Hacker News new | past | comments | ask | show | jobs | submit login
Craig Silverstein, Google Employee #1, Leaving for Khan Academy (allthingsd.com)
318 points by jc4p on Feb 9, 2012 | hide | past | favorite | 79 comments

I'll mention that we're still hiring for both full-timers and interns. Here's the job post:

Our mission is to provide a world-class education to anyone, anywhere. We already have millions of students learning every month, and we're growing quickly.

Our students answer over 2 million math exercise problems per day, all generated by our open source exercise generation framework (http://github.com/khan/khan-exercises, http://ejohn.org/blog/khan-exercise-rewrite/), and Sal's videos have been viewed over 117 million times. We're just getting started feeding this data we're collecting back into the product to help our users learn more (http://david-hu.com/2011/11/02/how-khan-academy-is-using-mac...). If you're interested in data, analytics, and education, this is a dream gig.

Plus, it's one of the highest educational impact positions you can imagine. We're hiring all types of devs -- mobile, frontend, backend, whatever you want to call yourself. Big plans ahead.


I'll just add that even though the job ad isn't up yet, we're specifically looking for you quantitative big data folk.

Apply through the ordinary software developer position for now -- but we're hunting for some passionate data engineers to do more stuff like this: http://david-hu.com/2011/11/02/how-khan-academy-is-using-mac...

I'm from Kaggle. We'd be happy to help you anonymize your data and create and run one or more data mining competitions for free. Ping us through www.kaggle.com if you're interested. Might be a good way to get started with deep data science, and also leverage the data engineers you hire once you have them.

can some of KA data be made available for 3rd party analysis?

We've thought about it, need someone to go through and properly anonymize everything, etc -- not at the moment.

Sounds like a once-in-a-lifetime opportunity to see how first hand how some pretty famous developers work. The company now employs both the creator of jQuery and the first Google employee, and that's just the development side!

Even though I currently work remotely from school at a fantastic company that I used to work for full-time before I re-entered school, being able to help KA in anyway would be a dream come true for me. Sadly I've had to resort to creating a crontab entry for e-mailing me when any non-known IP visits the projects page I set up for my application, no visits yet :(

I'm kicking myself for not making something like the URL kamens tweeted couple days ago for the kid who was looking to intern, but it's kind of hard to make a beautiful website when the thing I'm most comfortable with is backend design and MySQL.

I'm just hoping kamens' tweet about "throwing away half your resumes" so you don't hire unlucky candidates doesn't end up being the case!

but it's kind of hard to make a beautiful website

A $20 Wordpress theme solves this problem.

Seriously. I don't understand why programmers insist on trying to do design work themselves when the price of quality designs is so low.

Sure, if you want to noodle around with design for its own sake (learning is fun!), go nuts. But if you just want to get on with programming then spend the $20 and be happy.

Slight downside: having to put up with Wordpress...you can't have everything.

Just yesterday, I was browsing the exercise generation framework - Khan exercises on github yesterday and had a question. Can the framework be used independent of KA (ie for other exercises and youtube videos) ?

Thanks for the great job you guys are doing of bringing quality education to everyone.

Absolutely. The server-side code is also open-source and can be found here:


Of course, we write the software specifically for the Khan Academy site so you would need to change some parts of it to adapt it to other uses, but everything is out in the open for anyone to work on and use.

Its post like this that make me sad to be a .net guy living on the east coast, even though I have 10 years of web app experience.

Remote work can be a possibility for the right candidate -- there are a couple devs who work remotely right now. If you're interested, it can't hurt to apply.

Do you hire people from other countries? To work from remote.

I wonder when they'll start hiring, well, educators. While engineering challenges may exist, Khan Academy's biggest scaling problem seems to be Khan's ability to produce solid content. His biology stuff, for instance, is decidedly worse than the mathematics.

Other than that: great move.

This is an ongoing falsehood. Several Khan Academy employees were real honest-to-goodness teachers, and we work with teachers directly in-person every single day to make the product and the content better. In addition, we are looking for more folks to make videos. Our standards are extremely high, and that will take time, but I disagree that this is our biggest challenge in scaling.

Furthermore, your opinion of the Biology content may be spot-on, but there are thousands of students who say it's making their lives measurably better. Maybe it would be 100,000 if the content was better; it's hard to know for sure. I'm just not sure that it is directly affecting scale at this point, or that there's not a greater effect on scale trying to get more people to use our existing content.

