Based Camp： Our Attempt to Fix Education
Episode Stats
Words per Minute
205.64442
Summary
In this episode, we focus on the new educational system that we've designed for the Collins Institute, and how it aims to revolutionize the education system in the United States. It's a system that uses a skill tree, or tech tree, to help students learn and improve their skills, and rewards students who create content that helps them improve their grades.
Transcript
00:00:00.000
hello gorgeous hello so this is a follow-up episode we had done an episode on the education
00:00:07.340
system why it's terrible the forces at play here how how little genuine innovation is happening in
00:00:13.880
the space it is terrifying i suggest you check that episode out but you really don't need to
00:00:18.780
watch it before this one this one we're going to focus on the system that we've designed
00:00:23.120
for the collins institute you can learn more at collinsinstitute.org and yes just one of those
00:00:28.380
spaces where we've decided to put some of our effort to try to fix something because we saw
00:00:32.240
something broken and that means we got to clean it up the the fun premise here is this is basically
00:00:39.480
we discovered that in the world of education no one is really trying anything innovative also when
00:00:46.060
people have tried to experiment around different outcomes it did it wasn't actually trying to
00:00:50.960
experiment for earning potential or changing the world or making an impact or even thriving as a
00:00:55.720
it was pretty much around self-esteem and in the few isolated instances in which people did say
00:01:02.040
i'm gonna try to create geniuses they did so basically we have reason to believe that people
00:01:08.000
who come in intentionally trying to innovate new models have the potential to genuinely change the
00:01:13.940
world because there's nothing more fundamental than the way that you educate the next generation
00:01:17.760
and equip the next generation to take on new challenges so the premise here is okay we see this and
00:01:23.020
we're like all right we're gonna throw our hats in the ring and this is what we have decided to do
00:01:28.660
few caveats to what you just said montessori is pretty good for pre-secondary school so before
00:01:35.580
middle school montessori is actually a pretty solid system i think it can be improved a little but it's
00:01:40.240
it's decent for what's out there um and acton schools are actually a pretty innovative and
00:01:47.040
interesting model that we respect outside of that pretty much hate everything would you like to know more
00:01:53.020
okay now to to our system so what we've done is we have taken the entire educational system
00:02:00.380
secondary school so middle school and high school as well as we're eventually moving to colleges we've
00:02:05.560
talked with this new college that might actually implement our system as the primary way that
00:02:09.760
learning happens at that college so hopefully we can go from middle school to the end of college
00:02:12.860
where we have divided it into individual nodes that work like a skill tree or a tech tree in a video game
00:02:20.460
if you're not familiar with what that looks like i doubt that much of our audience doesn't know what a
00:02:26.060
skill tree or tech tree in a video game is but it's actually a pretty hard concept to explain to
00:02:29.020
someone who's never seen it think of it like a ancestry tree that you progress through like you
00:02:34.680
would have nodes you complete one node it unlocks nodes above that sometimes you need to unlock two
00:02:38.920
nodes to unlock one node above that okay and whenever a student clicks on one of these nodes
00:02:44.760
they get a place where they can book a time to take a test on that subject to complete the node
00:02:50.720
under that it's like you would have in hacker news although someone always says i i say reddit as well
00:02:55.460
but hacker news is a better direct analog because it's all links where you have links to all of the
00:03:01.780
places a student could learn that information that exists online and students can upload their own
00:03:08.120
youtube tutorials after doing something or their own notes after doing something then students vote on the
00:03:14.340
sources that were most useful to them those votes are modified by how well a student did eventually
00:03:19.560
we want to begin to build profiles of students so that students can be shown resources that were
00:03:25.080
most useful to students who have similar voting patterns to them on these these tests and we also
00:03:31.120
want to build some sort of reward or remuneration system where students can receive some sort of reward
00:03:36.380
for creating content that ended up being useful to a lot of other students now a quick side note on this
00:03:42.520
one of the things that we really believe in with our educational system is to not have a single metric
00:03:47.580
of measurement and by that what i mean is the existing education system 80 of students are not in
00:03:53.120
the top 20 and that's a problem when nobody wants to hire somebody who's not in the top 20 or no one
00:03:58.040
wants to accept someone who's not in the top 20 so when you are measuring students on how far they've
