This French tech school has no teachers, no books, no tuition
venturebeat.com
venturebeat.com
In this type of system you learn how to search knowledge it changes the pasive role of the student into an active role. Also you learn that healping others and teaching them is good for everyone and never a waste of time. I think the school works really well and one would need to measure the succes of their students in the happiness and sense of acomplishment they have.
The school is not a factory producing identical products. It's a human journey of knowledge seeking and personal developement and everyone developes diferently. At times it's very intensive and people that are not really passionate and motivated drop out.
As an example of what we do this year the first project we worked on was to recode the malloc function in c. To do this we have some videos explaining the basics as well as some introductory excercises. After some research i found that the projects at harvard cs61 are very similar to what we will be doing in the unix module so i used the book they recomend there Computer Systems: A Programmer's Perspective. I read the relevant section and afterwards i coded a lovely implemenation of malloc. I finished in time the project. Some times it is more dificult, for example for the assambler module i found that it is more dificult to find good explanations online. But this time another student explained some key things and afterwards i was coding in intel assembly.
The most dificult information to come by is with respect to parsing and compiler theory or coding in Ocaml. Sometimes i think that if i had a teacher which was an expert in the field this sort of thing would be easy. But the truth is that learning is a personal journey and no one will solve it for us. Education needs to evolve to generate mature people that study and learn because they find that doing so fulfils them and not because they are searching for validation or an external reward waiting for someone to clasify and rank them. Clasification and ranks are subjective and are only means and not ends.
> As an example of what we do this year the first project we worked on was to recode the malloc function in c [...] after some research i found that the projects at harvard cs61 are very similar [...] so i used the book they recomend there Computer Systems: A Programmer's Perspective. I read the relevant section
So you used the materials put together by the teacher responsible of CS61 at Harvard, and you read the book recommended to the class
You say Epitech and 42 are similar and it describe 42 as: "This French tech school has no teachers, no books, no tuition"
Yet to succeed in that case you used the Harvard teacher information(maybe some lecture notes, slides?) and read a book.(borrowed from the library? or 'found' a pdf online)
Also on a side note: you had to be able to read and understand that book so your English level had to be at a good working level, level that you do not really have when you're straight out of French high school, does 42 expect their students to be proficient in English?
I don't think anyone imagines that human beings can educate themselves into pushing the state-of-the-art in technology, especially in the XXI-th century, without benefiting from actively transmitted knowledge from former generations.
What's disrupted by 42 isn't the concept of mature people helping younger ones acquire knowledge and later expand it; it's the school setup, its organisation around academic authority, its mass production mentality (use exams to filter uniform kids--socially and intellectually--as input, so that you produce normalized batches graduates as output, through a highly scalable and repeatable process). The teacher's knowledge, and his ability to condense it into a synthetic form (here a book), remains as valuable as before, if not more. Teachers currently have to use the Ivy league bureaucracy in order to turn those skills into a livelihood, but that can be changed.
Schools destroy kids' curiosity and self-confidence, although those are the best individual learning drivers. That's not (only) willful sabotage: the problem is, those drivers are well suited for tailor-made education, but they're considered unmanageable to handle batches of hundreds or thousands of kids. What 42 questions is: can't we claim those drivers back, now that Internet has deeply altered how humans can interact? I'm thrilled to see the answers they come up with.
42 innovated on that, I think it is the only privately-owned school with no tuition. Schools with no tuition (or a very low one that you can pay in one or two months of internship) are not rare in France though, it is actually the norm for the public "Grandes Écoles" and universities.
Here are the lecture notes: https://courses.engr.illinois.edu/cs421/fa2014/lectures/inde...
Hope they might be helpful to you.
> France already has a reputation for creating great engineers (in software as well as in many other fields).
Engineering schools[1] however have a lot of teachers, students have more than 35 hours of class a week. The tuitions are already very low compared to US schools (and there is already no books). Their process is a lot different from what École 42 tries to do.
[1] http://en.wikipedia.org/wiki/Grandes_%C3%A9coles#List_of_gra...
Because of the shortage of coders, big french consulting companies (Atos, Cap Gemini, ...) hire since decades graduates from every domain more or less related to science/technology and give them a six months training (in the best case) before sending them to the clients.
So I'm not that surprised that such a school will be successful as long there is a coder shortage. When the next bubble blows, these graduates will probably be the first to be layed off.
The goal of this school and by extension of 42 (the school mentioned in the OP link) is to create real hackers, people able to learn by themselves. Most of the projects are coding projects right but it's oriented in a way that forces you to learn all along the way if you want to succeed. Learn from internet, books or your peers. RTFM is the motto of the school.