Disclaimer: I'm the lead designer for KA.

  > This is an ongoing falsehood.
I realize that now, and apologize for spreading a misconception. We don't hear much about non-engineers through the biased HN-filter; hiring Silverstein gets you headlines. Part of my point, however, stands.

  > but there are thousands of students who say it's
  > making their lives measurably better.
I'd prefer KA to emphasize content -- quality content made by solid educators in their respective field -- over delivery method. That the latter enjoys priority right now (as Khan's comments make clear) may be a necessity, but an unfortunate one. But I'm pretty sure that you'll get there.

I'd prefer KA to emphasize content -- quality content made by solid educators in their respective field -- over delivery method.

Depending on what you mean by "educators," I might prefer that Khan Academy (and a bajillion different competing providers) offer up content by actual domain experts rather than content by "educators." It is, of course, possible for a person to be both a domain expert in actual fact and a secondary school teacher by occupational category--I've seen Richard Dedekind described as an example, although I'm not sure I'd describe his teaching position as one resembling that of a high school teacher in the United States. But anyway the correct idea that content has to be both factually accurate and appropriate to guide the development of young learners does not constrain content-creation only to persons with the formal credentials of schoolteachers. Many of the best learning materials for young people today were produced by authors who were not K-12 schoolteachers in any stage of their career.

  > Depending on what you mean by "educators," I might prefer
  > that Khan Academy (and a bajillion different competing
  > providers) offer up content by actual domain experts
  > rather than content by "educators."
Don't worry. I chose the term "educator" over "teacher" or "expert" for that very reason: namely that neither having a degree in Education nor a PhD in physics makes you a particularly suitable physics educator. KA needs people who can do for biology, neuroscience, history, and so on what Khan can do for college physics and maths.

I suspect that they're going to be PhDs and practitioners as opposed to K-12 schoolteachers -- but I really don't give a damn about credentials. There are fantastic high school teachers out there. There are PhDs and masters in their respective fields who are ridiculously bad educators, especially at the level of teaching that KA provides.

That's a worrisome attitude if your goal is to make the best site possible. Great sites aren't made by people that get defensive in response to user feedback.

Rather, that's a recipe for stagnation and complacency.

His response seems reasonable in the face of factually incorrect user feedback based on widespread misconceptions. Don't overgeneralize.

User feedback: "Materials for topic area X is poor compared to your flagship content."

Employee response: "It's good enough/better than even worse options."

That's a poor response, and not what I would expect from a team with the culture to make a great product.

That's not defensive, that's just a response to the statement...

> His biology stuff, for instance, is decidedly worse than the mathematics.

I dunno.

I've been studying cell biology on my spare time. I've been using The Molecular Biology of the Cell (http://www.amazon.com/Molecular-Biology-Cell-Bruce-Alberts/d...) that in my opinion is a great book, and as far as I know, is well-respected in general.

Still, watching Sal's biology videos has been of great help. Although I think e.g. his organic chemistry videos are better executed (~10 minutes per video, a bit less packed, better splits by content), I still think that his biology videos are a super-valuable resource. I've been watching some MIT lectures too, but visualizations, which are necessary for biology, aren't properly visible in them. Of the free biology content in the web, Sal's videos are the best IMO.

Recent KA hires have included:

Beth Harris and Steven Zucker (http://smarthistory.khanacademy.org/)

Vi Hart (http://www.youtube.com/user/vihart)

Brit Cruise (http://www.youtube.com/user/ArtOfTheProblem)

Do they need to actually hire educators? I understand that it would raise the level of knowledge but if I remember correctly from the last Google Hangout Khan said that they are currently working on a platform to make it easier for anyone to add things to KhanAcademy without having to do github pull requests.

Everyone working there IS an educator.

People saying "they don't hire educators" really mean that they don't hire as many people with Masters degrees in Education, which is known to be one of the degrees that attracts the lowest possible GRE scoring students, and which produces legions of incompetent "teachers" that have created a nation of ignoramuses.

It's a good thing they don't hire more of these folks. They are the ones that have destroyed education in America, why would we want them to have anything to do with Khan Academy?