00:04:03.200
gotten in the tree how good their grades are how many students liked the content that they were creating
00:04:09.380
and we're using that these are metrics that can have different students succeed within them
00:04:13.900
which means that no longer are you dealing with a student where 80 of students are not in the top
00:04:18.360
20 okay yes so now back to the tech tree one of the major problems in the educational space is what
00:04:27.460
we call the extrinsic intrinsic reward problem this is a phenomenon where if you pay someone to do
00:04:33.760
something that they love they start liking it less and less and less this is why when people take a job
00:04:39.380
one of their hobbies they will sometimes grow to really resent that hobby and the school system
00:04:44.180
is right with this problem because that is what grades are they're an extrinsic reward and so when
00:04:48.580
a student has a genuine passion for something that passion can be dulled by continually applying this
00:04:53.420
extrinsic reward this is why kids who go through unschooling this is a schooling system where different
00:04:58.940
from homeschooling it just lets kids do whatever they want all day these kids often end up doing just
00:05:05.440
astoundingly at the things that they love but then unfortunately really bad at the things that
00:05:10.560
they're disinterested in and this is the the core problem of the unschooling movement and it was a
00:05:16.040
big inspiration for us when we were designing our school system we said can we design a school system
00:05:20.960
that instead of trying to fix the problems with public school instead uses students in a control
00:05:26.100
scenario students who are just doing whatever they want all day and see where those students end up
00:05:30.820
struggling and fix those areas and only those areas while allowing students to do basically
00:05:37.660
whatever they want so they often end up with crippling test anxiety so that is why we incorporate
00:05:42.520
tests throughout our entire system because the truth is that the real world will test you it will put you
00:05:47.400
in testing situations if you never go through any of those that's that's a problem because that's
00:05:51.440
functionally in the real world i mean we're optimizing our system around outcomes so that makes
00:05:55.560
sense second is the extrinsic intrinsic reward problem so what you need is a system that can
00:06:00.800
organically detect when a student doesn't have intrinsic motivation within a subject and then
00:06:07.080
imply that extrinsic motivation in and only in those scenarios and proportional to their lack of
00:06:14.760
intrinsic motivation in that space now in project-based systems which is what a lot of people move to but
00:06:20.020
are unsustainable you have teachers that watch the students and push them where they need to be pushed but
00:06:24.460
you can't do that at scale so it's not really useful for a scalable system unless you're doing it with
00:06:29.480
ai which we do implement in our system and might be a counter system to ours that we can talk about
00:06:35.180
so what we do is the further behind a student is in a part of the tech tree the higher a multiple is
00:06:42.820
applied to the credits they get for the score they get for completing a test in that subject and they
00:06:48.520
need to get a certain number of credits per semester to stay within our school system and this is the
00:06:53.940
thing where people are like oh that's really harsh what you kick kids out just for not getting a
00:06:57.060
certain number of credits even though this is like a public system that's meant for everyone
00:06:59.820
in the adult world i mean theoretically this is the way it's supposed to work when someone can't
00:07:04.100
handle their shit what what do we do with these people we send them to prison that's that's what
00:07:08.700
we do it's it's it's not the way prison actually ends up working prison system also has a lot of
00:07:13.460
flaws i'd love to work on that someday when i get time and simone has some really interesting
00:07:17.200
theories around that i want to do yeah our charter cities one of you if we ever get around to
00:07:21.260
publishing that we recorded a while ago but anyway what the dynamic that we hope can eventually
00:07:27.220
evolve out of this is that our school system becomes available to any students who want it
00:07:32.800
but the students who lack the ability to find a way to motivate themselves they end up having to go
00:07:37.280
to the existing public school system because that's basically prison for kids and that's what it
00:07:41.440
already is it's using the same contractors it's the same system it's it's it's terrible it's terrible
00:07:47.440
anyway so that's the way that aspect of the system works now what do we do when a student
00:07:51.160
gets really far ahead in a subject three four years ahead of where they are well if somebody
00:07:55.320
really loves playing with legos the best reward is more legos by that i mean it's the best yeah
00:08:00.260
reward for somebody who's intrinsically motivated for something is more of that thing at a more
00:08:04.880
