That is great news Sam! This is really exciting to see an incubator starting it's own research lab and on top of that, keeping it totally in the open source. I'm fascinated by Research work folks have done at Parc and Disney, you should have a look here and perhaps speak with someone at Disney as well: http://www.disneyresearch.com
You are making a wrong correlation. Putting more money won't necessarily drive research forward. Most breakthrough researches and inventions were done with little money.
I am a professional scientist and I have to disagree. 10 million is just seed cash that I'm sure will be magnified. But do not underestimate the cost of research. While a particular project that pays off seems (and is) a great return, salaries, equipment and conference travel consume a lot of money.
It probably depends on the type of research that you do. Drug research is expensive. But I think that you can do something like Xerox PARC (it seems like YCR might be going for something similar) with relatively little money. It appears that in 2001 their budget was $65m (http://abcnews.go.com/Business/story?id=86872). I'm not sure if their budget was larger during the 70's but that's not that much money considering what came out of there.
From what I can tell, back in the 70's Xerox PARC's annual budget that focused on computing was $10-14.5 million in today's dollars, which seems fairly cheap. But it took multiple years to develop it.
This also feels like cheating because we're targeting extremely successful research. If we knew ahead of time which research would be successful it would be a lot cheaper to do.
But it is inevitable to have a lot of unsuccessful research projects before one breakthrough. Similar to startups where most fail. Those "failures" add up to the cost pretty quickly, and you cannot avoid them.
For now they are going to be hand-selected by me (the area we are starting with requires pretty specific domain expertise). Not sure what we'll do in the future.
@sama - Interesting, I definitely think the toss up between meaningful research and high paying engineering work is very broken. You mentioned how compensation will not be determined based on conferences and papers, do you have any specifics on how you would measure "quality of output"? Will researchers be compensated differently based on said quality?
(As often times the quality of research output is hard to gauge immediately without the greater context of time passing)
ARPA-E (The Advanced Research Projects Agency for Energy, http://arpa-e.energy.gov/) has some interesting models on how to encourage meaningful innovation tied to continued funding. Might be worth a look. Same with DARPA.
Nuclear makes sense. The fluid-fueled nuclear company he's the chairman of (Transatomic) is considered at least slightly heretical by the nuclear industry, which is used to its solid fuel water-cooled reactors. They also use used nuclear fuel (i.e. transuranics in nuclear waste) as fuel, which could be called heretical because recycling nuclear waste was once outlawed in the US by Jimmy Carter (but it no longer is illegal).
I think $10M can go pretty far in basic feasibility studies needed for the Transatomic reactor. You need a bunch of diffusion couple experiments and hydride stability studies. Lots of furnaces, fume hoods, and surrogate fission products. But you're right that when you get to major component qualification then yeah, $10M is nothing.
He's also chairman of the board for two nuclear power companies. And YC itself has been branching into biotech and hardware and other science-y companies. Anything seems possible to me.
A 25-year estimate for how long it will take them to develop an AI that will be able to create biological organisms that utilize nuclear power directly? Sounds reasonable.
My money's on nuclear. Advanced nuclear designs regularly claim 25 year timeframes for design, qualification testing, licensing activities, supply chain development, and construction. Transatomic is proposing concepts that require long qualification and licensing developments, many of which could make meaningful progress in a small lab with a $10M initial budget. You have to produce your salt fuel and do materials testing with surrogate fission products (e.g. stable isotopes) over long periods of time. They also have to do hydride transport experiments which can be done on a pretty small scale.
25 years is a reasonable number for how long it took deep learning research to begin getting traction. I would think AI would be a strange choice since it seems like there are a lot of research opportunities in academia and industry for this topic, but for the same reason I would say AI research is ripe for disruption.
My bet is on renewable energy/sustainability, agriculture or biotech, with energy being my first choice. Sam has expressed serious concern about powering the future. His first(only?) deviations from the "No Board Seat" policy were two energy companies.
Safe AI [1], on the other hand, has [2] running on < $1M and [3] running on $10M (!) from Musk. Throw in Yudkowsky's recently expressed interest in startups [4] and you can guess what I am guessing.
I think "we’ll have a process in place to address technology that could be dangerous" is a pretty strong hint, but it would apply about equally well to AI or nuclear, which are already the top guesses in this thread.
"For now they are going to be hand-selected by me (the area we are starting with requires pretty specific domain expertise). Not sure what we'll do in the future."
Two words: baseband processor.
The world really, really needs a fully open baseband processor.
It's true that we need an open baseband processor, but I wonder if this qualifies as the sort of blue-sky research that I'm guessing they'll want to focus on?
My feeling is that the software to accomplish this is already largely there, but the hard part of getting a design to the fab and actually getting anyone to adopt it isn't.
I think there would also be a number of regulatory hurdles to cross, since this would end up being essentially an SDR, and the FCC (for example) already seems to fear wide public availability for those.
They only care if you can make it work on channels it shouldn't, they don't care whether the code that controls it is open or not.
But yes, an open processor could help with a lot of open networking problems, even though it seems more like an engineering matter than research per se.
It really does. But that seems like a mission for a startup, not a research organization. The challenges are more engineering and business than science.
If someone wants to apply with an open baseband processor startup, applications are due Oct 13!
Hand-selecting a bunch of famous folks results in IAS Princeton effect which has the criticism of producing nothing significant. Usually the approach of hiring a towering/inspiring figure with an emerging research view works well (like FB did). They bring along a bunch of rookies and experts, and set up research groups quickly and rather smoothly.
FWIW, as a researcher at a university, it is only my decision whether I apply for a patent or not. In contrast, patenting inventions is an inherent part of the research workflow at research labs such as MSR, Yahoo Research, etc.
MSR has published lots of great papers but it has not yet replicated the success of PARC or Bell Labs by any means. I am not sure why this has happened, but it has been pretty obvious to everyone in the research community. Google, which came in with a research agenda pretty late, has more or less single handedly pushed the boundaries of industrial research in the last decade.
Eh. That a pretty huge jump. Look, most new "tech" companies research is mostly PR, internal and external. If you're doing something you have little experience in and therefor can't evaluate, which success you aren't dependent on or necessarily benefit from and everyone will say how cool it is regardless then it's extremely hard to not cut corners and achieve real success.
Still exciting, but I wouldn't underestimate the challenge.
MSR has been around for a while, with quite a bit of freedom, and still is a far cry from the accomplishments of Bell Labs (maybe time will tell, though.)
It is possible to structure it as "patent granted to everyone, but expires the moment they bring up a lawsuit against us or any of our licensees" to do doubly good for the world.
YCR is a non-profit. Any IP developed will be made available freely to everyone.
We’re not doing this with the goal of helping YC’s startups succeed or
adding to our bottom line.
you could allow free use of patents under a license that requires the entity doing so to also allow free use of their patents. If you believed the patent system is largely counter productive (as I do), this would be an awesome way of trying to address this a constructive way.
what YC could get out of it ( besides making the world a better place ) is very good knowledge of the capabilities of potential entrepreneurs. Imagine one of these researchers then founding a startup. Funding that would be a lot less risky, so that's quite a bit of edge
This is great. I am hoping things like this will shake the dynamics of publication-oriented credentialing and ultimately the value proposition of academia/academic research.