The average teacher might not be vary qualified, but there are some incredibly intelligent people that work in the education field. The main problem with education in America is inconsistent teacher quality and social issues that take place outside the classroom. So, having people on staff that understand say the cognitive neuroscience of memory would be a great idea, but adding someone that happened to get a Masters because it's a pay bump and they had to take continuing education credits every year anyway is probably a waste.

PS: The Head first series is an example of what happens when you add educators into the mix. They might seem fluffy but they sell well and do teach people surprisingly well. http://my.safaribooksonline.com/book/programming/java/059600...

Regarding the design of the Head First series, neither Kathy nor Bert have education degrees, Kathy's degree is in physiology and then she later switched to programming. Bert does say that he talked to teachers over the years while developing his philosophy of pedagogy, but so have a lot of people.

If you are saying that Kathy and Bert are educators because they educate, then I agree, and that was my point in saying that obviously Khan Academy is populated by educators.

he talked to teachers over the years while developing his philosophy of pedagogy that's actually exactly what I am advocating. I am all for subject matter exports being the people on the stage, but they need to be informed of the basics behind learning. There are many great online lectures, but finding ways to add repetition without simply watching the same video again is critical.

PS: Funding is clearly an issue, but there is a long history of education and some extremely valuable lessons have been learned the ridiculously expensive way. Leveraging that history does not take a lot of resources just a willingness to use them.

  > Khan said that they are currently working on a platform
  > to make it easier for anyone to add things to KhanAcademy
That strategy appears to ignore what made KA as successful as it is -- quality content. His 10 minute lessons on, say, calculus are genuinely well-made and effective; not too rigorous, but helpful and to the point. If they fully open up, KA becomes YouTube with exercises.

Not what I find exciting about KA. Educational standards should be enforced, which necessitates the hiring of educators. I'd be disappointed if Khan falls into the good ol' Silicon Valley trap: believing that it's all about technology and platforms.

I don't know, as if they have a good enough method of evaluating lectures and only promoting the good ones. After all, look at the success of Wikipedia as compared to other encyclopedias. There is an important difference, of course, in that the ways you can incrementally improve a video lecture are much narrower. Still, though, there must be many qualified people who would be willing to record lectures for free, which doesn't have to be a full-time job like the technical side.

Yes, they should. However, I think they should hire doctors of education in the same way they're hiring engineers: the engineers are creating the technical infrastructure to deliver lessons, and the educators should be creating the pedagogical standards and informing the engineers about what kinds of tools to create.

A doctorate in education is of dubious worth. As an academic field education research doesn't yet have the tools of structures to answer interesting questions. In place of rigorous theories there is currently a swamp of poorly verified or unverifiable notions.

Hire teachers, hire experts, hire anyone who either has deep knowledge or experience transferring knowledge. Hopefully both. A doctorate in education, alone, provides neither.

(this comment brought to you by small sample size interactions with real-life Ed.D. dunces)

The most exciting thing about being involved with software development is that it is a practical discipline that has both scientific testing (A/B, unit tests), and fast evolution that is discovering successful organisational structures and working processes.

Education seems to test the wrong things (knowledge, ability at tests, not practical nor personal skills), and changes at glacial speeds (conservative, and aiming for politically motivated outcomes).

I was speaking about people like Dr Stroup or Dr Petrosino (I've had a class with the latter and read papers by the former) who are experts and have written compelling papers. However, you're right, my wording was bad. Anyone with a proven track record or some accomplishment in teaching should be considered. Both of these people have PhDs, I'm pretty sure, not EdDs.

I highly doubt they're proceeding in the dark. While they may not have teachers on staff, I'm sure they're consulting with the brightest minds at the best universities to develop optimal delivery.

Ever read Ivan Illich's _Deschooling Society_? I like his take on this, and it doesn't particularly jive with what you're saying here.

In general 'educators' generally are not so great.

The biggest scaling problem is Sal. IIRC he is the only one making the videos.

There are a couple others now

Craig Silverstein is someone I've had the distinct pleasure of interacting with one several opportunities over the years. More important than being Google employee #1 is that he's THE Google open source person. He's always involved in Google Summer of Code and personally maintains a number of very cool open source projects on Google Code.

I wish him the best!

I'd like to second this opinion. I've had the pleasure of reading and using a lot of his code in the form of the sparsehash, ctemplate and tcmalloc projects and I know for a fact that my own projects would have been a lot harder to write and test without this code.

He's a top notch open source advocate and a great encourager of contributions from non-Google employees, so hat's off and all the best at Khan Academy!