advanced level but we take it beyond that and this this is the my personally favorite part about
00:08:09.980
this system is that rather than orient people around just pure academic achievement or outperforming
00:08:15.820
students in the mainstream industrial schooling world we're really oriented around instead of
00:08:20.800
outcome being get into good college or get scholarship or get good grades or just graduate from high
00:08:26.180
school our outcome our desired outcome is self-sufficiency upon graduation so this could mean that you can
00:08:32.560
enter academia basically already with a grant-based like postdoc kind of situation or with a scholarship
00:08:38.500
or you graduate with enough income from work consulting or business you create or even a non-profit
00:08:45.700
to support yourself fully as an independent living on your own paying rent adult and that means that
00:08:52.640
one of the rewards that we give to people when they start excelling in some domain is basically actual
00:09:00.100
real world work in that domain and it doesn't matter how old you are to start doing this i mean
00:09:05.280
benjamin franklin had a full-time job and was contributing under a pen name to a newspaper
00:09:09.820
in the early colonies when he was 12 years old so this is about competence not age and we really hate
00:09:15.940
how students are infantilized these days so it's really about starting your career so instead of
00:09:20.660
let's say you're super into mushrooms you're like oh my gosh mycology i love this and you get really
00:09:24.960
really advanced well then ultimately what we start rewarding you with is okay hey we're gonna help you
00:09:29.420
pitch a peer-reviewed paper maybe some research in some weird like jungle with some weird mushroom that
00:09:34.600
only five people know about we're gonna have you pitch this to them pitch a peer-reviewed research
00:09:39.440
paper and hey maybe they'll say yes maybe you can fly to this strange country and research this
00:09:44.400
mushroom and and end up as a major player in this field and that has me so excited because one i really
00:09:50.700
think that people at much younger ages can and should be major movers and shakers in different
00:09:56.980
industries but this also reveals one other part of this goal that i think is really important
00:10:00.740
in that we do think that you can reliably cultivate genius and world leaders but you can't choose what
00:10:08.560
someone's passion is going to be like a parent can't be like i'm going to give birth to the best
00:10:12.980
violin player in the world or i'm going to be laszlo did but no no no laszlo polgar didn't plan on
00:10:19.340
creating chess geniuses his daughters really liked chess and based on his limited funds that was what was
00:10:23.920
doable for them and that's how it works when it comes to cultivating genius genius is is a is a
00:10:29.740
combination of disposition of each individual student or family and also what connections and
00:10:34.900
resources can be possible for that family you can't choose that so parents can't call a shot and say oh
00:10:41.860
i'm definitely going to have the world's best golf player i'm definitely going to have a brilliant
00:10:45.800
doctor you have to look at what can be possible and bring out the passions in a child and then make
00:10:52.260
it possible for them make the connections and that's what this system is about that has me really excited
00:10:56.620
and i think that this comes to one of our things which is admitting that genes play a role in humans
00:11:01.200
and humans do have some pre-coded predilections and you can't make an apple tree create pairs you
00:11:07.480
your goal is to give it the most fertile environment possible to be what it's going to be that is our
00:11:14.460
philosophy on education but this aspect of the system this reward and intrinsic motivation to play
00:11:20.160
with legos with more legos we call it our democratized nepotism system it's something that we
00:11:24.840
haven't talked we'll talk about secret society someday on on this platform but simone used to be
00:11:28.580
the managing director of dialogue which was the secret society that was originally funded by peter
00:11:32.260
tiel and arn hoffen and then schmidt futures hired us to build out their secret society thing that they
00:11:38.520
were building and then like we did matchmaking for one that the ea community did called future
00:11:42.020
forum and so we've been able to interact with a large number of people who are very classically
00:11:48.240
successful by society's terms and able to interview them and look to find out what is a
00:11:53.980
commonality among them and uh nepotism is is something that you see throughout this but not
00:11:59.580
nepotism in the way that people think it is not that they had people at they did have people advocating
00:12:05.960
for them yes yes but it's not necessarily that these people were family connections or something
00:12:10.260
like that it's that they made the effort early in their career to reach out to people who were high
00:12:17.020
level people in something they had a passion for and engage directly in those fields and so we are
00:12:23.460
trying to build systems where one using our rolodexes we built this amazing mentor network
