Please Fund More Science
blog.samaltman.com
blog.samaltman.com
I've also seen, repeatedly, that most biologists (both wet-lab and computational) will just waste more money rather than using it to scale up productivity.
The real question is ultimately "how do we scale scientific productivity in a way that benefits humanity directly and in the short term", and few people have any answer to that. It seems like the current state of the art, and hard to improve on, is to fund lots of people, give them time to be creative, and then send them to meetings to get tipsy with people who have money, while demonstrating their posters. I'm not kidding.
I think the rich folks are more interested in making Institutes- the Allen Institute and CZI being two examples. Institutes which are not affiliated with a university have a lot of advantages to the folks they hire- tend to be in popular locations, with higher salaries, more resources for crazy ideas, less overhead on grants (just a guess, I don't really know for sure), and NO TEACHING REQUIREMENTS (many top research profs hate teaching because it eats so much time).
We collaborated with them quite a bit by the way, and they faced almost exactly the same issues. Salaries aren't that good really (maybe 10-20% higher than in US national labs), they struggle to hire people who usually can make twice the money at Amazon or Microsoft. And you still need some ties to academia.
From my point of view, as a developer who didn't understand anything about the science my platform was used for:
- I think you probably want to have a few PhD students, postdocs and professors. They are the ones who can be creative given the right environment: no pressure to publish, access to sufficient resources (material and human) that enable their work.
- You need more money, but more importantly to spend it wisely. Pay people enough (in my case 20-30% more) and you'll cut the turnover by 2-3x.
I think the answer lies in taking a good look at where the money is going. Particularly, I think I think we need to give more to the very productive people and institutions. I know it's the million dollar question, by what is Salk doing that others are not? Or CSHL which has something like 8 nobel laureates there.
One last observation is that a career in science just doesn't make sense anymore. Why slave away making $40 or $50k as a postdoc or an instructor after years of studious study when you can pull 6 figures in tech right out of school? This right here is one of the things that tells the bright ones to stay away from academia.
Back in those days, you needed the term "business model" in your grant submission in order to attract funds. I walked away, thinking I had interests that could better generate a business.
One major change I would make from public money is publishing of results good or bad. There's this competition to get into top tier journals (which I don't think is bad or you can get rid of), but this is done at the cost of dumping good science just because it wasn't good enough for a top tier. Using impact as a metric is good but not great. The mark of a good scientist is their work and their passion, not where they publish. While there's a correlation, there's also Goodhart. If this is the only metric that we really concentrate on, then it will lead to slower science. We all know that we learn much more from failure than success. We all know that most of science is banging your head on the table for weeks out of frustration. But that's also not common knowledge.
So I'd say that those funding science need to be more risky and those in science need to be more open about their failures and be realistic in their discussion of science.
> The mark of a good scientist is their work and their passion, not where they publish.
You are overlooking an invisible but equally important mark of a good scientist: a scientist’s impression on peers aka reputation, which usually comes down to two types: prestigious publication (in a prestigious journal or with a prestigious co-author) or having a good number of citations. The first kind of reputation is bequeathed almost immediately, the second kind of reputation is deferred — it can only be gained after the passage of time (years).
If a scientist is known to have co-authored a paper that peers later on characterize as a seminal contribution to the field, peers are much more likely to pay attention to future papers from that scientist. The converse is true.
The number of published papers, whether in prestigious journals or otherwise, far exceeds the number of people with free time and relevant qualifications to assess them.
At the end of the day, human attention is a scarce resource, which is why it is the metric those aspiring to be regarded as “good scientists” in their field optimize for. Taken together, a scientist’s publication history is used as a proxy for authority in a field, which in turn is used to filter which papers to read or send to the trash can.
So it’s really a question of which metric you’re using to evaluate each system not some universal efficiency metric. Global impact over the next 20 years vs global impact over 200 years etc.
I don’t know the history of crispr, but I doubt it was just a single check, single project completed from start to end. I’m sure it took a lot of iteration and success of many small short-term goals
A COVID-19 vaccine is extremely valuable today, but it becomes useless even 5 years from now. On the other hand our experience with COVID-19 is likely to spawn a great deal of research into developing vaccines faster which while useless for this epidemic could be extremely valuable for the next several.
Clearly a COVID-19 vaccine could save millions of lives and is worth investing in, but it’s a specific finite number. So called basic research can save more people’s lives than are alive today.
You're describing a 40+ year body of research as if it were a project to develop CRISPR as an endpoint. But that's nothing like the process that actually produced CRISPR.
Take the Hsu et. al. paper describing CRISPR-Cas9 (https://www.ncbi.nlm.nih.gov/pmc/articles/PMC4343198/), for example. That paper cites 100 other papers, with multiple citations in each decade going back to the 1980s. (The first paper cited was published based on research that likely happened 10 years before before the sequencing of the entire genome of an organism.) And those 100 papers lean on other papers, etc.
Science is not an engineering project.
Also, CRISPR‘s roots go way past 40 years. Watson and Crick for example helped pave the way well over 60 years ago. A decade after their research was published a doctor may have found it interesting, but it was still a long way from being directly useful.
Can you start with any evidence?
Correct me if I'm wrong: Gates projects are less productive than NIH, counterproductivey harming education policy, and focused on different, cheaper problems like non-US basic health and malaria.
While I agree to an extent, it's not necessarily possible to 'scale up'. I helped run a project off around $40,000 of grant money. Our goals were research and education. About half went to pay my coworker and I about $15/h, another large chunk went to gas. The rest went to equipment and stuff. We couldn't charge for anything, we had no product to sell. We made brochures and buttons we gave out by donation, our work was in no way marketable. We had to justify everything we did by proving the value of the animals we were studying to agriculture.