You'd be surprised by what the students are able to achieve.
If that matters some Epitech students work in very successful companies (fb, twitter, google or microsoft) and the Docker founders were in Epitech.
The problem of most people coming out of grande ecoles is that they lack experience, after a few years on the job, they can become great, but newly graduated engineers from Grand Ecoles are not very efficient at the beginning on average.
On the other hand, I've seen great candidates from Supinfo and Epitech who learned a lot of the CS theory on the side by themselves and were really good...
I think learning CS Theory without a good practical background is actually wasted a bit and that more Grandes Ecoles should try to get students to have more practical experiences which would help them connect the theory to actual situations they've had.
It was for Rails web application which I have to say requires somewhat less pure CS theory (although it is very useful to reason with).
That sounds like theory, probably closer to RTFSO (read the fucking Stack Overflow)
The creators of Docker and React Native come from Epita/Epitech, so it does produce a few gems :)
Docker comes to mind, as a silicon valley success story founded entirely by monkey coders.
Graduates from Epitech, or 42, are very far from having the same ability with mathematics and quantitative sciences in general. That being said, it probably doesn't matter since most tech projects don't require such skills.
The only way that wouldn't happen is if the grandes écoles brainwashed their students into stupidity, while Epitech and 42 had a time-distorted super saiyan training chamber.
That doesn't say much about the quality of the curriculum.
I agree.
However, this thread is not about the schools themselves, but their output, the graduates (c.f. parent and GP: "produce monkey coders", "yield better results").
Are you considered more apt to be a leader, or qualified to be a finance quant (even if you never trained for it) with a top school or your resume ? Yes you absolutely are.
Also "most of them don't actually work in tech". Yes, that. Programming is considered a dirty job in France by most people from grandes ecoles. They would all rather be architects or consultants. It sounds better on the business card. But someone needs to actually do the job. And this is why there are only tech services companies, but nothing close to a Google.
The thing is that the french system lacks a proper CS education. "Grandes Ecoles" has the best student but offers pretty limited courses in CS (except maybe TPT, not that sure) and not that much coding experience. Universities are the only offering deep theory education and sometimes good coding experience but their degree are not well recognized so they don't attract good students. As a consequence private schools try to fill the gap by offering a great education in programming, however lacking the CS foundation.
My school's undergrad curriculum focused very heavily on theory and I appreciated and thoroughly enjoyed it. I think I've had a much stronger theoretical foundation than most I've seen with similar experience. On the other hand, I don't really feel like any of my jobs have really required that knowledge whatsoever.
I certainly don't see a problem people learning any theory necessary as they go if they are just planning on working in the industry. I certainly appreciate the theoretical background but I really would have a hard time believing having this theoretical knowledge is necessary to becoming more than a "monkey coder" whatever that means.
It's not as though these people are going to a research university or planning on doing research; they are going to learn to program and maybe gain the skills to find a job, and that's certainly what they are getting.
Epita is an engineering school and it is absolutely not focused on producing programmers. The school teaches electronics, physics, OS, embedded systems, compilers, management, security,... _and_ programming.
Epitech, is not an engineering school and is focused on creating programmers (but still provide some other courses).
If you select the very best few, you could get them into an empty barn, or have them walk the Appalachian Trail or whatever and the "school" will have great success measured by what the smart motivated students accomplished.
The question to me is
a) how well do they learn things and understand the deeper principals such that their understandings generalize?
B) what do they end up learning/creating? Is it like the mit media lab where they're encouraged to innovate 5-10 years out, or are they building stuff that's existed for a long time like Ray tracing?
C) if everything is team based, do the smarter/type a ppl end up doing all the work or does everyone contribute equally enough to learn?
D) is there any kind of mentorship? The stuff I've learned from my mentors across school was never what was being assigned in class; understanding that there are things that you would never think of as entire fields needs guidance for the not yet educated
Sounds like a cool concept for self starters however.
b) first we are building stuff the way the school wants, basically you have to do a programm that do certains things (like recoding a wolf3d for exemple) and you have to follow the coding norme of the school, you are free to do it the way you want as long as you follow the rules for the project. But there is also free projects, during 6months you can do what you want the way you want as long as it is a real project and not 6months of vacation.
c) not everything is team based, except for rushes and some projects, all projects are individuals, we learn together yes but we don't code all the time together.
d) You don't have one mentor, every student can be your mentor, the concept of the peer learning is exactly what you describe, talking to other students to get help, another approach of the projects, learning new things, etc.
