just be a data source. The actual software should be parent choice, much like we have school choice in many states. If one piece of software gets a bad reputation the parents can choose to use different software. (Much like paying teachers more, this may not be possible in todays political landscape)
https://www.reddit.com/r/bayarea/comments/6aadft/math_teache...
The money quote (for me): “That sounds like a low bar [that their apps did not harm students’ educational results],” Ms. Woolley-Wilson said. “But with the history of education technology, it is not.”
That is to say, you can throw all the money and toys and methods you want at it. What you need, is people actually paying attention.
And paid to pay attention. So that they are not overwhelmed with workload nor how they are going to pay their own bills and take care of their own families.
In other words, invest in the people providing the education. Choose them wisely, and empower them. The rest is just various forms of b.s.
This is literally the first time I've read an article with a line similar to:
“They have the power to change policy, but no corresponding check on that power,” said Megan Tompkins-Stange,
“We should be asking a lot more questions about who is behind the curtain,” Ms. Davis said.
I've seen such outrage as "He only gave away a $1,000,000 so that he could cut $350,000 off of his taxes!" :/ "The deal eventually, indirectly feeds some money back into the project in the long term making it self-sustaining!" :/ "It not anonymous, so he's only helping thousands of people to be famous for helping thousands of people!" :/ "He's rich, so giving away huge amounts of money is easy for him. Helping thousands of people should not be recognized unless it was hard!" :/ Then there's the classic "So, he's helping this part of the problem. But, why doesn't he help the rest of the overwhelmingly huge collection of tangentially-related problems?" :/
Sometimes it seems that the standard we hold people to is "Helping people must be anonymously self-immolating to avoid disparagement." Then we disparage the same people because, as far as we can tell, they are not helping enough.
IMO, the conclusions people jump to mostly tell about people doing the jumping.
Also, https://blog.jaibot.com/the-copenhagen-interpretation-of-eth....
I'm speaking from observations made sometimes on Hacker News and frequently on Reddit.
This is the opposite of corporate social responsibility, yall. Do your freakin job, Benioff!
My beliefs are somewhat extreme in this topic, but I will share anyway. In broad strokes:
1. One of the biggest problems is that education has come to be ruled by the administrators. They are not servant leaders, rather they are aristocrats. Their pawns are students, faculty, parents, and taxpayers. These stakeholders need more influence in the decision-making process.
2. Mandatory schooling (both explicit and implicit) could easily stop at 8th grade with the goal of basic numeracy and literacy as well as fundamental learning skills. These skills will provide enough education for the graduates to function in society as well as move forward in the education system if/when they choose to do so. Currently, most of high school (and middle school, frankly) is just very expensive and inefficient baby sitting.
3. That said, secondary, tertiary, and trade-based education should be widely available, of high quality, and low price -- probably subsidized heavily by the state -- throughout a person's life. Many people do not realize the value of education and learning until they are older, and our current system does a relatively poor job accommodating these people.
4. Overall curriculum design needs to focus on autonomous learning (touched on in the article). Note that this means that the learner takes responsibility for their learning, not someone else (e.g., teachers). Also note that this does not obviate teachers, rather it just makes them one of many sources of knowledge/learning. Anecdotally, I have seen this type of system pull off what many would consider to be miracles.
5. Learning in general should be much more contextualized and (when practicable) experiential. Many/most students do not see how the current curriculum content applies to them. The current content seems to be related to, and largely is a product of, the type of education upper middle class people want. Folks from other walks of life actually have a lot of curiosity and a high capacity to learn, but taking part in the current curriculum would turn them into an alien within their community (e.g., non-native English speaker being beaten/abused for speaking English well). Few people are willing to do this.
6. Assessment should lean much more heavily on portfolios than on standardized tests. Don't get me wrong, standardized tests have their place, but currently they are overused and are often times misused.
This is an issue I very passionate about. Please let me know via a response here if you would like to discuss this more.