There's a ton of scientific research that isn't directly valuable in the short term. Especially when it comes to biology. As it is, much of the work is just gathering data because we have painfully limited knowledge of our ecosystems. But the more we learn, the more we understand how important they are for the long term functioning of the planet.
It's exceedingly difficult convince a logging company or an oil company, the government they bribe, or the public in general that caring about the long term gains of researching these things should come before the short term gains of exploiting them. At least in my experience.
There's also some insight here: https://www.quora.com/Why-was-Bell-Labs-so-successful
This is actually pretty straight forward and if you read my comment history, the clues are in there. If you have $500M+ you'd like to spend on curing cancer, feel free to reply with a means of contacting you and be prepared to provide proof you're good for the funding.
Double over 5 years is just shy of 20%, risks have grown by much more than that amount. Current budget is a paltry ~30B. Progress is better than perfection.
To his point, "Waste is more than just a waste of money and resources. It can actually be harmful to people's health."
> He backed his staggering statistic with these additional stats: 50 out of every 100 medical studies fail to produce published findings, and half of those that do publish have serious design flaws. And those that aren’t flawed and manage to publish are often needlessly redundant.
What we need is NOT more funding, but we desperately need to improve the research and funding processes to make them more relevant, more efficient, and more reliable.
1. Publicly funded studies should yield open source research data that is freely available, so that studies can be repeated and experimental methodologies be improved and scrutinized.
2. We need to prioritize randomized clinical intervention trials over weak and questionable epidemiological surveys that often only muddy the waters and hinder our ability to draw sound conclusions.
3. Consequentialism should drive research funding. We need better and more formalized ways to identify gaps in our current knowledge, and to identify the potential impact of research before funding it. We don't need to allocate our current proportions of funding into research on subjects that are already very well understood or unlikely to drive policy and decision making. For example, more studies showing that exercise is good for you aren't likely to have a large impact moving forward.
Here's a more in-depth link to Dr. Bracken's speech:
https://justthenews.com/politics-policy/coronavirus/while-ni...
Sometimes scientific progress goes "boink". Consequentialism is dangerous. Researchers, like everyone, need to be able to fail.
I mean, look at the proportion of software projects that fail, which I'd estimate to be 50 % at least. And software engineering operates with much fewer unknowns compared to research.
Physics research is similar: Much of the research does not yield world-changing technology, or anything useful at all. I wouldn't say it's useless though, as unsuccessful projects still can provide inspiration for new research avenues and even if the research fails, the researcher (hopefully) gets better at doing research in the process, so the chances of producing something good the next time he/she tries increase.
I think the most promising avenue of increasing research productivity is to make it possible for more people to do quality research. Talent is everywhere but opportunity is not, so let's create more opportunity.
The Manhattan project and others like it succeeded because they gave resources to scientists and researchers who were not adapted to gaming a government-funded and bureaucratically-run grant system. Major government funded research projects (especially international ones like nuclear fusion) are not stalling because of lack of money, it is simply the system and incentive structure. Pouring more money in will not help us.
A prize system like what Tyler Cowen is advocating may work better, though I can't say I'm sure about that.
Sure, we spend a lot of time figuring out how to advertise and promote our work to get the most return from it. However, when it comes to running experiments and figuring out what projects to support, we only care about what would be important to our field. In fact my advisor funded me to spend a few months on a project he didn't even have a grant for. Even though he couldn't get money for it, he thought it was an important piece of work and wanted to see the results.
On the other hand, the lack of funding for public research is very real. Most of the equipment I use hasn't been updated since the 80s. My lab has had to grab used and broken equipment from around the whole state just to get our work done. This is the primary reason why I'm not interested in staying in academia. You have to beg for your funding and what you get isn't ever really enough to support the people on the ground. My lab computer for instance is a thin client that I'm forced to run CAD software on. It takes a solid minute to open some menus in it. I had to spend a month of my time rebuilding some ultra high vacuum pumps because we couldn't afford to pay a professional to do it and because I get paid pennies as a grad. student. I stick around, however, because I care about what I study, not because of citations.
> The way the incentives play out is that they create a survivorship bias. You can’t do science today without at least basically paying your own salary, either through grants (many biomed researchers have to pay a large component of their salary directly through their grants) or through making “having you in the university” make sense for the university so that it pays your salary. That again basically means either grants, or adding substantially to the university prestige and the ability to charge money for tuition and get high quality students etc.
> So, anyone with tenure in academia today has spent say 5 to 10 years minimum creating a “brand” that somehow enables them to pay their salaries, and/or funds for lab work etc. If the primary kind of “brand” that universities respond to is a “fake sciencey hypey overconfident, bullshit brand” then the primary content of academia today is people who survived that cutoff…
> It’s not so much that individuals see the incentives and change their own behaviors to fit it (though this happens to some extent) it’s more just that lots of people go into the filter, and the ones that come out are enriched for that particular behavior.
> Also note that this mechanism doesn’t require the individual scientists to be primarily money-driven. It’s sufficient for the MBAs running the dean/provost/president offices to be primarily money driven, which they definitively are.
I would recommend reading chapter 4 of the book "Disciplined Minds" by Jeff Schmidt (a physics PhD). While I disagree with a large fraction of the book (in particular, I don't think the problem is inherently political), I think chapter 4 is mostly right and shows how a researcher can feel self-motivated while actually being under strong external influences.
http://disciplinedminds.tripod.com/
I can give examples from my own experience in academia if anyone is interested.
I might just be really lucky to be in a clean department, but Machiavellian researchers accumulating in academia by filtering or otherwise just isn't my experience, sorry.
I'm interested in hearing your experience though because it sounds like it's very different. I'm studying accelerator physics by the way, an extremely applied sub-field of physics. It may be that it's easier to weed out the people that don't pull their weight here.
Academics are incentivized to publish papers in recognized journals and bring in external funding. I'll focus specifically on publishing here for simplicity. I could write about funding if desired.