We talked to Alan at great lengths in the process of putting this together. He is the most insightful person I've ever met on how to structure an organization for great research.
In the article, he suggests finding Michelangelos. I imagine there are many who are clearly the right type, but many more non-obvious Michelangelos. Do you have (or intend to discover) a reliable way to foresee the best people even while circumstances influencing their lives have diverted them from the traditional outlets to look for such individuals? I assume there will be enough obviously great candidates wanting to "work" at YCR that you won't need to dig for hidden gems. I was just curious to know if it was something you'd considered.
Sam, when I met you at Startup School last year, I wanted to be brief. The most important thing I wanted to communicate to you was my gratitude, and I thanked you for sharing your startup class. I'm blown away by and so grateful for the stuff you've been doing since. I'm very excited to see how these new projects work out. Thank you again man.
At first I thought about that too, but as I got further I was reminded about an article about the person to first theorize the Higgs Boson, and how he had only published a few articles in his entire career.
> We’ll especially welcome outsiders working on slightly heretical ideas (just like we do for the startups we fund) and we’ll try to keep things small—we believe small groups can do far more than most people think.
What's the process for an outsider to get involved? This sounds like a really cool and much needed movement. Will more information be coming soon on how to apply/join/contribute?
This sounds interesting. I'm curious to see what areas the research will be focused on.
Are there plans to provide laboratory space for areas that require it? Would researchers be housed in some space in MV, or would they work out of their own space like YC startups?
The first group will be in SF. TBD on other groups. But unlike startups we fund, these people work for a YC organization, so we will provide space etc.
As I keep chiming in whenever YC and location come up - please seriously consider Australia's East Coast. Like the valley, there is a big supply of world class universities and their graduates, and there's also great weather and quality of life that keeps people from wanting to leave. What's missing is an American style risk tolerance amongst the holders of capital. That's changing, but it's something YC could change even faster with an organisation like YCR.
Our government has historically underfunded research, and even more often one side funds it then the other gets into power and cuts it again. As a result, there are lots of ex-academics looking for more interesting ways of earning a living!
Yes, but we're coming in on the very high end of the comp scale. We're also considering doing things like YC guaranteeing 100% mortgages for the researchers--housing prices in the bay area are a joke.
We're starting relatively small (on the order of 10) and grow from there at a reasonable rate.
While I'm somewhat skeptical of the idea (I'll admit I'm an academic researcher) this seems like a very good step - the Bay Area is becoming cost prohibitive.
I like the idea of YC funding research. Bias free and politics free research can go a long way. A couple of questions though for @sama:
1. Where will the research be based? Is YC providing space to setup lab?
2. One of the problems I see with this is lack of fellow researches present on site making casual collaboration harder (compared to a research university campus). Have you thought of that?
3. Are you planning on funding researches with track records or people with high risk/high reward ideas that are relatively new?
Interdisciplinary work is incredibly hard to pull off in academia due to tribal affiliations. I think the MSR NYC lab (where I work), which consists of computer scientists, economists, and other social scientists, is one of the few places I've seen it actually flourish.
I don't think interdisciplinary work has to be between fields that disparate to begin with either. From personal experience, I think there are definite benefits from just having different fields of engineering collaborate and share ideas.
1. Try set up a lab in Cambridge at some point down the road.
2. Pay particular attention to CS theory: it's (arguably) high-impact and very cost effective --you can practically fund an entire generation of researchers in a subfield for a fraction of the cost of, say, a major interdisciplinary initiative. See for instance what the Simons foundation is doing at Berkeley[1].
Best of luck with what you're doing. You can end up actually changing the world in this way.
Can't stress this enough, but a lot of TCS research only becomes valuable once you put it into the wild, and that takes a lot of interdisciplinary work.
A lot of the stuff in TCS is theory for the sake of theory, and one would want to avoid too much of this.
Serious question: In what ways is CS theory high-impact?
I'm particularly interested in any arguments that complexity theory is high-impact (beyond the very useful insight that there are some problems for which no polynomial time algorithm is known). I have a pretty good idea of the impact of cryptography and randomized linear algebra (sometimes considered CS theory), but am also interested in hearing about other fields considered CS theory with useful applications.
type theory has broad applicability, since it serves as the basis for all modern programming languages design and research. another nice example is kripke logical relations, which can be used to give a model for a programming language in which you can prove certain safety lemmas for critical code.
Graph theory is a good example of this - the asymptotically fastest minimum spanning tree algorithm was made possible by Hopcroft and Karp just drawing weird data structures on a chalkboard until union-find popped out, which gives you near-linear time MST.
It's not really that different from work in another corporate research division (IBM, PARC, MSR, Bell Labs..). Even in academia you are still getting paid, and your work is influenced by funding frameworks, citation index and faculty interests.
Peer review has long been the traditional validation mechanism in science. I don't see why it wouldn't be applicable here.
Because as much as it is true that funding influences how some projects get started in academia, professors do get tenure. And they can and do things which would be totally offensive to a corporate parent. There is a totally different kind of publication coming out of even the old research labs and an academic institution, much less a modern corporation. A typical research lab does not need to worry about a bottom line - which a company (like YC) is obligated to care about.
Interesting. I am curious though, as to why YCR researchers will also receive equity in Y Combinator as part of their compensation? Especially given that both are organized separately and YCR discoveries won't be funneled through or used in YC. I am also assuming that YCR will pay market rate salaries.
We want to get the best people in the world, and I think YC equity is a pretty compelling thing to get. We're competing against e.g. Google, and Google stock is also a pretty compelling thing to get!
Gotcha. So this is more about getting brilliant minds in media res, as opposed to discovering fledgling great researchers of tomorrow. Good lure; YC equity is probably increasingly more compelling than Google's!
Seems like an effective system-level path around tech-transfer and licensing headaches. Open inventions feed innovative companies and back into research. Well done.
Just replace "flux" with "flow" and you get the picture. If you're at YC you (presumably) encounter lots of different problems in lots of different contexts, which is often a good environment to foster creativity and inventiveness.
Problem flux means the rate of new problems that need solving.
Theoretical breakthroughs are often inspired by real problems. For instance, information theory arose from practical problems of wartime communications. YCR will be in close proximity (geographic, organization, and intellectual) to lots of startups with interesting and worthwhile problems to solve.
@sama is this targeted towards pre-PhDs who are deciding between pursuing their PhD or working in industry, as well as post-PhDs who are full-time researchers?
Not sure if this is driving your thinking but this would open a desperately needed alternative to Academia if you can scale this idea as successfully as you've scaled startups.
I've watched my genius brother-in-law (PhD Materials Science & Chemistry / Biomimicry) be consumed by a very broken UC system over the past decade.
Started with an exceptionally bright, curious, and inventive man who created breakthrough science in self-healing materials for the "benefit of the world"
10 years of grant fights, personnel struggles, under-served licensing resources, conflicting lab-vs-student priorities, etc has put him out the other end hopeless and disenfranchised.
The world will probably lose one of their best "public" researchers to the bowels of commercial science as a result.
Go for it Sam, good luck and excited to see what your first projects are.