00:12:29.100
and then through that mentor network we can connect students directly with these really high profile
00:12:34.720
individuals to work with them but even when students can't we don't have somebody for a student
00:12:38.740
and and to make this system genuinely scalable one of the things we teach students is how do you do
00:12:44.160
cold outreach how do you get people who are already in a field who you respect to engage with you
00:12:50.380
and to include you in their projects in a way that can get your name out there because when we're not
00:12:56.180
optimizing around the students graduation like just generic score but optimizing around their ability
00:13:03.060
to compete in this real world environment the thing that matters most is why our school will have an
00:13:07.700
in-house we've already worked on this in-house PR firm we've actually already built a partnership with
00:13:12.080
the 1517 fund which is a fund that specializes a VC fund that specializes in companies started by high
00:13:18.020
schoolers so we can have our students when they have ideas they know how to raise capital that's
00:13:22.820
one of the things we're going to teach them how all of these systems work that are often obscured
00:13:27.900
by the existing educational system that is essentially geared around creating slaves it's around creating
00:13:33.380
slaves for the bureaucracy and and so we can get into how all that works so now we want to talk about
00:13:39.480
how we handle socialization so one of the things that people often ask was this is they're like okay so the
00:13:45.460
first iteration of this is going to be predominantly online everyone knows homeschoolers have
00:13:49.740
socialization problems how are you going to handle that and hold on and then it's that's the bad toupee
00:13:54.540
problem you're talking about right there meaning that you only notice the bad toupee so there's a
00:13:58.480
perception that bad toupees create that you know look bad and there's actually been a lot of research
00:14:03.540
on this you can look at the research on this and while like all research it's all over the place
00:14:08.400
the majority seems to lean towards homeschoolers are actually slightly more better socially
00:14:13.260
adjusted than non-homeschoolers yep and it should be obvious why this is the case are kids who have
00:14:18.880
learned their social skills by modeling adults going to have better social skills than kids who
00:14:23.160
have learned in this weird lord of the flies setup we call public school what what you probably are
00:14:28.660
learning a bit more in public school is resiliency but i mean and social resiliency and resiliency to
00:14:33.860
bullying and okay that's all good but i mean can we maybe do something can we can we find
00:14:40.400
better than that but but outside of that i actually think that the existing school system is
00:14:44.040
terrible at teaching social skills well and actually i think most of the research says as
00:14:48.320
as parents at least with really really young children like toddlers when they're thinking
00:14:52.300
do i keep my home my kids home with me or with a nanny if i need one or do i send my kids to daycare
00:14:57.960
sort of the the trade-off is okay if you send them to daycare their language will develop more quickly
00:15:04.320
but they're also going to pick up a ton of bad habits so i think it's also understood already in some
00:15:08.980
peer-reviewed research that school is a source of bad habits more than it is a source of maturity
00:15:15.060
and ability to socialize plus of course the problem is school socialization teaches you to what
00:15:21.200
befriend those who are stuck in a room with you to befriend your inmates essentially and so this is
00:15:27.160
how you end up with people who only know how to make friends with their classmates with their workmates
00:15:32.640
with the people who they like live with in a dorm so in in the future they're limited to
00:15:37.280
what we would call i mean what what was called convenience friends we call them convenience
00:15:42.800
friends and parks and rec they're called workplace proximity associates but but the point being here
00:15:49.520
is is i've heard this before from people i know yeah is they they moved to the new city for the first time
00:15:56.420
and they realize no one ever once taught them how to go out and make random friends yeah and that's bad
00:16:02.460
that's really bad and so the the actual useful skills for going out and meeting totally new people
00:16:07.900
that young kids are learning they're not learning them in school right now they're learning them
00:16:11.760
in these online environments that they're often being scolded for interacting with so we we do have a a
00:16:18.820
course of the skill tree that is dedicated to this but it is really focused on making your own
00:16:26.360
friends going out there joining a sports team like a local sports team running in local elections
00:16:31.580
building local groups and your grades are based on how you actually perform at that how you do in the
00:16:39.760
election how you do with like how many people do you get to regularly go to these meetings that you
00:16:44.440