Also, they have no tuition but neither do the public universities in France ("Grandes Écoles"), and the most selective and prestigious schools are public. So while the "no tuition" thing is impressive for a private school and may sound like a dream for Americans, it's not really rare in France considering the choice you have with public schools.
Finally, not really related to the way they teach there, but here's the director (same guy in the picture on the article) spanking a girl in a classroom during off-hours (nsfw). http://www.19h59.com/WTF/10238_Le-directeur-de-lecole-42-don...
As koolkat said, 42 is a fork of Epitech. In those schools, programming is taught by projects, and supervised by teaching assistants. The teaching assistants are older students selected by their CS knowledge and their human abilities.
This method is preferred by the school because a TA costs less than a real teacher ; and preferred by students because the TA have a closer relationship with them and have a better understanding of the projects (they coded this project 1 or 2 years before).
Personally, I was head of the teaching assistants in EPITA for 1 year. EPITA follows the same principles of projects and teaching assistants. We were a team of 30 students and lead the projects for approx. 300 students.
However, I am wondering if these students would learn the computer science fundamentals which are usually found in more or less dry books. Programming analysis, logic, fundamentals of programming language implementation etcetera. You can't say: look on the internet and find the relevant papers in all those fields and synthesize the theory by yourselves. That would be highly inefficient, a problem we tackle with education. If you make the comparison to Harvard you aim for more than educating people to become good/ exceptionally effective developers in my understanding.
Yet, they do create people that are truely amazing at coding.
The kind of people going out of this school will most likely build OS implementation stuffs, or security appliance, or VM or compilers. But probably not a new compression or distributed concensus algorithm, unless they already have a personnal interest for the fields.
The greatest French hackers like Jean-loup Gailly (zlib) and Fabrice Bellard (ffmpeg, QEmu and much more) actually went to Polytechnique (X), where they actually got something 42 students are very unlikely to get, an extremely strong math background readily applicable to multiple fields. Fabrice Bellard was already an excellent self taught programmer at age 17 when he released LZEXE. What got him to superhero level was likely more the well rounded, traditional French generalist engineering coursework he followed at X.
I completely fail to see what would attract such bright people to the 42 school, when X, ENS and a few good engineering schools and university programs are available.
42 is an excellent opportunity for the people that for some reasons didn't strive in the more traditional system, but I find the possibility it start to attract top people very remote, because being taught by students two years older than you isn't nearly as attractive as being taught by top researchers, and being on a campus and mingle with top mathematicians, physicists, biologists chemist is IMHO a lot more stimulating than being among only rejected by the system computer geeks.
For both zuck and page, their uncommon technical accomplishments came from people they hired: a small army of very-well-educated engineers with traditional education.
Google seems much more "bleeding edge" though, and i doubt anyone not working in a university on the field could have come up with the ranking algorithm.
Like others on the thread, my hunch is that there will be a small group of self-driven students who will teach themselves using the same formal education methods: replicating the courses given in other universities, reading textbooks/papers/research, researching online, and watching MOOC style videos.
This group will end up teaching the others in their class through project collaboration. But unless everyone in the class is self-driven, the result would be a couple of successful self-taught developers, and a whole lot of below average ones.
Sounds like most CS schools I know then.
The project looks awesome obviously. What you need to know out of the context with it's director (Xavier Niel), is the guy is an ace of communication.
Looking a little more into details and feedbacks from people who did the school, it's a VERY different story. I would highly doubt anything that comes to the communication about 42.
Probably the vision is achievable, I'm not sure it is by this school.
I'm curious to know where are those people who came out of42. The school opened like 4 years ago right? So the first graduates must have been last year.
If you have any stories to share I'm curious :)
I did 42 for 1 year. Now I'm working as sysadmin.
During the first year, we had projets with limited time (depends on the difficulty, from 1 to 6 weeks). We had 2/3 projets at the same time.
This school is kinda like Epitech. The staff is composed by students (I was one of them, I joined them after 3/4 month of learning at 42) and people who came from Epitech (sysadmins, 'teachers', web dev). Yes, there is a kind of 'teachers'. They create the first videos, the subjects (mostly taken from Epitech too :))
But this year, there was a big change. Now student can create subjects, videos. The projects are not limited in time. We are not working for 100hours/week, but a lot of people are at school for more than 50hours/week. This depends on their faculty to learn and their motivation.
But...
There is some bad points for 42.
There is no labs. There is no books. There is no English class (as you can see with my poor english :) ), no math class, no French class (and a lot of people really need it....). When you see people reading the man pages in french, you feel sad for them. Whe you see people not able to do some math, you feel sad for them too. And, seriously, a library or some labs for the students can be a good thing.