Seconded. This is brilliant. As a physics PhD student at MIT who's leaving to build a startup, I'm finding it a relief to leave a broken system.
The academic community is fraught with politics, mismanagement, and inefficiency. Researchers are driven by a need to get published and get grant funding, not to expand the frontiers of human knowledge.
Impact is currently measured by citation pagerank, and broader impact is undervalued. Outreach efforts (blogging, education) are consistently stigmatized [1]. This needs to change.
I would love to see more transparency and accountability in research. I've seen Nature papers continue to be cited despite completely wrong conclusions (the 'erratum' was published as a separate paper in PRL). The arXiv was a great first step at opening up access to papers, but the large publishers continue to hold cartel-levels of control over scientific knowledge.
I'm super excited about YC research. Would love to see a quantum computation division!
Wow, I knew things were bad, but the education system must be well and truly broken when the students at one of the world's top research universities are leaving to build startups, and calling it a broken system on their way out. (No sarcasm intended; I couldn't find a way to say this better).
It's not necessarily the 'brokenness' of the system that forces people out but the consequence of producing PhDs faster than new positions are being created. "You don't have to go home but you can't stay here". No one expects that all high school students will become high school teachers. Why should it be the same for PhDs?
A lot of what is wrong with the academic system is this expectation, which is bewilderingly prevalent in both the student and faculty bodies despite abundant evidence of arithmetic skills. Faculty are often quite unaware of what happens outside their building, and students are, well, students and absorb a lot of the mindset of their role models. So there becomes a glut of post-post-docs applying for a scant few professorship, which makes getting hired and then getting tenure ultracompetitive and stressful--so many people want the jobs that the administration can ask just about anything out of them, including pretty unrealistic expectations, bad working conditions (80 hr/weeks for 75 k/yr in the sciences at many good schools) and so forth. The conditions of tenure-track professorships (when professors are still young enough to be most creative and energetic, and are still skeptical of their fields' dogma) are poorly suited to creativity and focus, so their grad students are tasked with the actual science while the professors become managers.
This isn't to say that academia is actually 'broken'. If my coffeemaker is broken, it doesn't make coffee. Academia is still making the things it is supposed to, but never in the amounts or quantities or prices that everybody wants (as it turns out everybody's wants are wildly divergent). So for this and other off-topic reasons, it's not exactly broken but definitely not optimal.
I'm of the opinion that a more optimal academy would do more to prepare grad students to find their own path, be it in academia, industry, or elsewhere. Sort of like what the US liberal arts education is supposed to be, but at a higher level and with the requisite specialization as is fitting a PhD. The critical, analytical, creative, scholarly, and technical skills that grad school can (and often does) bestow or hone are a great fit for a variety of pursuits, including entrepreneurial ones. But there is little guidance here, in part because the faculty themselves have little idea how to go about doing anything but what they do 80 hours a week.
I left the academic track after a post-doc to become an independent researcher (doing funded and unfunded research as well as R&D consulting) and it's been really surprising to see how many professors think it's an unworkable idea. Not an idea that many people try and flame out, but one so unthinkably scary that no one has seen anybody try. Which smelled like an opportunity to me, and still does...
When I was in grad school I think most people were aware of how tough it is to get a tenure track faculty position. Many students had no aspiration of going into academia and planned to go directly to industry. This was in mechanical engineering, an inherently practical subject, so maybe the field makes a difference.
Kudos on trying to make it as an independent researcher though. That's a really cool idea. How has it been applying for grants without a university affiliation? What's the breakdown of your time spent on funded research, unfunded research, and consulting?
I've registered with SAM (System of Award Management) so I can get grants directly. I still collaborate sometimes with academics, writing papers and grants, and minor consulting for them. I'm also still an affiliate of my post-doc school so I have an email address and journal access, but I think you can basically buy library access from a school for a decent price.
I spend about half my time or a little less on consulting, but it's been very research-oriented consulting so far, with few restrictions on publishing results. There are and will be ups and downs (most of my consulting is in energy) but I bill a lot more than I pay myself so that smooths it out a bit.
So far most of the non-consulting research has been unpaid but I try as much as is practical to get funding for it at some point--make a minimum viable product to write a grant upon and if it doesn't work out then make it publishable and iterate or move on. I like strategy, it combines well with dreaming. I have some moonshot ideas too, but most things can be broken down into units digestible for peer /grant review.
I'm finding that there are a lot of theoretical problems that are also practical problems. I also feel like I am doing research all the time, but a lot of time is spent building tools and skills to do new things, and less of my time is spent actually learning new things about the earth than I would like. But I guess that's research...
You've said it better: it's not completely non-functioning, but just far from optimum. I've learnt so much from grad school: I wouldn't exchange my years here for anything.
Every time I explain my experience in academia to VCs or those in business, they don't believe me. "It can't be that bad!" It is. I've worked at 3 of the top schools in the world and it's a total mess as the OP mentioned. It's also a mess that hurts everyone, so I'm happy to see this type of initiative.
I'm in a similar situation and I don't know a single person who doesn't think academia is horribly broken. Internet rants from people leaving academia are practically a monthly occurrence these days, and nobody ever disagrees with them.
In fairness, all systems are broken, academia isn't much more broken than anything else involving large numbers of mortals.
The basic problem is that we need to rank one researcher's merit against everyone else's, and there's no perfect way to do this because nobody can familiarise themselves with every individual's research. Rely on peers' opinions to measure merit and you just reward back-scratching. Rely on objective measures like citation counts and h-index and you incentivise people to focus on these instead of research quality.
The thing is that macro-scale works pretty well. As an overall system, academic science produces new knowledge at an astounding rate, and in terms of dollars in to research produced I'd say it works pretty efficiently. The measures we use are correlated with research quality even if they don't measure it perfectly.
It's only at the detailed scale where it looks badly flawed.
edit: and of course whenever I spend a long time writing a thoughtful comment I get the "you're submitting too fast, please slow down" message. Time to leave this comment to sit for an hour before pressing reply again...
> Rely on peers' opinions to measure merit and you just reward back-scratching. Rely on objective measures like citation counts and h-index and you incentivise people to focus on these instead of research quality.
In the meantime, Google seems to have developed a pretty good ranking system that is difficult to game.
Don't citation indexes work pretty much exactly like Google's PageRank, with cites taking the place of href links? A parent comment seems to suggest the same thing, so it looks like the ranking is gameable.
In practice I understand this happens via citation-swapping (similar to link swapping, or paid links in blogs) and poor-quality journals accepting articles for a fee, or without review (much like paid content, or content farms) and papers that cite other articles just for the increase in rankings (similar to link farms) so the analogy to SEO (in its various hats) seems to fit.
This would suggest that the only way to fix the 'broken' citation ranking system is the same as Google, and bolt on empirically determined heuristics above the pure mathematical model of PageRank, and black-list or penalise certain journals or types of citation?
> and bolt on empirically determined heuristics above the pure mathematical model of PageRank
The impact of Google doing this is, in financial terms, probably much larger than if the same type of approach would be applied in the academic situation. So if Google is allowed to do this, why can't we allow the same approach for research? Apparently, the system works quite well.