The source of many problems I have with the current system is that I want to do high quality research, but the current system doesn't make that easy. The most important research I've done so far has been to make a large data compilation and use this to test different models and hypotheses in my subfield. But my advisor never liked this, and mostly discouraged me from doing this research. He would have preferred instead that I stick to the more established path of skimming the literature for a problem some review article called open, doing a naive experiment about that, and then publishing the experiment. (Yes, I am being facetious, but unfortunately this characterization is accurate in my experience.) My experience is that review articles are far from comprehensive and frequently contain misinformation. It's important to actually assemble a large amount of data to see where the real gaps are, not just what a review article called a gap. This sort of data compilation is very rare in my experience.
It really was clearly about getting publications as quickly as possible. My advisor repeatedly pointed out to me what others recently published in my field and exclusively what he pointed out to me I'd call naive. His point was "This is good enough to get a publication" and my point is "So what? Just because other people do a bad job doesn't mean that I should." To him, other people are getting publications when he's getting nothing! He didn't seem concerned that I thought most of these papers weren't making a contribution. That's not to say that he doesn't care at all about quality, but in practice he's more concerned about getting publications than quality. (He might not see the two as in conflict.)
In the end, I think I am making several large contributions, even if the "gestation" time was longer than people would like. To get tenure, you need to publish often and regularly, but a lot of valuable research takes longer than people would like to get results. The data compilation took a long time. I started this data compilation in roughly Jan. 2017. I had a conference presentation on it in Nov. 2017. I published a conference paper based in part on it in May 2018, and it wasn't until December 2019 that I submitted two journal papers using the data. (I prefer to take my time to perfect an article rather than rush it.) This sort of delay isn't acceptable in the current system, regardless of the value of the work at the end. In terms of total publications, I think I'm actually coming out ahead of most PhD students in my department, but by now I'm seen as unproductive by my advisor so that doesn't matter.
At this point, I'm waiting for one of my papers to be accepted because my advisor won't allow me to graduate until I have a paper accepted at a peer-reviewed journal. My original plan was to publish a few years after my PhD. I've been told multiple times by professors that the basic purpose of a PhD is to learn how to publish in peer reviewed journals, and they'll let you graduate once you've demonstrated that you can do that. And this does actually seem to be the crux in my case. Doesn't seem to matter that nowhere in my department's rules does it say that peer-reviewed publication is required.
It would have been much easier to just pump out a naive paper every year. I would have been done my PhD years ago if that were the case, and science wouldn't have advanced much if at all.
Also, don't mistake my advisor's position as "the incentives say to publish papers so that's what we should do". No, his position is something along the lines of "publishing papers is part of the point of a PhD". He's not thinking about incentives. He genuinely believes that publishing tons of papers is a good thing. The incentives selected him; he is not responding to the incentives as far as I can tell.
I think you should consider listening more to your supervisor. I know nothing about your field, but I've seen a lot of PhDs go off the rails because they chased after something they thought was a great idea only to find it was a problem no one really cared about.
Young PhD students are really smart, but I've certainly seen that experience does seem to count in choosing good questions.
The incremental "find an open problem mentioned elsewhere and work on that for a while" isn't a bad way to start a PhD career. You learn a bunch about the field doing that.
At this point, I'm waiting for one of my papers to be accepted because my advisor won't allow me to graduate until I have a paper accepted at a peer-reviewed journal. My original plan was to publish a few years after my PhD. I've been told multiple times by professors that the basic purpose of a PhD is to learn how to publish in peer reviewed journals, and they'll let you graduate once you've demonstrated that you can do that. And this does actually seem to be the crux in my case. Doesn't seem to matter that nowhere in my department's rules does it say that peer-reviewed publication is required.
This seems an odd thing to complain about. Surely a PhD is all about peer review, and should be so. Peer review means people who know about the topic have reviewed it - how else should a PhD be done?
Unfortunately only so much can be communicated in an online comment. I tried going into detail, but I'll leave it at this: I do listen to my advisor, but I disagree with him about how science is best done.
> The incremental "find an open problem mentioned elsewhere and work on that for a while" isn't a bad way to start a PhD career. You learn a bunch about the field doing that.
Reading your paragraph makes me think my point wasn't communicated clearly. I'm not arguing against working on open problems. I'm saying to make sure that the problem is actually open, and don't start doing expensive things like experiments without having a good idea of what you're doing. Given the existence of a large amount of archival literature, your great idea is likely to have been done already. There is value in reproducing previous experiments, but people should do new research too.
> This seems an odd thing to complain about. Surely a PhD is all about peer review, and should be so. Peer review means people who know about the topic have reviewed it - how else should a PhD be done?
For what it's worth I don't believe peer review as currently practiced in journals provides much value. The reviews tend to be superficial at best, and are provided far too late in the research process to have a strong influence on it. In my experience the reviews mostly help to make the paper more clear, and not because the reviewers give good suggestions on writing themselves, rather, because the reviewers seem confused about things I thought were written clearly.
However, my advisor is definitely a strong believer in peer review as currently practiced. I mis-characterized him before. For him it's less "publishing in journals is the right thing to do" and more "publishing in journals provides a strong check on the work".
In my view, the most important thing someone can do to improve their research is to be deeply critical of it. Unfortunately few people are critical of their own work. Getting feedback from others is valuable, but I believe that the best way to do so at present is to actively solicit it yourself, which is what I've been doing. Peer review in a journal should be the last step, not the only step aside from one's supervisor looking over the work.
Also, I removed a paragraph about how I strongly dislike the current academic publishers. That's a major factor here as well. My field does not have good open access options at present. I'd really prefer to wait a few years as I think the landscape could change dramatically.