Do you want to sell advertising the rest of your life, or do you want to come with me and change the world of education?

If he's the first employee, doesn't that mean he helped create the Google search engine, and thus changed the world? I think that's a lot closer to the truth than what you're implying. I don't think Google started worrying about advertising until years later.

They also own that thing called YouTube which may or may not have been a part of Khan Academy's success.

And AppEngine, the infrastructure it runs on.

I'd be interested in trying to figure out just how much Google Search changed the Internet, and by extension the world. How different would the Internet look if we were all still using AltaVista? Certainly it would be difficult for quality content to find viewers, as everyone used to pad their articles with thousands of keywords to jack up their rankings. What else would be different?

Google's been a great boon for the Internet, but i'm curious just how much Search changed things, on its own.

Websites would probably be stuck with "pay for inclusion" and all the other nastiness of the early days of search. Google also labeled our ads clearly; the FTC had to warn ~8 other search engines not to label paid results as "partner" listings and similar things. We also took a stand against pop-up ads in the early days when many other major search engines were doing pop-ups on their site. Not to mention that search engine spam was pretty bad in those days.

Certainly I'm glad that Google disclosed DMCA complaints (instead of dropping them, which every other major search engine did), push backed on overly broad DOJ subpoenas that tried to get 2 months of user queries, worked hard not to partner with scumware/malware companies, helped to push back on things like SOPA, and launched a transparency report to shine a light on government requests to take down information around the world.

Google has also been a major proponent of open source (e.g. Summer of Code, Android, Chromium), not to mention espousing principles like data liberation: http://www.dataliberation.org/ . Overall, I think the web would have been quite a bit worse without Google: slower, less organized, more closed, and definitely spammier.

I remember that when I was doing my startup, it occurred to me that I simply could not do it if Google did not exist. Several times a day, I'd Google for obscure library error messages or documentation for obscure libraries. Heck, I wouldn't even have known that those libraries existed, or that what we were trying to do was even possible, without Google.

One of my colleagues wrote this in the fanfic style: http://www.distilled.net/blog/seo/the-internet-without-googl...

We don't know how things would have played up in that alternative future...

This is a different topic but I think that while PageRank was a much better form of text search, the impact was minimal. Text search itself is quite silly when you think about it, and in the coming decades when we look back, I doubt if Google will be remembered at all.

Don't read into it too much, Google's mission of organizing the world's information is noble, but you gotta pay for it somehow.

And Khan Academy will probably eventually have to start selling advertising or user data or something similar too. Nothing is free.

My comment probably should have been downvoted, it was off-the-cuff and a bit snarky. Google is nowhere close to Pepsi.

Is that really a fair comparison? Soda literally is just sugar water. Google provides lots of really good and useful products that people genuinely depend on and they push technology forward.

For that matter I bet there's an awful lot of science that goes into that bottle of Pepsi in your fridge. Everything from the chemical formulation of the sweeteners and preservatives, to the purification of water, to the design of the bottles for maximum strength per gram of plastic. And then there's the whole organization of the distribution channels and salesmen to get Pepsi product into every restaurant and convenience store in the entire world.

Well, combinging ads and educational content is a whole new area.

I am going to make a wild prediction right now. Taking into account that 2012 seems to be the year of the "legit" free online schools. I am going to predict that within 5 years a person will be able to get a high-school or college degree (possibly accredited) from Khan Academy...and it will be recognized as legit by peers and employers alike.

I think that is a fair prediction.

A follow up to that is to ask whether accreditations mean anything at all at this point. A high school diploma from a public school is known to be not worth anything presently as it doesn't even certify if the recipient is literate or knows basic math. Employers and colleges know better to rely on a high school diploma as evidence of anything, this is why they still have to test for skill level.

Are college degrees worth anything either? That is uncertain. Few employers validate whether someone really has the degrees they claim. Tech shops don't accept a bachelors in lieu of a programming test during an interview. Whether one has a degree is considered quite irrelevant since it is known that there are developers with no certifications or diplomas whose talent and skill surpass that of Computer Science diplomates from the finest institutions.

Why even bother with credentials with they don't really indicate anything useful.

In my 15-20 years of work history, to my knowledge, I have never had an employer confirm my academic credentials. Credit score? Yes! Criminal history? Yep! Whether I am academically qualified for the position? Nope! In particular (in my younger days) I "inflated" some of my academic credentials and achievements. They were simply accepted at face value.