put together like well but here's the big thing too and it's it's such a notable thing that in school
00:16:49.680
no one like gives you credit for building a social network building a network when those things like
00:16:54.920
when it comes to getting promotions when it comes to finding a spouse when it comes to having good mental
00:16:59.740
health like those things like being able to socialize to build a friend network to get people
00:17:04.460
together i don't like even within thriving with your in your own job like building coalitions to
00:17:09.920
get a project done or managing sales or hiring people like you need to have these skills and schools don't
00:17:16.760
teach it another big thing that is a big part of our school is rather than have any teachers or anything
00:17:22.640
beyond ad hoc tutors when you need them and request them what we the only sort of personal touch point you have
00:17:28.480
the only sort of regular personal meetings you have are with what we call a proctor who's essentially
00:17:33.760
someone to check in with and just make sure that you are meeting your social goals that you're doing
00:17:39.260
well that you're not getting stuck somewhere like there's basically like a kind of like a life coach
00:17:43.640
without all the woo if you know what i mean yeah so and this is something we've really worked uh on
00:17:48.340
honing because if you've watched our video on how psychology became a cult it is very dangerous
00:17:54.000
whenever you put one person in charge of another person's mental well-being it is very easy for that
00:17:59.180
person to build dependency and the person who they're in charge of and use that to their advantage
00:18:03.940
and we already have a huge problem in the public school system i mean if you're watching like
00:18:08.060
democratic elite controlled media you don't know that like rape and molestation is a big problem in
00:18:12.060
the public school system but it's enormous it's enormous it's actually there was this study done
00:18:16.880
some people have debated whether this is accurate but it was done during a democratic administration
00:18:20.560
it's under in the clinton administration they showed that the rates of molestation in the public
00:18:24.180
school system are higher than they were at the height of the catholic church scandal but the unions
00:18:31.320
have a have a reason to cover this up so they're gonna cover it up and it doesn't fit the the generic
00:18:37.240
narrative so cover it up yeah but anyway so it's a big problem so we really don't want to put people
00:18:43.160
in a position where they can victimize students and we want to and we we spent a lot of time doing this
00:18:49.360
actually our system is a modified version of cognitive behavioral therapy that is less easy
00:18:54.100
to screw up less easy to mess up so we want to have a lot of safeguards on this system we want to
00:18:58.940
have it on rails so even a fairly stupid person can perform it without messing someone up actually
00:19:03.580
took a lot of inspiration speaking of catholics from the catholic confession system which i think
00:19:08.460
is a much superior system than modern psychology in that it's really hard for even a dumb person to
00:19:14.360
screw up it's got this anonymity which let's say unskilled team unskilled person to screw up you've
00:19:20.480
got this layer of anonymity which makes it a bit harder to get this person to build dependency
00:19:24.780
specifically on them and instead of telling you like oh there's this trauma in your past that you
00:19:29.180
got to work on or something like that they're like yeah what you did is wrong here's how you can
00:19:33.040
feel better about it don't do it again it was genuinely wrong and then you go out you do the thing to feel
00:19:37.740
better about it and you go okay i feel better about it now i did the thing obviously our system is more
00:19:42.200
complicated than that but what i'm saying is it's not impossible to create a system like this
00:19:46.240
that isn't based around uh a system that could be easily abused or messed up in a way that is
00:19:53.240
dangerous to students so we've worked a lot on that system very excited about that and that but the
00:19:58.300
reason why it's important and i think we just want to emphasize this is that right now in public school
00:20:03.300
we are seeing a mental health crisis we are seeing all these really huge issues with with teens and
00:20:08.760
i mean especially teens not being able to thrive in terms of anxiety in terms of depression i think
00:20:13.740
the bigger problem here and this is something that we want to emphasize is that the current
00:20:16.960
industrial school system is oriented around this really limited and increasingly obsolete outcome
00:20:22.000
we want your grades to be good we want you to go to a college we want you to go through this system
00:20:26.600
and that's not what predicts success anymore that's not going to predict good mental health that's not
00:20:30.840
going to predict that you thrive that you are mentally well or that you are physically well or that
00:20:35.720
you're getting a good job so it just it's bizarre that we're throwing our kids into this this shredder
00:20:42.040