This is where I am in my PhD. Things are culminating near the end. It can be psychologically difficult to go long stretches without publication, while your peers are firing off publications. I've seen this with other graduate students who take on large projects.
Many non-academics seem to have this view that science is full of senior researchers with political motivations stifling good ideas and gaming the grant system for personal gain. I can't say that people don't game the system, but in my experience it was very much a symptom of an already broken system.
As I see it, there certainly isn't enough funding (in the US) but that's not actually the primary problem. The issue is the processes used to allocate the funding (IMO). The funding cutoffs are _extremely_ competitive (because there isn't enough to go around) and renewals are typically every 4 to 5 years, but the system doesn't seem to be designed with funding stability in mind. It creates an incredibly stressful and uncertain climate, so of course people group up and game the system to varying degrees.
(I almost want to describe the climate as feast and famine, but that's not right because there doesn't seem to be much in the way of feasts - just slightly less famine.)
Academics and other government funded researchers are responding (personal communication) but they can't do so at the speed of companies who already have their ducks in line.
In addition, consider how hard it is for an academic to get regulatory approval. The biotech companies already have relationships with the regulators, while this is completely new ground for those academics, all while the regulators are completely swamped with requests from their preexisting network.
Most scientists aren't funded well, but I think a big problem is pouring money into the right holes. Some things will help a lot, some won't. I think getting more investors looking at the system can help with figuring out the right holes to put money into, and that will help just as much as the actual money.
Additionally, the same (or another) startup could make it easy for companies to pay to allow customers to be tested with a 15 minute test. I would not mind waiting outside a restauraunt or supermarket for 15 minutes while my test is run, if it means that I could enter knowing everyone else in the building is uninfected.
I imagine you would supply each business with one or more testing machines, test kits, and some number of technicians to operate the tests and test machines correctly.
The hard part is probably getting test machines/test kits while supplies are limited.
The problem is bulk ordering tests - you need a massive investment (about 750k for minimal order size of 100k tests), which he's been trying to find some philanthropic money for.
Those finger prick tests are convenient, but they only show results if your body is already producing antibodies. The really good test is an RT-PCR or RT-LAMP test, which uses a pretty invasive swab technique. It's pretty unpleasant.
Colormetric RT-LAMP is the right method for that startup (probably) because you can deploy it with pretty few resources and get rapid tests. However, it can be finicky, and isn't approved by the FDA AFAIK.
What I do not hear people talking about is this: Getting tested once is not enough. If you do not have COVID-19 today, it doesn't mean you are not going to have it tomorrow. The only test that can tell you that is an antibody test. Also, we don't even know yet (for sure, anyway) if people can be re-infected.
Then we can fund the science with it.
Honestly, I heard a comment on Slate Money to the effect of "Why aren't the billionaire class doing more to fund COVID research"
All the epidemiology, all the WHO plans for last decade, all the incredible co-ordinated response globally (#) has been publicly funded. Give over on the idea private companies are good for anything but taking existing research in the hands of hundreds and putting it into hands of millions and billions.
Pay tax. Vote for competence.
(#) yes I happen to think that shutting down most of the planet, in an attempt to save millions of lives, in a matter of weeks, is an amazing feat of global co-ordination. Yes there are lots of local differences and many many ways things could be better but humanity could have done so much worse.
So the next time your insurance pays $1000 for a drug, know that $200 of that goes back to R&D.
What you’re really saying here is it’s wrong that some entities choose to direct more money to research than society does collectively, and they should defer to society’s decision to spend on other things instead.
For starters they can ask companies they fund to start paying taxes in some other place than Jersey, Cayman Islands, Luxembourg or any other tax heaven.
Even more, they can start paying taxes themselves without passing their income through bogus "foundations", "funds", etc.
One off donation that many companies are doing now "for the sake of the virus" is a nice PR stunt but does not change anything in the long run.
This is why early research is mostly funded by governments. Sadly, a lot of conditions are not properly researched due to lack of funding. Some governments are better at distributing funds than others. The current US administration isn’t.
Although the sentiment resonates, the business logic does not.
Consider antibiotic drugs. Big Pharma and biotechs alike have been jettisoning programs. Why? It has nothing to do with science. These drugs simply aren't as profitable as others:
> Several major pharmaceutical companies recently shut down their antibiotic and antiviral research projects, backing away from the growing threat of superbugs, which may kill more than 10 million people a year by 2050.
https://www.beckershospitalreview.com/pharmacy/big-pharma-ba...
How much do you think a cure for COVID-19 will make its creator? Be careful when you answer, because you'll need to include all of the R&D effort in the calculation, and also consider the opportunity cost of not pursuing something more profitable.
The next pandemic (or even this one) won't be cut short by science moonshots. It will take a fundamental shift in how the highly-unprofitable activity of developing drugs taken for a very short period of time and only once gets funded.
>The next pandemic (or even this one) won't be cut short by science moonshots
Why not? We are basically in wartime conditions. Even if the government doesn't relax it's rules, there's a high likelihood that enough hospitals/doctors will to conduct high risk trials on critically ill patients. Now is the best time for moon shots.
Ok, so let's just socialize the pharma industry and treat it as a matter of national defense rather than consumer luxury. With respect to infectious and transmissible disease, the paradigm of treating health as a matter of personal comfort and subjective preference has very plainly failed.
I also would like to counter - science doesn't need any more money. Money distorts the progress of science (for example, look at the grants system - what a dumpster fire). What science needs is proper primary and secondary education and a healthy population. Start with fixing schools and healthcare, and you'll have so much science you wouldn't be able to stop it.
I'd much rather see the governments fund coronavirus-related research after this is over. The primary reason why SARS (v1) still doesn't have a vaccine and research was abandoned was due to the incentive structure (SARS is gone, and so is funding). I wish we knew more about coronaviruses now, it's been over 10 years since SARS ...