Honestly, funding a research team to develop a science ranking engine that was both being continually developed to thwart gaming and effectively independent of university and science-funding politics would be a truly excellent use of YCombinator's money.
Indeed. Simple link counting based search algorithms got gamed a lot in the past, by regular website operators and most heavily by spammers. That's partially why the current search algorithms are much more clever and secret than just using the Page rank. Human raters and machine learning off the data they generate are used as well. This helps to make the rules more uncertain and nasty SEO tricks much harder to discover and use.
Compare this to the situation in rating of academic research, where the papers are rated by the community of authors itself and the ranking is essentially a weighted counting of references. Of course it leads to lots of self-serving spam.
If the idea of independent raters (and possibly machine learning on top of it) could be implemented for academic papers as well, that would be a big improvement, I think. Academia has got a lot to learn from search engines.
The problem is not that it is gameable or not. Even if it is, most researchers won't try to game it. The problem is that they will focus on optimizing these counters by necessity (they have no choices, it is that or goodbye), rather than doing what they want to do: great research.
I wouldn't call prioritizing quantity over quality gaming the ranking system, it is merely playing by its rules. I would call gaming it if there was some cheat (like you cite me and I cite you even if really we shouldn't), which may happen but I believe is rare.
It seems you have reduced my argument to the last part only. Google's system is pretty good even if you don't focus on gaming. By the way, even when it comes to websites, most users (admins) don't try to game the system.
>As an overall system, academic science produces new knowledge at an astounding rate, and in terms of dollars in to research produced I'd say it works pretty efficiently.
I wouldn't call throwing out your trained experts and starting with fresh trainees in a five-to-seven year cycle all that efficient.
Many fields of work don't pay much, but tell you you've got a Mission. Many fields of work make you work very long hours. Many fields of work require extensive training and skills.
Academia doesn't pay very much, makes you work very long hours, and requires extensive training and skills, justifies it all by reference to Mission, and then inevitably fires you and tells you getting fired was your fault for not being in the top 8% of the entire profession.
Top research universities have more money, and smarter/harder working students. Neither of those are necessarily solutions to the problem of university research.
Don't get me wrong: our lab makes awe-inspiring machines that can see individual atoms held in vacuum by lattices of laser light. Our research leads to incredible discoveries, but we're held back by inefficiencies. It's a rusty system, but to any mechanic, rusty = broken.
Also, leaving to build a startup is rather uncommon at MIT physics, and most students (including me) absolutely love research. The academic career is just not attractive enough to spend another 4-12 years in a lab becoming increasingly good at doing exactly one thing.
Academics isn't really that bad. It's full of many of the smartest people in the world trying to do things which are fundamental but aren't seen by the market as important and can't be funded any other way besides charity. Sometimes it doesn't go that smoothly and there's some bureaucracy like in any organization (people are constantly fighting very hard to cut what little funding there is so academics have to fight over it) but it's really not all doom and gloom like some people make it out to be.
Graduate students tend to love to complain as much as possible about things that aren't that bad (I know I was one) and people leaving something idealistic often have a compelling need to try and put it down in any way they can to justify their decision to give up on the idealistic goals and try to make some money -- which is why you see so many "goodbye academia, you are broken" posts and blog entries. And some people have genuinely bad experiences like at any job -- bad manager, particularly low funding, bad co-workers, you're going to have a bad time.
That being said YC research sounds amazing. Basic research was funded by charity from the superwealthy (or only carried out by the independently wealthy) for most of human history, the model works. Researchers tend not to look into areas or find results which are bad for their backers. It's good to have some public and some private sector; we need to balance out the conflict of interest a bit.
Graduate school was one of my best experiences, and I wouldn't trade my years here for anything. My co-workers were fantastic, and my advisor is golden.
My complaint is with those inefficient bureaucratic systems that you mention in passing. Life's much easier without them. :)
I think there is a big disconnect between industry and academy. When I was a grad student (got a masters in CS) there was a huge lack of problems to solve at school. I went to Sun and after about a year I had 9 fantastic PhD problems.
Be nice if we could fix that. School is great, industry is great, somehow connect them.
I've been wondering if academia is really fundamentally broken, or whether it's just horribly distorted due to some combination of underfunding and overpopulation, resulting in way too much competition for scarce resources.
I was reading the article, and it made my blood simmer, as articles like this do. Then, I looked at the comments which referred to Sean Carroll, and I couldn't believe it. The fact he claims his textbook, a fantastic intro to GR and now a standard accessible text, was a mistake with regards to his life in academe is so disheartening, it makes you wonder how science actually continues in the present, hideous environment.
Totally agree. Just wanted to stress how the grant funding systems is totally nonsensical. While as a PhD you could get funding for your entire 3/4 year program (at least here in the UK), if you decide to go for a post-doc things get dramatically complicated as your employment is (usually) tied to a particular grant. When the grant expires, you're out if there isn't another source of money. Is that a decorous situation for a highly skilled person who devoted much of his life to studies? Not sure about that.
It really is nice to see an organization like YC take an interest in research. I applaud Sam for making that considerable donation and hope the outcome is better than expected.
The amount of financial and administrative skullduggery within the UC system 2009 (after the financial crisis)-present (I was at UCSB as a grad student 2009-2011) is almost uncommunicable to people who weren't there. People simply refuse to believe some of the stories I tell them.
The best story I heard about this was from a math grad student / instructor about how you had to jump through multiple hoops to even re-up on chalk for the blackboard.
When I was at Michigan, if someone brought cookies for a seminar and wanted reimbursement then anyone who wanted a cookie had to sign a sheet and the cookie bringer had to enter it all into Concur.
I think there is a bit of exaggeration on the non-availability of grant money for research in applied science/technology. @sama acknowledges the adequacy of funds for applied research explicitly, "[1] Funding for technological development is actually relatively high, but funding for fundamental research keeps getting cut. Investors want to fund incremental progress—and the world has gotten very good at delivering that. This is more valuable than it sounds; incremental progress compounds quickly." This is true of industrial research and to an extent, academia.
It conforms with my experience.
That said, the skills required to write great grants and schmooze with grant-givers are quite different from that needed to do quality research in science or engineering. So the academic researcher community in science and technology is whittled down to the small set of {those good at research}∩ {those good at writing grants} + ⊂{those good at writing grants} + geniuses
They are losing me. Almost 15 years and twice as many grant submissions. I'm writing a grant atm (a fucking good one), and I hope not to be around to be turned down for it. I'm literally pitching investors between experiments.
Reminds me of a great historical figure of Hungary, István Széchenyi.
>Széchenyi gained a wider reputation in 1825, by supporting the proposal of the prime minister, Nagy, to establish the Hungarian Academy of Sciences; Széchenyi donated the full annual income of his estates that year, 60,000 florins, towards it.
>The height of the bubble was reached in the winter of 1636-37. Tulip traders were making (and losing) fortunes regularly. A good trader could earn up to 60,000 florins in a month— approximately $61,710 adjusted to current U.S. dollars. With profits like those to be had, nothing local governments could do stopped the frenzy of trading. Then one day in Haarlem a buyer failed to show up and pay for his bulb purchase. The ensuing panic spread across Holland, and within days tulip bulbs were worth only a hundredth of their former prices. The tulip bubble had burst.