of their best years in many cases when they could honestly be building their careers right away they
00:20:47.540
could be living great lives yeah so we built this as a certification system and the reason we built it
00:20:53.220
as a certification system is our rollout has a few plans we want to start with the children of the
00:20:58.320
elite i'm sorry like that's people are like why are you starting it was like elite gifted kids or
00:21:03.000
whatever and it's because that will make the system easier to get state approval for if we start in
00:21:08.620
really low income communities and then we try to move to middle class communities people are going
00:21:12.420
to be like yeah i don't want that system that's the one that the like the super ghetto system i don't
00:21:16.360
want that yeah but if we're like oh we're bringing like the the one that previously only the elites
00:21:20.720
could get and now like a gmail rollout is basically what we're aiming for but our goal is i want this to
00:21:25.920
be in every state and free and we're trying to build it so that it can eventually be free so with the
00:21:31.220
initial one we're looking at doing like a minerva like system well so first we'll just do a completely
00:21:35.180
online system then we'll do a minerva like system where we'll have some campuses in developing
00:21:39.720
countries where people can do what we call dynamic boarding school which means you can send your kids
00:21:44.540
to go 11 months a year six months a year however much you want and they're not going to slow down
00:21:48.360
because they're all learning through the same platform but if you also want somebody else to handle
00:21:52.180
raising your kids which a lot of people do especially wealthy people well that's a area of price
00:21:56.160
discrimination for us there and then for the general rollout one of the problems you have so
00:22:01.360
you look at a place like texas the truth is why doesn't texas have a good voucher or charter system
00:22:05.880
it's because the public education system has become a jobs program especially in rural communities
00:22:11.140
and the rural voting block is very important to the republican party which is typically the party
00:22:14.880
that's trying to innovate within the educational space you don't have a good voucher system in texas
00:22:19.080
and so how can we keep the money in the community instead of having it go out to these large
00:22:23.160
private equity companies that don't care because that's often what happens when you get these
00:22:26.360
charter schools coming in and the answer we came to was well what if we can utilize counter cyclical
00:22:34.160
community assets by that we mean community assets that aren't being used at the time when students
00:22:38.900
are in school and with the understanding the schools are often food programs for kids they're often
00:22:44.280
daycare for kids they're often a place where kids can be watched so can we certify the people at
00:22:50.580
your local public library or your local museum or your local religious center to be able to take
00:22:58.460
care of kids during the day and use those spaces during the day which will further distribute the
00:23:03.720
money throughout the community than just going to the local public school so because we operate so
00:23:08.400
inexpensively we can take a cut of that and then the rest can go back to these community institutions
00:23:12.940
community centers whatever hopefully make them better places for the entire community and support local
00:23:17.840
jobs and i yeah i mean i think it's really insensitive for anyone to be like oh i have this
00:23:21.440
great academic solution and then to forget about the food services about the safe place for students
00:23:26.680
to be about the fact that parents can't necessarily be at home watching their kids all day so we we
00:23:32.040
acknowledge that and we plan for it and then the final part of the system we haven't talked about
00:23:35.600
which really comes to our larger thing which is one of the things we hate about the school system is
00:23:39.600
the way that it imposes its morality on people and the way that it has become a center for
00:23:45.000
ideological indoctrination that terrifies i think a lot of parents and it's a big problem with if
00:23:51.640
you're even just talking about this as a business opportunity people are like why are the pro natalists
00:23:55.900
also working on school systems and it's look there is it yes school systems are becoming increasingly
00:24:03.480
irrelevant boarding schools all over going bankrupt all sorts of private schools are going bankrupt but
00:24:08.100
these are often super super lefty institutions school systems for people who don't want their kids
00:24:12.820
brainwashed into this urban monoculture that's actually an exploding market right now because
00:24:17.840
the people who are having kids are often from these more conservative cultures that don't want their
00:24:21.840
kids homogenized into this monoculture right so how can we create a school system that never imposes
00:24:28.260
our ideology onto students and that is modularizable so families can ensure that it stays within their value
00:24:35.540
system and the way that we have done this is our school system has two types of tests one which is a