No.
It's too late to get us out of this. We can prevent the next case. But no, it's too late now.
What was needed with coronavirus was prevention. That's a matter of politics and economics. No science.
I work on that problem in close collaboration with a lot of community lab spaces, and it is surprising how many have the capacity and knowledge to implement rapid testing in their local communities and the inability to do so because of regulation. Biotech, if you run proper controls, isn't actually that hard. A lot of those spaces are now seriously struggling because of the shut down.
The main thing that is stupid about how the FDA is approving tests is that they approve protocols, even though they have lots of controls for input->output. Coming from a computer science side, I'd imagine they'd so something along the lines of unit tests, but they aren't - they're still regulating how the protocol is done itself.
More people want to build than destroy IMO. If we tested in the synbio dream, this outbreak would have never happened in the first place.
https://crosscut.com/2019/03/seattles-only-community-lab-put...
That article mentions the cost of equipment. In academia, we used to semi-seriously discuss building our own stuff. We generally didn't because even basic engineering, fabrication, and firmware require a lot of time and effort. (We were occasionally forced to when we needed something extremely specialized that simply wasn't available off the shelf.)
It seems like open source equipment designs don't exist largely because they don't usually make sense for academia or industry to fund. I wonder if that approach might be viable for this type of endeavor, since it would allow designs to be shared between groups and would lower the cash barrier to that of the raw materials? I'm not sure how you'd navigate the patent minefield, but if successful I think it would benefit society at large.
Lots of medical and academic labs liquidate, and you'd think that the market would compensate and labs would pick up used equipment, but that typically doesn't happen - probably for regulatory reasons and because (I've found) academic labs aren't typically in the position to take risks on equipment.
For example, an eBay seller ain't going to be able to test a PCR machine, so they sell it FOR-PARTS, which makes it so academic labs won't touch it. Then, you can swing in and buy perfectly fine $2,000 equipment for $150. Best deal I ever got was $350 for a fully functioning lab robot.
The SoundBio people are great, and there's lots of labs locally! They're really suffering right now though, because a revenue stream is often education.
Biology is what all of medical research eventually hinges on and any significant breakthrough in biology tends to make vast swaths of medical science irrelevant and in a sense a waste of money.
It is super easy to contribute to the search for treatments for particular diseases but how do you contribute to the fundamental research that ultimately fixes problems?
>any significant breakthrough in biology tends to make vast swaths of medical science irrelevant and in a sense a waste of money
But most research in pure biology is irrelevant. So you're effectively mistaking a type II error for a type I error in your argument.
https://www.cdc.gov/nchs/pressroom/sosmap/firearm_mortality/...
A couple of issues contribute which are the pricing isn't transparent like hospitals in the USA. Also there are only 2-3 major suppliers and for certain items it's essentially a monopoly.
I'm hoping for this to provoke a rational conversation and not a political flame-war – so, please respond fair-mindedly.
I've worked in international dev where normally government grants go to a handful of familiars who subcontract out. It's often a question of visibility for everyone else (smaller groups either win subcontracts or have a shot at getting their work accelerated directly).
I know science funding is different, but I wanted to map all the orgs & relationships for visibility. The best list I can find are previous grantees of https://wellcome.ac.uk/
I’m not saying that _has_ to be the case but "tax-avoidance / tax-optimisation / complex international structure / negotiation with government keen to offer low tax deal" is a common point made against large tech companies. I’m surprised why Sam doesn’t connect that point to his point, or confront that issue directly.
Going beyond that for a moment, some corporations voluntarily foregoing these loopholes will not even be enough for the US federal government to be able to afford their existing obligations, let alone increase spending on research or anything else. Furthermore, in a competitive environment, any company that does forego these loopholes is likely to find itself losing out to a competitor that uses these loopholes to undercut their prices.
That leaves closing the loopholes altogether, which is something the government would have to do. But that's not even enough to address the deficit--the government would have to either cut spending or raise taxes elsewhere.
And it's entirely possible that closing corporate tax loopholes isn't even a good idea regardless. Corporations are legal straw men that, in most cases, ultimately pass on their earnings to their shareholders. A corporate tax effectively taxes all shareholders the same--in other words, the retired grandmother pays the same amount as the billionaire investor. In other words, corporate taxes can actually be far less progressive than individual taxes.
[1] https://taxfoundation.org/us-tax-revenue-by-tax-type-2020/
Citation needed.
The atrophy of the thymus is a major reason why the adaptive immune system declines. The evidence from sex steroid ablation in prostate cancer patients strongly suggests it is possible to provoke the naive T cell component of the adaptive immune system into regenerating itself in a matter of months, provided that the thymus is restored to more youthful activity. Also evidence from the Intervene Immune trial with growth hormone (not advisable as a strategy, but congratulations to that team on getting interesting data) for the same proposition.
To be clear this is a years long process for Repair Biotechnologies that would see us into trials around 2022, but at some point fixing the aged immune system will and must become a practical concern. At that point, we'll all be a lot less concerned about pandemics of this nature, as the mortality and hospitalization rates will be much reduced.
There are other things that need to handled to restore all of the aged immune system: getting hematopoietic stem cells back into line and functioning properly, regenerative medicine for lymph nodes, some form of targeted destruction for malfunctioning immune cells. But each of these items will give incremental benefits on its own.
I know nobody wants to hear this nihilistic take but humans are the most destructive invasive species in existence.
I wish biomedical and health science research instead focused more on how we can lead rich lives with the little time we do have, knowing nature might take us back to dust for any reason, and leave the world a better place with more knowledge for the next generation. Instead we have old wealthy (mostly white) men driving the interests of science research for them to live longer and have more time to spend their money destroying the earth.