Education credentials (especially college degrees) represent the ability to apply yourself towards, and accomplish, a stated goal over the course of multiple years.

I'm not sure about your prediction.

I have been hearing impaired since birth, and I wear hearing aids. However, I 'evolved' to lip-read and read body language. School was invaluable to me, for making friends, figuring out how to deal with jerks and asking lecturers and tutors the right questions.

Not until I was in my 4th year at uni that I realised something was _very_ wrong - how did I know what to ask, if I didn't realise I'd missed it? :-)

These days, I am in a programming job, however I quite often hear something and go, "what the hell was that". However, with Khans offering captions for everything I pick up such tiny details that suddenly blow my mind and I feel very humbled. It feels like I'm learning backwards.

I hope that in the future, people don't rely on online classrooms, because then they will miss so much other things, that they simply won't realise until it's rather late.

"I have to... TALK to my colleagues?!"

I'll take that bet. Care to put your money where your mouth is?


the change starts with us

if you have a startup looking for engineers, I would troll the reddit boards devoted to those classes and see who's interested in joining you

you might be able to come up with some pretty good underpriced talent

I hope they don't try for traditional accreditation.

I think the "recognized as legit by peers and employers alike" is the important part

I think an alternative to the current model of getting yourself deep in debt to prove you're employable would do society alot of good

Well as I understand it, one could study on Khan Academy and use that to get a GED, right?

Then you would only have to be in school when you took whatever test was necessary to get that degree.

Although you could conceivably try directly for university and skip High School completely. Or even skip university, learn to code in your basement and be done with it.

Why not? I would hope for some sort of certification program.

To me, Khan Academy represents a break from the traditional educational model. I would rather they focus on innovating than proving themselves as being just like every other school.

Is this another sign that the new sought-after perk is meaningful work? The time of beer and ping-pong tables maybe coming to an end?

I sort of see this as the natural progression of people's careers

Totally - if the supply of talent becomes more: productive / wide applicable / sought-after - it only makes sense that they'll move up Maslow's Hierarchy into the highest fulfillment: meaning itself.

Absolutely amazing for the edtech space to see high profile, tech elite join the industry.

Wow, First John resig and then This, salman Khan must have some great charisma

I think it is more the desire to have an impact. Seriously as you get older and interact with more and more people you see the choices and opportunities that lead them to the current space-time co-ordinates. And you get a little bit sad sometimes that folks who could have been great have become de-railed due to some addressable issue into crime, or poverty, or drug addiction.

It is hard to sit at a company working really hard to fill jobs and to listen to all the folks who are out of work but not qualified for your jobs. Its hard to listen to the stories of good, decent people who work hard and are certainly capable, unable to break the chains of poverty because they can't both get an education and stay off the streets. Its hard to listen to young people, now young adults, who suddenly realize that it wasn't learning they hated it was the school and now they are trying to make it as adults without a solid foundation.

KA is has three things going for it, the founder is passionate about the cause (so its not going to sell itself suddenly), the technological sub-pieces have finally come into existence across a critical mass, the availability of 'self funded' (which is to say folks who no longer have to work if they don't want to) individuals to volunteer efforts for it.

So being remembered for lifting some fraction of the world (no matter how small) out of poverty by providing educational opportunities, or being remembered as someone who worked at one of the companies that formed the Web as we know it. Easy choice.

I think it's the cause that attracts people. Silicon Valley repeatedly talks about Changing The World and all that people do in the end is to create another web app. Khan Academy gives a lucky few the means to truly change the entire world by working on education, one of the most basic aspects of modern societies.

I wonder if the chance to work for a good cause is a natural filter for exceptional engineers.

When you've already made millions, or when your reputation allows you to get any job you want, money surely becomes a much smaller work incentive.

Or these high profile people really believe in the idea of KA

What they're doing is what carries the charisma.

I get that the cause itself is part of the reason, but there are plenty of opportunities that have great causes that could need help of aforementioned individuals. I really think its Salman Khan, because people believe in him. I can see him being the next leader in SV

I hope they can make learning biology, calculus and programming as hours-consuming addictive as modern first person shooters are to teenagers.

Guidelines | FAQ | Lists | API | Security | Legal | Apply to YC | Contact