00:24:41.880
multiple choice test which are constantly generated by ai then filtered by a human and then they go into
00:24:46.640
the school so less bias there and we use ai with prompts that decrease its progressive bias because
00:24:52.160
right now ai has created a lot of progressive bias then we measure those questions against authentic
00:24:58.220
assessments which are the mastery courses within our educational paradigm and by that what i mean is
00:25:04.200
occasionally students need to do something to show i understand all of the concepts being taught here to a
00:25:07.840
high level so what is an authentic assessment this is something in educational pedagogy which means
00:25:12.200
something that tests students ability to compete in a real world environment for example in english you
00:25:16.740
could test them through how many five-star reviews that they got on a fan fiction they wrote and posted
00:25:22.500
online or how many upvotes their scary story that they posted on a reddit got which is fantastic because
00:25:28.060
you're testing their real world writing ability and one of my favorite things is we mentioned this and we
00:25:31.580
there were there was this teacher who we were talking to and she goes oh well i'm sorry i am a
00:25:36.920
author the failed author obviously and english teacher and let me tell you what didn't you know
00:25:42.480
that 50 shades of gray used to be a fan fiction and it is terribly written and if she was in my class
00:25:50.020
i would have given her a c and i go is there anything more indicative of the problems with the current
00:25:54.960
education system and this failed author masturbating to herself about this idea of punishing one of the
00:26:02.240
most successful books in the last century in all of human history actually because it doesn't fit
00:26:09.480
what she considers good writing which obviously isn't performing well in a real world environment
00:26:15.060
and this is the problem here we because we haven't been forced to measure people and how they perform
00:26:20.140
in real world environments we have unfortunately conflated a few different things in some
00:26:25.760
measurements so what we're really measuring in english scores is a student's ability to show class
00:26:29.740
status through how they write and a student's ability to write for a reader as soon as you
00:26:33.280
disintermediate these into two tests like to test how a student writes to show class status you can have
00:26:38.020
other students rank students writing samples by the class status they think the student has based on
00:26:43.380
that writing sample and then use that to grade the students so then what you do is you take the way
00:26:47.660
students perform in these and then you test the individual multiple choice tests that were created
00:26:54.280
the individual questions i can see within every individual question how much they correlated with
00:26:59.320
the student's ability to compete in a real world environment and measure our own questions and then
00:27:03.360
constantly create new questions so i have a system that's constantly self-healing and improving but
00:27:08.460
here's where it gets really interesting so we've got a partnership with metaculous which is the major
00:27:11.560
prediction market player where people go and they make predictions about future events and you can win
00:27:16.000
tokens and stuff like that really cool suggest you check it out but anyway we have a partnership
00:27:20.760
with them to do student-based questions around more political outcomes because remember if you had
00:27:26.900
written a political theory paper leading up to the trump hillary election that said trump has a chance
00:27:31.960
of winning and we saw this was like pundits who said trump has a chance of winning they would get told
00:27:36.360
that basically they were crazy people they were idiots what they were doing was malpractice
00:27:40.060
because the industry itself was wrong the industry had become so ideologically indoctrinated they
00:27:45.540
weren't able to see truth anymore we saw this with the ukraine war as well was the ukraine war everyone
00:27:50.180
saying oh russia's going to steamroll ukraine and then that doesn't end up happening right
00:27:53.580
dogma in the field had hidden people who might actually have a better understanding of the field
00:28:00.380
than the status quo so by having students compete in these metaculous like environments for their
00:28:06.360
authentic assessments we can see which students actually understand political ecosystems better
00:28:10.000
actually understand economic systems better and then correlate how those students do within those
00:28:15.700
tests to how they answered individual multiple choice questions to understand which of those
00:28:19.460
questions are actually testing a student's real world ability instead of their ability to parrot back
00:28:24.120
ideological indoctrination and here is where this gets really interesting with some of the more
00:28:28.640
controversial stuff so you talk about something like evolution right i can't test a student's ability
00:28:32.880
to actually understand evolution but i can't test a student's ability to convince somebody who has a