It is an ethical mess to know how to cure a disease or give one better health outcomes and intentionally not apply it but there has to be a balance. I worked in a hospice for a few years and I saw elderly folk often abandoned by their (alive and well) families and dieing alone. It sounds crazy to think but I feel like too much of anything, even longevity, is not good.
I currently teach at a community college and have taught at some major universities. What I consider passing today is far less than what I would have considered passing 25 years ago. My anecdotal experience is that too many people are going to college. Standards have decreased and we are passing people through the system who should not be graduating. Even so there is increasing pressure from administration and politicians to increase the passing rate.
At my college tuition is increasing because state funding is decreasing. We are advertising to a population who realistically aren't college material. But we need the enrollment and I need a job. So I pass people who shouldn't pass. I strongly disagree with the notion that we need more colleges.
EDIT: What I perceive as a degradation in undergraduate degree quality has led to the present state where a Master's degree today has the same intellectual signaling value that a Bachelor's degree had 40 years ago.
I believe, education needs to be distangled from jobs. People should not be educated for the sole purpose of slaving away 8 hours a day somewhere. Education is about learning, deeper reflections, aquiring skills you enjoy, developing reasoning, identifying pitfalls, dealing with abstractions, practising scientific methods, becoming desciplined. Only as side effect of all these, you might be able to also do well at some job. This means everyone is entitled to the highest level of education. In an ideal world, all humans should obtain a PhD in something they are passionate about, something they believe in and something they are interested in.
Now here's the problem. We have mutilated our education philosophy to the extent that it is no longer recognizable from what it used to be. When you step in modern day institutes, the education is all about getting good grades and passing exames. I have met inenumerable humans, starting from Kindergarten to PhDs where they are forced to memorize things, instead of really understand things in order to pass exams. People who don't have great memory powers are forced out of the system. Other day I was looking at 1st grade history lesson on Columbus. What was the test exam questions? Name the ships! Why does name of ships matter at all? Why not teach kids instead reason he needed 3 ships?
My core hypothesis is that all humans are curious, we all want to learn something, we all have some interests. We destroy most of these attributes as we grow up by making it a massive memorization contest with prizes given as pass and fail. Then we make it all about jobs so folks are forced to learn things they were never interested in the first place.
Proof > Opinion
If there were thousands of small research labs taking tiny amounts of money, it would be worth it purely for the massive amount of trial and error, but currently, the system is just collapsing.
I'm not being snarky, I mean it. Bucky Fuller calculated that we could solve most of the world's problems for about $25B (this was in the 70's, so adjust for inflation and whatnot).
He was thinking big: houses built in factories (like cars) and delivered to the site by helicopter. You could roll out a whole neighborhood in a week or two. They had waterless toilets that captures BMs in little plastic bags, hyper-efficient "fog" showers. ( Legend has it that the project was spiked because it would put too many electricians and carpenters and so on out of business. )
Or see his Old Man River's City project ( https://en.wikipedia.org/wiki/Old_Man_River%27s_City_project https://solutions.synearth.net/2002/11/24/ ) A city in a single building with ten million square feet of enclosed open space. Add some applied ecology ("Permaculture") and you've got an arcology ( https://en.wikipedia.org/wiki/Arcology ).
We have pretty much all the answers, the problem is to align our power with our values.
Eastern block panel building settlements were built from per-fabricated segments mass produced in a factory. Then assembled on site from these segments by cranes rolling on rails.
https://en.wikipedia.org/wiki/Panel%C3%A1k
You can actually tell quite easily in many cases where the crane rails were & that their most efficient location was the factor, not stuff like ergonomy or aesthetics.
The end result was indeed a lot of housing capacity built at record time. On the other hand the quality of the flats was often disputable and the new high density neighborhoods often lacked basic public amenities, at least initially.
In a more modern context there are also issues with lack of parking space (no one expected so many cars back then) and lack of thermal isolation (usually fixed by replacing the windows by modern ones & isolating the sides).
>hyper-efficient "fog" showers. ( Legend has it that the project was spiked because it would put too many electricians and carpenters and so on out of business. )
Modern Japanese showers are pretty much like that. Japanese toilets are also much much more water efficient than what you can see here in Europe, not to mention mostly including bidets and other civilized living amenities still mostly considered extraodrinary around here.
When the wall fell it became a free-for-all overnight, all that security went out the window and not much was done to replace it. A couple of people got obscenely rich and plenty of people dropped well below the poverty line. It is that which powers the rise of assholes like Orban: the people that yearn for the bad situation they were in before because at least they knew just how bad it was.
It isn't just parking spaces, the neighborhood was designed with walking in mind. Many streets had shops at the bottom floor, and I could do all my errands on foot.
My elderly grandparents still live there. In a small radius around their building is a grocery store, medical clinic and a bus stop. They could live independently for a long time after they lost the ability to drive.
https://factoryos.com/press/housing-development-erected-in-w...
> The Dymaxion House was completed in 1930 after two years of development, and redesigned in 1945. Buckminster Fuller wanted to mass-produce a bathroom and a house.
Add his Ultra-High Voltage World Electric grid to your list.
I'm currently writing a series on designing my own home and contrasting it with some soul-crushing new construction around here. It's very obvious that the places cherished and loved in New Hampshire were the villages built slowly, and the places least loved are the ones flash formed overnight by contractors trying to max out $$. Building on site is not very expensive but I'm not sure that making it cheaper is fixing any of the major problems with housing. It would just give more life to planning for lifeless sprawl.
If anything, we need ways to build more beautiful houses better or cheaper, in smaller, odd and non-level lots. Not assembly line houses.
But, yes, most houses around me have so many flaws that it seems problematic to count them as assets. Christopher Alexander has a lengthy calculation on this in 'The Nature of Order.'
They don't have to be identical copies at all, in the computer age.