42 is a cool thing for people who have not the 'baccalauréat'. I dont have this degree but thanks 42, now I have a great job. If I had this degree, my choice for a school might be something like ENSIMAG / ENSEIRB, but definitly not Epita / Epitech.
Btw.
> In fact, 40 percent of École 42’s students haven’t finished high school. Others have graduated from Stanford or MIT or other prestigious institutions.
Why someone from a prestigious school would like to come to 42 ?
He's one of the only french businessmen having actually disturbed anything in France, a country where "innovation" is usually driven by publicly owned big companies and public grants. He has driven telecommunication prices down dramatically with his internet provider company, and then his mobile phone carrier. Internet is cheap in France mainly thanks to him.
I trust him probably more than anybody else in France to actually achieve something disturbing the system in place, precisely because he's one of the very very few that has done it before and he does not come from the elite making system of "grandes écoles".
Amusing fact, he started his fortune with sex messaging newtorks on the french precursor of internet, the Minitel.
And I'm pretty sure we have one of the best offers in the world for mobile plans thanks to him as well. I haven't seen a lot of places in the world where you can get pay on the go unlimited calls/texts/data for such low prices (10€/m)
France used to have some of the best internet access in the world, only behind the like of South Korea or Japan. Now it's down to #30 [1]. Same (or even worse) for mobile networks: 4G took forever to come to France. Investing in better infrastructures is expensive, and ROI is just not there to be on the cutting edge anymore.
[1] http://en.wikipedia.org/wiki/List_of_countries_by_Internet_c...
We can have a look at USA with probably the most expensive internet / mobile plans and worst internet (excluding undeveloped countries). On the other hand, there are many emerging EU countries like Lithuania, where you can get stable 4G across a country for 5 eur/month.
FFS, I just received a bill from the hussier about a service SFR never provided, billed me for, and I've been arguing with the useless CSRs for over a year now.
The old mobile cartel in France used to have the best speeds but now they are more interested in paying their investors instead of delivering what their clients have paid for.
I think, if you consider that the job of a school is to educate people, then if 42 can turn someone with little education into an average to great programmer, that's a great accomplishment. In comparison, Ecole normale sup turns extremely bright and knowledgeable people into great programmers, not such a great accomplishment IMO.
Well, that's not what the article says.
"Last year, 70,000 people attempted the online qualification test. ... In the end, 890 students were selected for the school’s inaugural class"
But I agree that it's difficult to judge the quality of formation considering that the student populations are different. But still, I don't see how how throwing programming challenges at CS beginners with no supervision can be better than university professors.
""" Last year, 70,000 people attempted the online qualification test. 20,000 completed the test, and of those, 4,000 were invited to spend four weeks in Paris doing an intensive project that had them working upwards of 100 hours a week on various coding challenges. In the end, 890 students were selected for the school’s inaugural class, which began in November, 2013. """
Not true :( Even though I thought I completely blew the preliminary tests I then received a mail telling me I wasn't accepted.
Although as an aside, my female friend who gave up before finishing and completed less of the tests than me got in, so I'm not sure I would even want to be accepted now but still, there is some heavy pruning going on.
I don't buy it. First, there's a large panel of CS formations in France, most of them virtually free. It goes from highly selective schools "grandes écoles", to universities or undergraduate programs that anyone with a high-school diploma can attend. There's room for many different student profiles there.
Then, I don't see why it would be such a progress to get rid of the teachers and the books, are they that bad that they prevent the students from improving?
> The students are given little direction about how to solve the problems, so they have to turn to each other — and to the Internet — to figure out the solutions.
Usually, this is what professors do when they are too lazy to cook up a real project.
Meanwhile, we also had projects that we had to figure out by ourselves. For instance in the first semester of the first year, there is a course that explains theoretical stuffs that are supposed to have a relation with hardware (like 2-adic numbers [1]). The project for this course consists in programming a watch that displays time. But we have to program the watch in an assembly language that we have designed ourselves. And that asm has to execute on a processor that we have designed ourselves. And we have to write the processor code in a hardware/netlist language that we have to design ourselves. And we have to write ourselves the interpreter we will use for that netlist language.
That was what we were told to do, with no additional instructions. Some students learned C by themselves for that, some other C++, some other OCaml… And all of them learned or reinvented how to design memories, or a CPU (for instance, me and a friend decided to invent a single-instruction processor, just for the fun, we got it working but it was rather slow compared to other who decided for instance to learn how to implement a simplified MIPS).
All this to say that there is absolutely no incompatibility between having professors and teaching (partly) by projects.