00:28:38.300
high level of quote-unquote understanding of evolution that they understand evolution we believe in evolution
00:28:41.840
but i i do recognize that this is a true thing right so this means when students are taking the multiple
00:28:47.900
choice tests on evolution it is made clear to them that what they are being tested on and what they're
00:28:53.440
going to later be tested on is their ability to convince someone that under understands what we call
00:28:58.940
evolution that they too understand evolution not understand an objective reality because we can
00:29:05.220
never claim to be testing a student on more than what we can measure the student on which is really
00:29:11.240
exciting to me i i really like that this system is self-constructing because students can begin to
00:29:16.980
create their own learning materials they vote on these materials and it's self-healing to an extent with
00:29:22.420
the students but here's where it can get really interesting which is some of the weirder stuff we've done
00:29:27.880
around education one is we have a partnership with praxis which is this organization that's trying to
00:29:33.660
create a new country and one of the things that we've theorized about doing is this partnership
00:29:37.040
is having a member of their government who is an ai amalgam of all of the student writing so the
00:29:45.840
students contribute to decisions that are being made in the government by how much they contribute to
00:29:53.680
their academic work which i think is just a really cool way to give students a vote but also encourage
00:30:00.040
students to get out there and create more content the other another really interesting idea we had here
00:30:07.020
was can we build prediction markets like get all students to constantly be predict engaging in these
00:30:13.620
prediction markets and because prediction markets are so successful get companies to actually pay into the
00:30:19.980
school to get access to this prediction market data so a company could because this has been shown in
00:30:24.900
companies prediction markets are actually fairly good at like determining outcomes within companies
00:30:29.900
so like if a biotech company is going to do some big uh experiment right or or try to produce some drug
00:30:36.660
prediction markets can determine with a high degree of accuracy whether or not that test will be worth
00:30:41.080
the money they put into it they can then basically pay for access to once we're running in all the u.s states
00:30:46.340
to access to the cumulative brain power of all of our best biology students for example and then this
00:30:53.140
can be a source of funding which then can go to things like scholarships and stuff like that so the
00:30:58.460
school itself is acting as a source of cumulative brain power which i love we also want to build a system
00:31:06.480
that can track students after they graduate how those students are doing in terms of mental health and
00:31:12.540
career outcomes so in the long term not just we want to measure students off of authentic assessments
00:31:17.100
but we want to measure off of how they do in real world environments and throughout their entire lives
00:31:23.840
and how they're doing in terms of mental health throughout their entire lives when we are a b testing
00:31:28.020
different parts of our system because i do want our system constantly a b testing having students go
00:31:32.000
through different types of trees different ways of doing things to see which ones have better outcomes
00:31:37.160
and then choose those models for ones that we can then distribute throughout the entire system so we
00:31:43.780
can have this system that's constantly experimenting but that isn't just looking at how students do in
00:31:48.200
terms of graduation but how they do in terms of real world outcomes anyway love you saban sorry for
00:31:54.100
panting suffice it to say you're excited about this and you have every reason to be but now we have to
00:31:59.840
actually finish building it and make it work in time for us to at least have our kids run through this
00:32:04.480
because yeah we had a lot of our funding pulled for this at one point and then we got a bunch of
00:32:09.760
funding for the pronatalist stuff so the work on this slowed down a lot i mean it's still going in
00:32:14.860
the background we've got a team in africa working on it right now but man i wish hey if any of you are
00:32:19.440
big donors we might be able to pull something together here listen we're gonna get it done regardless
00:32:24.820
i don't care i mean honestly i'm i'm stoked about us owning all of it and running all of it with no
00:32:30.440
adverse incentives so whatever we're just gonna make it happen but i'm curious if anyone else is
00:32:35.680
building their own secret system for cultivating genius and building the world's future leaders
00:32:41.280
we want to hear what your plans are because again we feel like this area is rife for disruption or ripe
00:32:46.240
for disruption and the more players here the better honestly it's we're really sad about how little is
00:32:54.340
being done here so join us tell us what you're doing really want to hear about it and check in