I live in a newish housing estate in a relatively affluent area of the UK, where every single house has the same cladding, the same colour of roof, the same colour of doors, the same colour and style of windows, the same small patch of grass out the front, and each house is separated by less than 2m.
Walking around the place feels... wrong, hollow, eerie somehow. And yet, many people seem to love this homogeneous kind of place.
Thing is, as you point out, it doesn't have to be this way - it's local planning regulations and NIMBY neighbours that are the problem. Why does every house need to look almost identical? Why do I need permission to paint my front door a different colour?
It's not prefabrication that prevents diversity - it's policy.
But it doesn't conform to regulations and/or everyone that has a veto in the planning and permit process.
Also, the vast majority of the cost of a half million dollar home is the land.
There is no technology that fixes people problems. I would say Skynet from Terminator was onto something but still failed :)
Compared to building the same stuff in a factory, it is.
I'm looking to design my own home and very much agree with what you have to say. I'd be interested in hearing more of your thoughts.
TL;DR: Buildings and neighborhoods should be designed in situ by the people living there.
The PL volumes are earlier and seminal (both SimCity and the concept of "Patterns" in programming are two examples) while the more mature Nature of Order is nearly mystic.
I'm convinced that if we could combine the ideas of Pattern Language with applied ecology we could create a new form of civilization so beautiful our ancestors would consider it Heaven on Earth.
The truth tends to be that:
* It wouldn't be so easy, even if we did the right thing.
* Doing the right thing is actually the hard part.
* What you think is the right thing isn't actually the right thing.
Most systems in place today are there for a reason. Organizations that do the wrong thing tend not to survive for a very long time.
Also most of the time right thing is not "right thing" for everyone because there will be always group of people who think totally opposite is the right thing to do.
Personally, I want to see more organizations devoted to solving problems and boldly becoming obsolete. The predominant reward models do not really seem to reward that.
(Recent semi-related anecdote: see recent nytimes article about what happened to the project devoted towards designing and stockpiling inexpensive ventilators that came into existence after doing a risk analysis around SARS.)
> The curious task of economics is to demonstrate to men how little they really know about what they imagine they can design.
(https://en.wikiquote.org/wiki/The_Fatal_Conceit#Ch._5:_The_F...)
We can invent those things, sure, but we can't necessarily anticipate all problems with them. For example, a city with 10 million sq. ft. (according to Google, about 0.36 sq. mile) of open space would be very cramped compared to New York, which has Central Park (1.3 sq miles). A virus like COVID-19 in a more dense location than NY could be far worse. Or to take another example, plastic bags are certainly one way to capture waste, but there was a movement to produce fewer plastic bags.
We really don't have all the answers - we have people who think they have all the answers, and we have some other people that understand that the questions are hard.
A composting toilet is probably overall a much better solution. (And a bidet.)
edit: I found this linked from the Dymaxion House article: https://en.wikipedia.org/wiki/Packaging_toilet
Some of the world's most severe and urgent problems are caused by (western?) society's emphasis on individualism -- or maybe that's a bit unfair, arguably it is fairly human for humans to be short-sighted and selfish much of the time. E.g. global warming is a huge problem, but global warming is just a symptom of the underlying problems of overpopulation and overconsumption. I don't really believe we can fix those two underlying problems by continuing to operate within the same individualist market-driven philosophy.
E.g. most people want a comfortable place to live, children, opportunities for their children, a way to make a living, cheap food, cheap luxuries, cheap travel, medical care when they are sick. Governments, researchers and business strive to provide these -- but arguably most of these activities are harmful where they encourage consumption or discourage population reduction.
With a market-driven individualist approach we (as the collective human endeavour) just keep doing what we are already doing a bit more efficiently or a bit faster, which doesn't address the problem of what we individually demand -- which the market & research moves to supply -- is bad for the collective in the long run.
I think there's a clearer version of an argument along these lines in one of Yuval Noah Harari's books (although he may not have the same outlook or conclusion that I do).
Individualism has been the crux of human progress - Individuals voluntarily and radically change their behavior when convinced. The job is to convince them.
I know some people argue this, but I think this argument misses the point, particularly when it is made rhetorically and not spelled out in detail. I don't think if GDP is coupled to resource usage or not is relevant when we're talking about physical / ecological limits. GDP doesn't measure anything physical, so it isn't a particularly helpful thing to add into the discussion.
What precisely could we mean by GDP decoupling from resource usage?
Suppose the statement "GDP has now decoupled from resource usage" is defined like so: Let g(t) denote GDP and r(t) denote resource usage at some time t. Then by "GDP has decoupled from resource usage" perhaps we mean that the fraction g(t) / r(t) will becomes arbitrarily large (monotonically in t, say) as t increases.
That sounds superficially positive, but then consider, from an environmental perspective, the GDP g(t) quantity is irrelevant -- it measures economic activity in units of human accounting conventions, it doesn't measure any physical quantity about our world --- we need to be able to enforce upper bounds on r(t), resource usage, for key resources, where we express r(t) in terms of physical and environmental limits.
To cherry-pick a working definition of resource consumption, r(t), for a key environmental bottleneck resource, we could approximate r(t) as "consumption of remaining free storage capacity for CO_2-e in the atmosphere to have a 2/3rds chance of limiting global warming to at most +3 degrees C, in gigatons of CO_2-e". That's measurable, has real physical meaning, is a critical factor in the future of our environment, our society & our physical economy, and is still being consumed with no real sign of us all slowing down.
If you look at the trend of CO_2 emissions attributed to the US and Europe, I acknowledge there is a slight downward trend over the past few decades:
https://en.wikipedia.org/wiki/List_of_countries_by_carbon_di...