Florins appear to be pretty steady with the dollar...
I don't want to minimize the awesomeness of the YCR proposal, but $10m isn't really very much in the scheme of research. NIH alone funds to order of $25bil a year in science research. Public-private funding schemes probably fund an order of magnitude more every year.
$10m is pretty much just enough to fund one reasonably sized group, doing advanced research, for about a year to a year and half. Seems to me that Sam is hedging so YCR can do its first round in about a year and get a bigger ask by pointing to the starting groups achievements.
Alan Kay's group got $5M from NSF to "reinvent computing". The money lasted 6 years, in which they produced a reasonable prototype (called "Frank") of a complete personal computing system in only 20K lines of code.
Money in theoretical CS research goes pretty much entirely to overhead, not much goes to reagents / processes. Paying a little under a million a year in expenses isn't surprising there.
Holy crap, that was some kind of read! Thanks for telling me about this as I had never heard of it. My proposal, similar to what Racket and REBOL/RED attempt, was to combine simple DSL's on top of a LISP and low-level 3GL (eg high-level assembler) with optimization, safety, etc built into translation process. Seems like Kay already did most of that with way more clever abstractions and techniques. Gives me hope my smaller proposal might go somewhere esp as I wanted to build components & servers not desktops.
If you watch any presentation from Alan Kay from the past few years, it looks like he's using PowerPoint or something similar. But, occasionally he does something PP can't do, like run a physic simulation or have all the letters in the text do a swarm animation. He's using Frank.
Thanks for the tip. I'll try to watch for that. Confirms it's already practical for at least graphic presentations. Meanwhile, I wonder what the odds are of getting a hold of Frank to play with it or build on it. Might have to try to contact him.
Note: I thought about doing something similar in first secure desktop I construct where I do my presentation on it, let people counter the methods probably won't work, and then tell them after that. Always priceless faces when that happens. Also helps tell the contrarian bullshitters apart from those who think hard on your claims. Saves time in future. ;)
It may not be a lot in terms of research, but it's certainly a lot in terms of a voluntary personal donation to anything, however much money someone has.
>$10m isn't really very much in the scheme of research. NIH alone funds to order of $25bil a year in science research
Sure, but the NIH is funding near to 50,000 competitive grants a year. Some googling online gives median/average grant size at $300k/$350k. Even $25bil over 50,000 projects is $500k a project a year. An academic institution will take 50% or more straight off the top, so you're looking closer to $150k to $250k a year.
$10 million is a lot of money. Science can be expensive but that $10 million can last a long time.
I agree it's real money, enough to do some real stuff.
But:
> An academic institution will take 50% or more straight off the top, so you're looking closer to $150k to $250k a year.
Well, that 'overhead' is paying for things like lab space, electricity, health insurance (at least for the faculty, good luck grad students), liability insurance, etc. Not to mention the actual salary of the PI and/or other researchers -- which universities in STEM schools increasingly expect to be covered by grants too. The academic institutions would argue all the 'overhead' they take is legit funding of things necessary for the research to take place, but even if you think they are spending it inefficiently or 'stealing' money for unrelated things...
I think as a rough napkin calculation, the $500K is probably still closer to that same $500K when YC Research spends it too, not twice as much.
Maybe, maybe not. I debated bringing it up, and honestly YCR is likely to have a large, up-front facilities cost (regulations, certifications, etc.). It would make sense to start easy - don't start off with infectious diseases of humans - and build from there. But still, I'd wager that YCR will run far more efficiently than your average university. Someone above brought up the UC system: management-level positions have grown over 200% the last few decades (http://universityprobe.org/2011/03/new-data-on-management-gr...). Even if we limit the scope to department administrators and staff, YCR won't need nearly the same sort of personnel infrastructure.
For starters, I'm prejudiced to think any smaller org will run more efficiently than a larger one (even though I realize this is literally the opposite of conventional/mainstream assumptions!).
But TWICE as efficient?
Also, I think all those administrator salaries are probably being funded by tuition (that is, by government subsidized student loans that the students will be paying back for most of the rest of their lives), rather than research grants. :)
Or actually you didn't even suggest twice as efficient, you suggested virtually ZERO costs other than direct research costs -- would we call that INFINITELY more efficient with a zero in the denominator? :) (Not even including salaries? Aren't those also funded in part by the 'overhead'?)
It was a point of comparison. The NIH pool is large but spread very thin and, we can both agree, less efficiently than YCR is likely to be. Twice? Infinite? No of course note, but it's a point of interest in the "$10 million isn't so much" discussion.
And no, researcher's salaries (postdocs, tech, grad students, etc.) are not included. The indirect costs go to the university to pay for department/college-level utilities, facilities, infrastructure, and administrator salaries (which, admittedly, the lab does benefit from).
Well, salaries obviously WOULD have to be included in the YC case, there's no other pool of money for them to come from! So that would actually be a significant cost that the NIH grants _don't_ have to include (if it's not included in the overhead), but the YC Research seed money WOULD have to.
But I understand that research faculty in science and biomed are increasingly expected to bring in grant funding as a condition of employment/tenure. It sounds to me like the institution expects to help pay their salaries with it.
And with institutions hiring faculty _to_ do research, not caring about teaching really, evaluated solely on research, they are basically there to do research... it actually doesn't sound that unreasonable to consider a portion of their salary overhead. The research couldn't get done if they weren't getting the salary, after all.
But anyway, yeah, like I said, I do agree with your original point that this is real money that is enough to fund real research.
Point of clarification: Researcher salaries aren't included in the indirect costs aka overhead, they come out of the lab's funding. So 50% indirect costs of a $300k project means $100k to overhead/indirect costs, and $200k to research; researcher salaries, grad student stipends, healthcare, etc. all come out of that $200k. Then what's left of that $200k funds the lab.
$10m is a lot of money for some fields, but in the scheme of high tech and high reward biological/chemical/physical research it really isn't. Personnel overhead alone for a reasonably sized group (10-20) can range from $2m on the low end, up to 5 or 6 million on the high end.
For example, at many institutions it takes ~$250k to support a single researcher (Salary, health/personal insurance, institutional insurance) .
As a researcher, I've run the numbers on what it would take to start a new nonprofit research institute. Assuming some other sources of revenue (e.g. grants), 10M was actually the sweet spot to get a minimum sustainable group off the ground (about 5 PIs, with postdocs and grad students). This is for math/cs/physics theory, where there are no experimental costs. The IP stance is surprising -- for an independent group, it is tough to give up on a potential source of revenue. Seems like YC itself provides a way to monetize the ideas, though, since some of them will likely feed into the VC side and help the companies.
@sama: how are you thinking about evaluating progress of the groups? it seems all the bureaucracy in the current research ecosystem derives from the fact that it is really hard to evaluate progress in research/science. So you form a committee of experts to evaluate if a given researcher is doing well, and that eventually leads to politics and bureucracy. what are you thoughts on evaluation?
TIL that Sam Altman has considerably more personal wealth than I would have guessed, if I'd ever been inclined to guess.
Anyway, this is very cool. Exactly the sort of thing that needs to happen to push the world forward in deeper ways. I'll be very curious to see what sort of teams they put together and just how fundamental and long term and heretical they go with their research.