The ENS is a special case, as it does not wish to form engineers, but researchers.
This said, what I would be worried about with this system is how long it will take for the employment market to adapt to this/those schools. When it comes to jobs, French companies, especially large ones, have a really strong bias towards applicants from well known and somewhat elitist public schools, no matter how efficient those might actually be at their jobs.
Unlike the US, France doesn't really have a strong and major software industry competing for raw programming talent, so developers have nowhere near the salary, status, or lifestyle they could expect in places like SV or NY. Very few companies have a real incentive to hire a great programmer over a good one, which might reduce the efficiency of a school that focuses on training very good developers without giving them the access to the prestige of the (also free) traditional public schools.
It will be interesting to see how salaries and career paths available to programmers trained at École 42 will compare with those hired after they graduate from École Centrale Paris.
Nevertheless, I look forward to hire from this talent pool.
Not only French companies, but also banks, hedge funds, companies like Google and so on...
I was saying that in an other comment, but recently Google Paris organized a programming contest. All teams in the top ten came from top schools (mainly ENS, but also ECP and X) whereas 42 performed extremely poorly.
There's a reason why employers are biased towards elite schools.
Does this provide real proof that the model on which 42 is based is wrong? Moreover, how well does a score in this programming contest correlates to real world performance?
From experience, the strengths of school networks is a much stronger reason why employers are biased towards elite schools. Please note that I'm not saying elite schools aren't good (it's a pool I hired and will hire from), I'm just welcoming the initiative and wondering how it will pan out.
I think it's a good thing to disrupt prestigious universities that hire on credentials, but France is not exactly the country for that since education is mostly free and admissions (at least in engineering schools) tries to be more meritocratic.
EDIT: disclaimer: I did an engineering school in France.
I kind of doubt that giving a person with no programming/CS experience Google and a task like "write a concurrent distributed high-performance database system" would lead to a good outcome. There is a lot of trade lore, experience, knowledge of what works and what doesn't, etc. which is better passed by a teacher in an carefully constructed narrative than by random discovery by an uninitiated person.
But it is interesting to see what comes out of it, maybe they prove me wrong.
There is (more or less) all that in 42, thanks to the "peer to peer" approach.
It's in Paris, and the cost of living there is astronomic. Thanks but no thanks.
On top of that, what's the point of taking the smartest young individuals and teaching them how to be engineers ? It's elitist. I agree that the fact that it being free is great, but to me, those students would have been great programmers without that school.
It's easy teaching kids who already know almost everything.
The only programming experience i had before 42 was some html/css (far from knowing everything), i could barely do a static website (and i'm not young either, i am 27years old), there is no way i would have learn so much (c, php/js/mysql, assembler, ruby on rails, using docker) all by myself with no guidance at all and absolutly no idea where to start. Being in 42 gives us the motivation, alone at home with a slower learning, not so easy.
Yes living in paris is expensive, but if you live in the suburb, maybe part-time work for some people (no deadlines means you can actually work if you want or need it), you can manage it, it's just 3years, with a 6months internship during the first year (you can be paid) and a third year with possible sandwich course if you need it.
As to learning how to code, I bought a C++ book, and I learned the basics of C by myself, but never really coded anything until I got into some french BTS.
Yes 42 is a bit of a fork of epitech, but they don't have the same program anymore, the basic program is the same (projects for the beginner) but new projetcs are totally new, and students are encourage to create new projetcs too (we actually have some, like a project including docker, a student is preparing a OcamL pool etc...).
"All of École 42’s projects are meant to be collaborative, so the students work in teams of two to five people" This is not true, the main concept of the school is the peer learning, you are not suppose to do solo projects with other peoples, you are suppose to learn how to do the project with other people, this is actually a different concept, we do have group projects though but it's a minority.
> Pourquoi 42 s’appelle 42 ?
> On vous met sur la piste… regardez du côté du Guide du voyageur galactique
"Don't Panic" would certainly be an appropriate slogan for a school like this.
It's a good PR for Niel.
I'd estimate that well over half of my bootcamp cohort was over 30.
I would get in serious legal trouble if I posted an ad saying "Hiring software engineers; ages 18 to 30 only." Doesn't the same apply to schools?
the last big IT assets are being sold off, complete failures. europe as a whole basically is SAP plus its ecosystem. nothing else has survived. and even SAP is investing massively in its shift from Walldorf to the Bay Area.
so yeah, a lot of countries have elite schools. hint: all big founders never finished theirs....
Engineering schools do not have a lot of teachers in France, I don't understand where you get that from.