If we want to pick at the details I might argue that the graph attributes CO_2 emissions to countries like China for polluting economic activity required to produce cheap goods that are consumed in wealthy countries that industrialised earlier, so arguably this way of attributing CO_2 emissions against countries doesn't even fairly account for which countries are capturing the value associated with producing the pollution. You might argue that the manufacturers of exported goods captured a lot of the value, and imported a lot of stuff from other countries to help bootstrap their industry, so my argument that western countries likely are more responsible for CO_2 pollution in recent years than this graph suggests isn't entirely one-sided.
But, leaving that all those blame games aside, the slight declines observed in some countries CO_2 emissions is nowhere near fast enough compared to the remaining capacity left in the our shared global greenhouse-gas-atmospheric-sink definition of remaining resources r(t) . If you squint and look at the trends over a 50 year span for a country like the US, it doesn't look like greenhouse gas emissions have becoming arbitrary small, at a very crude approximation it looks more like that they're constant! US emissions now are similar to what they were in 1970. They need to go negative!
And the alignment has to be deeper in our society. A good example would be all the money spent trying to find a cure for baldness. Imagine if the funds were directed elsewhere if there weren't so many people willing to spend money curing baldness.
But does changing our values mean the people who are caring about not going bald turn their resources elsewhere. Or should we look into why these people care so much. Is it that hair, and especially natural hair, plays such a large role in attractiveness that we need to change? Maybe what we need to change is not the desire of people to be more attractive, but what society deems as attractive.
This gets even harder and more outlandish when you consider people with what even more deviant values. How much would it cost to end slavery?
Does this seem valid? I paused for a moment but only recall the concept, lacking any example off hand.
Both of the currently approved drugs for baldness -- finasteride and minoxidil -- were originally designed for other purposes: BPH and ulcers respectively. The actual research effort spent to find a "cure for baldness" is not that large. Plus, androgenetic alopecia is a likely symptom of prediabetes[1,2]; to that end, "high consumption of fresh vegetables" has been found to reduce the risk by more than half[3] and so research on baldness is not necessarily about hair loss per se.
1: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC5073072/
2: https://www.ncbi.nlm.nih.gov/pmc/articles/PMC6340244/
3: https://link.springer.com/article/10.1007/s00403-017-1799-z
Its shameful our high level civilization is threatened by a primitive virus.
We will now spend 100 times more on survival than we would need to spend to prevent a lots of deseases and pandemics in the first place.
Whole science power and money of our civilization is wasted on smartphones and wars. And health scientists work almost for free.begging for years for our attention. And we are where we are.
Smartphones have given millions of people access to information they would never have. Smartphones are the sum of 30 years of computer R&D, making them cheap enough that Kenya has more then 80% of there population online [1]
The issues you listed would take significantly more time then “five years” to cure.
I think there is room to argue how defense money is distributed and used.
[0] First link when searching for DARPA inventions: https://www.itpro.co.uk/technology/34730/10-amazing-darpa-in...
[1] https://www.standardmedia.co.ke/business/article/2001273507/...
And also - its natural army attracts some bright minds because sometimes only army has resources and money to fund research. So as long as it is used for good cause it makes sense as a symbiosis
If the money spent on space and military was spent on genetic engineering we might not have smart phones, but we would have fire breathing dragons!
The article only mentiones in passing - ARPA, as it was called back then - that DARPA was once ARPA, and doesn't explore how DARPA's rate of invention tanked after the passage of the Mansfield Amendment.
This is a longer story that's relevant to the war is great for research! claim.
Also, throwing money at a problem doesn't automatically solve it. Progress often depends on few individuals with crucial ideas, but you'll have to get those interested in your problem first.
I'm a great C++ programmer (or so I like to think) because I have been paid to write C++. I could write any other language if I had been paid (and I have but the rest were unique to the client)
Cancer isn’t “one thing” that can have one cure, and it’s silly to think otherwise.
A doctor and an oncologist won’t tell you these things because they’re not even thinking about them. They likely don’t even know how to answer the question “can vitamins help some cancers?” Instead, mine repeated word for word responses about how we get most of our vitamins from a healthy diet. Then said he’d have to look it up.
Ask them about IGF1 levels, fasting, ketogenic diets they mostly won’t know. They simply say there is NOTHING you can do to prevent getting cancer.
It’s the system’s fault, the doctors and surgeons are just human beings who have no incentive to be practical.
The past decade in America has seen the dominance of the capitalist tech utopians like Paul Graham, and we are seeing the results. For god's sake Peter Thiel was working with YC barely two years ago and now Sam wants to complain that the federal government isn't prepared to respond to crises?? what a joke.
This is the case with a lot of organizations like charities and government contractors.
When you have so many layers between the money source and sink, each layer takes some % cut like a parasite and each layer fights for its survival like a parasite.
It's like how healthcare.gov could have been made by 4 people in a garage being paid $100k/year each, but instead it required billions of dollars and thousands of people.
Less is more in software. Some of the best software ever written was written by one person (or a very small team). The more people you add to a project the less cohesive the code base becomes and the fewer people there are that can see the big picture.
It's like the company Simplifile - small company in Utah. They have only a handful of developers that build actual website front/backend infrastructure, and then they have a team of specialists whose job it is to become expert in all of the county laws across the USA and create requirements out of them.
1. Existing biotech companies are probably perfectly capable of addressing this with usual means. Vaccines are a known quantity, not novel disruption.
2. By the time any startup gets off the ground (6 months at least), the crisis will likely have passed
Other countries have scaled out thermal imaging for fever detection, face tracking for contact tracing, and have mobile apps that can warn people where a previous infected patient have visited before. I have yet to see any of that here.
These calls remind me of 9/11 when we permanently reduced our civil liberties for a momentary problem.
But they definitely need as part and parcel of the implementation, a 'self destruct' that goes into effect once the crisis is over.
But take a look at the Patriot Act. Much of it expired, but has been reauthorized time and time again -- with support from both parties, both Obama and Trump.