If every poster here had their net worth attached to their username, you'd probably cry.
We all realize founders maintain "a lot more" equity than employees, but it's almost like our brains aren't wired to recognize, remember, or rationalize how some people get really really rewarded more than others. With Sam being grand poobah of YC, he's on a life track to being a liquid-capable multibillionaire within 5-10 years after the current unicornpalooza goes through a few rounds of lofty exits.
My propositions for SF to require all equity-exited techies to tattoo their net worth on their foreheads keeps getting shot down by anonymous "concerned citizens."
HAHA - I thought the same thing. I ACTUALLY scrolled back up to the top of the article to see if it was truly posted by him after reading that line....
Sweet! We'll need more things like this if we want to see our share of societal progress. I have faith that more programs like this will blossom in the coming years.
@sama -is this only for severely resource intense research projects? And do you envision this turning into an alternative to graduate school too?
I like this idea; it's pretty novel. Since there is just one group starting this off, which makes a ton of sense from a funding perspective, is this an MVP or a long term bet with regards to this specific group? Meaning if this group has trouble getting their idea off the ground in the next 1-2 months, do they get scrapped or are they going to have years to conduct their research?
An idea for future research may be to do almost like a round of start-ups where a bunch of groups come to SF for 3 months but unlike start-ups you find the promising ones and keep them all and drop any that do not look promising. Basically thinking of ways to not put all your eggs in one basic. Naturally this becomes quite problematic if you're researching something highly complex that won't show signs of promise for a decade!
The better yet native is to focus on quality of research outputs rather than just number. "Grant committees can't read, but they can count".
Trouble with that is it doesn't scale. It's easy to do when you're funding a handful of researchers like YC, but impossible when you're funding thousands like NSF or NIH.
388 comments
[ 3.1 ms ] story [ 326 ms ] threadThis also feels like cheating because we're targeting extremely successful research. If we knew ahead of time which research would be successful it would be a lot cheaper to do.
This is awesome to see, maybe YCR could be the next Bell Labs or something (although Google probably has a hold on that now).
(As often times the quality of research output is hard to gauge immediately without the greater context of time passing)
But of course this is just my useless speculation :)
[1] http://blog.samaltman.com/machine-intelligence-part-1
[2] https://intelligence.org/donate/
[3] http://futureoflife.org/AI/AI_beneficial
[4] http://hpmor.com/notes/119/
Two words: baseband processor.
The world really, really needs a fully open baseband processor.
My feeling is that the software to accomplish this is already largely there, but the hard part of getting a design to the fab and actually getting anyone to adopt it isn't.
I think there would also be a number of regulatory hurdles to cross, since this would end up being essentially an SDR, and the FCC (for example) already seems to fear wide public availability for those.
But yes, an open processor could help with a lot of open networking problems, even though it seems more like an engineering matter than research per se.
If someone wants to apply with an open baseband processor startup, applications are due Oct 13!
We have OsmocomBB, but it's limited to the 2G standard and comes with a ton of usage caveats pertaining to regulatory compliance and legality.
(Source: I work at MSR)
[1] http://spectrum.ieee.org/view-from-the-valley/at-work/innova...
[2] https://thmatters.wordpress.com/2014/10/14/letter-re-closing...
Eh. That a pretty huge jump. Look, most new "tech" companies research is mostly PR, internal and external. If you're doing something you have little experience in and therefor can't evaluate, which success you aren't dependent on or necessarily benefit from and everyone will say how cool it is regardless then it's extremely hard to not cut corners and achieve real success.
Still exciting, but I wouldn't underestimate the challenge.
[0] https://news.ycombinator.com/item?id=10322929
I think Google managed to recreate something similar with their 20% projects
Other companies have had the 15% / 20% culture, etc. long before Google. Example, http://archive.wired.com/techbiz/media/news/1998/01/9858
Sam, when I met you at Startup School last year, I wanted to be brief. The most important thing I wanted to communicate to you was my gratitude, and I thanked you for sharing your startup class. I'm blown away by and so grateful for the stuff you've been doing since. I'm very excited to see how these new projects work out. Thank you again man.
What's the process for an outsider to get involved? This sounds like a really cool and much needed movement. Will more information be coming soon on how to apply/join/contribute?
Are there plans to provide laboratory space for areas that require it? Would researchers be housed in some space in MV, or would they work out of their own space like YC startups?
Our government has historically underfunded research, and even more often one side funds it then the other gets into power and cuts it again. As a result, there are lots of ex-academics looking for more interesting ways of earning a living!
- Will the researchers need to be based in the Bay Area - AFAIK: not the best place to raise a family?
- In the short term, how many researchers do you plan to hire?
We're starting relatively small (on the order of 10) and grow from there at a reasonable rate.
BTW the "Read more" at https://ycr.org/ point to the Y Combinator blog not to the post permalink. It will became confusing once you write a new post.
Yep, I'm planning on swapping in the permalink in a day or two, but just want to avoid making changes during this launch.
1. Where will the research be based? Is YC providing space to setup lab?
2. One of the problems I see with this is lack of fellow researches present on site making casual collaboration harder (compared to a research university campus). Have you thought of that?
3. Are you planning on funding researches with track records or people with high risk/high reward ideas that are relatively new?
2. Yes, eventually we'd like to have multiple groups in the same space just for this reason--I always think the interdisciplinary stuff is underrated.
3. Some of both, probably, but mostly we want to support people that the existing system does not support (e.g. people that are not yet famous)
1. Try set up a lab in Cambridge at some point down the road.
2. Pay particular attention to CS theory: it's (arguably) high-impact and very cost effective --you can practically fund an entire generation of researchers in a subfield for a fraction of the cost of, say, a major interdisciplinary initiative. See for instance what the Simons foundation is doing at Berkeley[1].
Best of luck with what you're doing. You can end up actually changing the world in this way.
[1] https://simons.berkeley.edu
A lot of the stuff in TCS is theory for the sake of theory, and one would want to avoid too much of this.
I'm particularly interested in any arguments that complexity theory is high-impact (beyond the very useful insight that there are some problems for which no polynomial time algorithm is known). I have a pretty good idea of the impact of cryptography and randomized linear algebra (sometimes considered CS theory), but am also interested in hearing about other fields considered CS theory with useful applications.
[1] https://www.youtube.com/watch?v=a0SUpwIAjQs&t=1m35s
MA or UK, though? Both would be awesome, I think, but I don't believe YC has any offices outside the US at the moment...?
What happens when some of the research that is done paints some of the activities of YC or their companies in a less-than-positive light?
Peer review has long been the traditional validation mechanism in science. I don't see why it wouldn't be applicable here.
We want to get the best people in the world, and I think YC equity is a pretty compelling thing to get. We're competing against e.g. Google, and Google stock is also a pretty compelling thing to get!
I assume that by getting google stock, someone can sell them easily in the public market. But how does it work with YC equity?
What does this phrase mean in English?
Theoretical breakthroughs are often inspired by real problems. For instance, information theory arose from practical problems of wartime communications. YCR will be in close proximity (geographic, organization, and intellectual) to lots of startups with interesting and worthwhile problems to solve.
Honestly I'd like to see Kickstarter take a stab at this too, providing the general populace with an outlet to support research projects.
Not sure if this is driving your thinking but this would open a desperately needed alternative to Academia if you can scale this idea as successfully as you've scaled startups.
I've watched my genius brother-in-law (PhD Materials Science & Chemistry / Biomimicry) be consumed by a very broken UC system over the past decade.
Started with an exceptionally bright, curious, and inventive man who created breakthrough science in self-healing materials for the "benefit of the world"
10 years of grant fights, personnel struggles, under-served licensing resources, conflicting lab-vs-student priorities, etc has put him out the other end hopeless and disenfranchised.
The world will probably lose one of their best "public" researchers to the bowels of commercial science as a result.
Go for it Sam, good luck and excited to see what your first projects are.
The academic community is fraught with politics, mismanagement, and inefficiency. Researchers are driven by a need to get published and get grant funding, not to expand the frontiers of human knowledge.
Impact is currently measured by citation pagerank, and broader impact is undervalued. Outreach efforts (blogging, education) are consistently stigmatized [1]. This needs to change.
I would love to see more transparency and accountability in research. I've seen Nature papers continue to be cited despite completely wrong conclusions (the 'erratum' was published as a separate paper in PRL). The arXiv was a great first step at opening up access to papers, but the large publishers continue to hold cartel-levels of control over scientific knowledge.
I'm super excited about YC research. Would love to see a quantum computation division!
[1] http://blogs.discovermagazine.com/cosmicvariance/2011/03/30/...
Academia is but one fruitful track for a smart grad student.
[1] https://www.insidehighered.com/news/2015/09/09/essays-academ...
A lot of what is wrong with the academic system is this expectation, which is bewilderingly prevalent in both the student and faculty bodies despite abundant evidence of arithmetic skills. Faculty are often quite unaware of what happens outside their building, and students are, well, students and absorb a lot of the mindset of their role models. So there becomes a glut of post-post-docs applying for a scant few professorship, which makes getting hired and then getting tenure ultracompetitive and stressful--so many people want the jobs that the administration can ask just about anything out of them, including pretty unrealistic expectations, bad working conditions (80 hr/weeks for 75 k/yr in the sciences at many good schools) and so forth. The conditions of tenure-track professorships (when professors are still young enough to be most creative and energetic, and are still skeptical of their fields' dogma) are poorly suited to creativity and focus, so their grad students are tasked with the actual science while the professors become managers.
This isn't to say that academia is actually 'broken'. If my coffeemaker is broken, it doesn't make coffee. Academia is still making the things it is supposed to, but never in the amounts or quantities or prices that everybody wants (as it turns out everybody's wants are wildly divergent). So for this and other off-topic reasons, it's not exactly broken but definitely not optimal.
I'm of the opinion that a more optimal academy would do more to prepare grad students to find their own path, be it in academia, industry, or elsewhere. Sort of like what the US liberal arts education is supposed to be, but at a higher level and with the requisite specialization as is fitting a PhD. The critical, analytical, creative, scholarly, and technical skills that grad school can (and often does) bestow or hone are a great fit for a variety of pursuits, including entrepreneurial ones. But there is little guidance here, in part because the faculty themselves have little idea how to go about doing anything but what they do 80 hours a week.
I left the academic track after a post-doc to become an independent researcher (doing funded and unfunded research as well as R&D consulting) and it's been really surprising to see how many professors think it's an unworkable idea. Not an idea that many people try and flame out, but one so unthinkably scary that no one has seen anybody try. Which smelled like an opportunity to me, and still does...
Kudos on trying to make it as an independent researcher though. That's a really cool idea. How has it been applying for grants without a university affiliation? What's the breakdown of your time spent on funded research, unfunded research, and consulting?
I spend about half my time or a little less on consulting, but it's been very research-oriented consulting so far, with few restrictions on publishing results. There are and will be ups and downs (most of my consulting is in energy) but I bill a lot more than I pay myself so that smooths it out a bit.
So far most of the non-consulting research has been unpaid but I try as much as is practical to get funding for it at some point--make a minimum viable product to write a grant upon and if it doesn't work out then make it publishable and iterate or move on. I like strategy, it combines well with dreaming. I have some moonshot ideas too, but most things can be broken down into units digestible for peer /grant review.
I'm finding that there are a lot of theoretical problems that are also practical problems. I also feel like I am doing research all the time, but a lot of time is spent building tools and skills to do new things, and less of my time is spent actually learning new things about the earth than I would like. But I guess that's research...
[0] In my profile.
The basic problem is that we need to rank one researcher's merit against everyone else's, and there's no perfect way to do this because nobody can familiarise themselves with every individual's research. Rely on peers' opinions to measure merit and you just reward back-scratching. Rely on objective measures like citation counts and h-index and you incentivise people to focus on these instead of research quality.
The thing is that macro-scale works pretty well. As an overall system, academic science produces new knowledge at an astounding rate, and in terms of dollars in to research produced I'd say it works pretty efficiently. The measures we use are correlated with research quality even if they don't measure it perfectly.
It's only at the detailed scale where it looks badly flawed.
edit: and of course whenever I spend a long time writing a thoughtful comment I get the "you're submitting too fast, please slow down" message. Time to leave this comment to sit for an hour before pressing reply again...
In the meantime, Google seems to have developed a pretty good ranking system that is difficult to game.
In practice I understand this happens via citation-swapping (similar to link swapping, or paid links in blogs) and poor-quality journals accepting articles for a fee, or without review (much like paid content, or content farms) and papers that cite other articles just for the increase in rankings (similar to link farms) so the analogy to SEO (in its various hats) seems to fit.
This would suggest that the only way to fix the 'broken' citation ranking system is the same as Google, and bolt on empirically determined heuristics above the pure mathematical model of PageRank, and black-list or penalise certain journals or types of citation?
The impact of Google doing this is, in financial terms, probably much larger than if the same type of approach would be applied in the academic situation. So if Google is allowed to do this, why can't we allow the same approach for research? Apparently, the system works quite well.
Compare this to the situation in rating of academic research, where the papers are rated by the community of authors itself and the ranking is essentially a weighted counting of references. Of course it leads to lots of self-serving spam.
If the idea of independent raters (and possibly machine learning on top of it) could be implemented for academic papers as well, that would be a big improvement, I think. Academia has got a lot to learn from search engines.
I wouldn't call prioritizing quantity over quality gaming the ranking system, it is merely playing by its rules. I would call gaming it if there was some cheat (like you cite me and I cite you even if really we shouldn't), which may happen but I believe is rare.
Maybe. But I can't really imagine how you'd measure that, or what you'd compare it to.
Even if the current system is pretty great, it can only get better with competition from alternative systems, like what's being announced here.
I wouldn't call throwing out your trained experts and starting with fresh trainees in a five-to-seven year cycle all that efficient.
Academia doesn't pay very much, makes you work very long hours, and requires extensive training and skills, justifies it all by reference to Mission, and then inevitably fires you and tells you getting fired was your fault for not being in the top 8% of the entire profession.
Don't get me wrong: our lab makes awe-inspiring machines that can see individual atoms held in vacuum by lattices of laser light. Our research leads to incredible discoveries, but we're held back by inefficiencies. It's a rusty system, but to any mechanic, rusty = broken.
Also, leaving to build a startup is rather uncommon at MIT physics, and most students (including me) absolutely love research. The academic career is just not attractive enough to spend another 4-12 years in a lab becoming increasingly good at doing exactly one thing.
Peter Thiel's lecture in sama's Startup Class is relevant: http://bit.ly/1ZdOwHQ
Graduate students tend to love to complain as much as possible about things that aren't that bad (I know I was one) and people leaving something idealistic often have a compelling need to try and put it down in any way they can to justify their decision to give up on the idealistic goals and try to make some money -- which is why you see so many "goodbye academia, you are broken" posts and blog entries. And some people have genuinely bad experiences like at any job -- bad manager, particularly low funding, bad co-workers, you're going to have a bad time.
That being said YC research sounds amazing. Basic research was funded by charity from the superwealthy (or only carried out by the independently wealthy) for most of human history, the model works. Researchers tend not to look into areas or find results which are bad for their backers. It's good to have some public and some private sector; we need to balance out the conflict of interest a bit.
My complaint is with those inefficient bureaucratic systems that you mention in passing. Life's much easier without them. :)
Be nice if we could fix that. School is great, industry is great, somehow connect them.
YCR is a great idea because there are issues in academia that need resolving. It's certainly not completely broken.
https://www.facebook.com/jfgariepyneuro/posts/46644277686075....
https://news.ycombinator.com/item?id=10177775
It really is nice to see an organization like YC take an interest in research. I applaud Sam for making that considerable donation and hope the outcome is better than expected.
e: Here's some hard facts - http://universityprobe.org
It conforms with my experience.
That said, the skills required to write great grants and schmooze with grant-givers are quite different from that needed to do quality research in science or engineering. So the academic researcher community in science and technology is whittled down to the small set of {those good at research}∩ {those good at writing grants} + ⊂{those good at writing grants} + geniuses
Holy cow. That's real money. Sincerely very impressed by the commitment.
>Széchenyi gained a wider reputation in 1825, by supporting the proposal of the prime minister, Nagy, to establish the Hungarian Academy of Sciences; Széchenyi donated the full annual income of his estates that year, 60,000 florins, towards it.
https://en.wikipedia.org/wiki/Istv%C3%A1n_Sz%C3%A9chenyi
Florins appear to be pretty steady with the dollar...
$10m is pretty much just enough to fund one reasonably sized group, doing advanced research, for about a year to a year and half. Seems to me that Sam is hedging so YCR can do its first round in about a year and get a bigger ask by pointing to the starting groups achievements.
http://www.vpri.org/pdf/tr2011004_steps11.pdf
Holy crap, that was some kind of read! Thanks for telling me about this as I had never heard of it. My proposal, similar to what Racket and REBOL/RED attempt, was to combine simple DSL's on top of a LISP and low-level 3GL (eg high-level assembler) with optimization, safety, etc built into translation process. Seems like Kay already did most of that with way more clever abstractions and techniques. Gives me hope my smaller proposal might go somewhere esp as I wanted to build components & servers not desktops.
Thanks again for the link!
Note: I thought about doing something similar in first secure desktop I construct where I do my presentation on it, let people counter the methods probably won't work, and then tell them after that. Always priceless faces when that happens. Also helps tell the contrarian bullshitters apart from those who think hard on your claims. Saves time in future. ;)
Sure, but the NIH is funding near to 50,000 competitive grants a year. Some googling online gives median/average grant size at $300k/$350k. Even $25bil over 50,000 projects is $500k a project a year. An academic institution will take 50% or more straight off the top, so you're looking closer to $150k to $250k a year.
$10 million is a lot of money. Science can be expensive but that $10 million can last a long time.
But:
> An academic institution will take 50% or more straight off the top, so you're looking closer to $150k to $250k a year.
Well, that 'overhead' is paying for things like lab space, electricity, health insurance (at least for the faculty, good luck grad students), liability insurance, etc. Not to mention the actual salary of the PI and/or other researchers -- which universities in STEM schools increasingly expect to be covered by grants too. The academic institutions would argue all the 'overhead' they take is legit funding of things necessary for the research to take place, but even if you think they are spending it inefficiently or 'stealing' money for unrelated things...
I think as a rough napkin calculation, the $500K is probably still closer to that same $500K when YC Research spends it too, not twice as much.
For starters, I'm prejudiced to think any smaller org will run more efficiently than a larger one (even though I realize this is literally the opposite of conventional/mainstream assumptions!).
But TWICE as efficient?
Also, I think all those administrator salaries are probably being funded by tuition (that is, by government subsidized student loans that the students will be paying back for most of the rest of their lives), rather than research grants. :)
Or actually you didn't even suggest twice as efficient, you suggested virtually ZERO costs other than direct research costs -- would we call that INFINITELY more efficient with a zero in the denominator? :) (Not even including salaries? Aren't those also funded in part by the 'overhead'?)
And no, researcher's salaries (postdocs, tech, grad students, etc.) are not included. The indirect costs go to the university to pay for department/college-level utilities, facilities, infrastructure, and administrator salaries (which, admittedly, the lab does benefit from).
But I understand that research faculty in science and biomed are increasingly expected to bring in grant funding as a condition of employment/tenure. It sounds to me like the institution expects to help pay their salaries with it.
And with institutions hiring faculty _to_ do research, not caring about teaching really, evaluated solely on research, they are basically there to do research... it actually doesn't sound that unreasonable to consider a portion of their salary overhead. The research couldn't get done if they weren't getting the salary, after all.
But anyway, yeah, like I said, I do agree with your original point that this is real money that is enough to fund real research.
For example, at many institutions it takes ~$250k to support a single researcher (Salary, health/personal insurance, institutional insurance) .
Anyway, this is very cool. Exactly the sort of thing that needs to happen to push the world forward in deeper ways. I'll be very curious to see what sort of teams they put together and just how fundamental and long term and heretical they go with their research.
We all realize founders maintain "a lot more" equity than employees, but it's almost like our brains aren't wired to recognize, remember, or rationalize how some people get really really rewarded more than others. With Sam being grand poobah of YC, he's on a life track to being a liquid-capable multibillionaire within 5-10 years after the current unicornpalooza goes through a few rounds of lofty exits.
My propositions for SF to require all equity-exited techies to tattoo their net worth on their foreheads keeps getting shot down by anonymous "concerned citizens."
https://en.wikipedia.org/wiki/Sam_Altman#Pre-YC_angel_invest...
@sama -is this only for severely resource intense research projects? And do you envision this turning into an alternative to graduate school too?
Congrats Y team! I'm can't wait to see what you come up for next announcement!
An idea for future research may be to do almost like a round of start-ups where a bunch of groups come to SF for 3 months but unlike start-ups you find the promising ones and keep them all and drop any that do not look promising. Basically thinking of ways to not put all your eggs in one basic. Naturally this becomes quite problematic if you're researching something highly complex that won't show signs of promise for a decade!
Trouble with that is it doesn't scale. It's easy to do when you're funding a handful of researchers like YC, but impossible when you're funding thousands like NSF or NIH.