On the referendum #26: How to change science funding post-Brexit [updated with comment by Alan Kay]

There was an excellent piece in the Telegraph yesterday by two young neuroscientists on how SW1 should be thinking about science post-Brexit. The byline says that James Phillips works at Janelia, a US lab that has explicitly tried to learn about how to fund science research from the famous successes of Bell Labs, the ARPA-PARC project that invented the internet and PC, and similar efforts. He must see every day how science funding can work so much better than is normal in Britain.

Today, the UK a) ties research up in appalling bureaucracy, such as requiring multi-stage procurement processes literally to change a lightbulb, and b) does not fund it enough. The bureaucracy around basic science is so crazy that a glitch in paper work means thousands of animals are secretly destroyed in ways the public would be appalled to learn if made public.

Few in SW1 take basic science research seriously. And in all the debates over Brexit, practically the entire focus is 1980s arguments over the mechanism for regulating product markets created by Delors to centralise power in Brussels — the Internal Market (aka Single Market). Thirty years after they committed to this mechanism and two years after the referendum that blew it up, most MPs still don’t understand what it is and how it works. Dismally, the last two years has been a sort of remedial education programme and there has been practically zero discussion about how Britain could help create the future

During the referendum, Vote Leave argued that the dreadful Cameron/Osborne immigration policy (including the net migration target) was damaging and said we should make Britain MORE welcoming to scientists. Obviously Remain-SW1 likes to pretend that the May/Hammond Remain team’s shambles is the only possible version of Brexit. Nothing could be further from the truth. If the government had funded the NHS, ditched the ‘tens of thousands’ absurdity, and, for example, given maths, physics and computer science PhDs ‘free movement’ then things would be very different now — and Corbyn would probably be a historical footnote.

Regardless of how you voted in the referendum, reasonable people outside the rancid environment of SW1 should pressure their MPs to take their responsibilities to science x100 more seriously than they do.

I strongly urge you to read it all, send it to your MP, and politely ask for action…

(Their phrase ‘creating the future’ invokes Alan Kay’s famous line — the best way to predict the future is to invent it.)


Science holds the key, by James & Matthew Phillips

The 2008 crisis should have led us to reshape how our economy works. But a decade on, what has really changed? The public knows that the same attitude that got us into the previous economic crisis will not bring us long-term prosperity, yet there is little vision from our leaders of what the future should look like. Our politicians are sleeping, yet have no dreams. To solve this, we must change emphasis from creating “growth” to creating the future: the former is an inevitable product of the latter.

Britain used to create the future, and we must return to this role by turning to scientists and engineers. Science defined the last century by creating new industries. It will define this century too: robotics, clean energy, artificial intelligence, cures for disease and other unexpected advances lie in wait. The country that gives birth to these industries will lead the world, and yet we seem incapable of action.

So how can we create new industries quickly? A clue lies in a small number of institutes that produced a strikingly large number of key advances. Bell Labs produced much of the technology underlying computing. The Palo Alto Research Centre did the same for the internet. There are simple rules of thumb about how great science arises, embodied in such institutes. They provided ambitious long-term funding to scientists, avoided unnecessary bureaucracy and chased high-risk, high-reward projects.

Today, scientists spend much of their time completing paperwork. A culture of endless accountability has arisen out of a fear of misspending a single pound. We’ve seen examples of routine purchases of LEDs that cost under £10 having to go through a nine-step bureaucratic review process.

Scientists on the cusp of great breakthroughs can be slowed by years mired in review boards and waiting on a decision from on high. Their discoveries are thus made, and capitalised on, elsewhere. We waste money, miss patents, lose cures and drive talented scientists away to high-paid jobs. You don’t cure cancer with paperwork. Rather than invigilate every single decision, we should do spot checks retrospectively, as is done with tax returns.

A similar risk aversion is present in the science funding process. Many scientists are forced to specify years in advance what they intend to do, and spend their time continually applying for very short, small grants. However, it is the unexpected, the failures and the accidental, which are the inevitable cost and source of fruit in the scientific pursuit. It takes time, it takes long-term thinking, it takes flexibility. Peter Higgs, Nobel laureate who predicted the Higgs Boson, says he wouldn’t stand a chance of being funded today for lack of a track record. This leads scientists collectively to pursue incremental, low-risk, low-payoff work.

The current funding system is also top-down, prescriptive and homogenous, administered centrally from London. It is slow to respond to change and cut off from the real world.

We should return to funding university departments more directly, allowing more rapid, situation-aware decision-making of the kind present in start-ups, and create a diversity of funding systems. This is how the best research facilities in history operated, yet we do not learn their key lesson: that science cannot be managed by central edict, but flourishes through independent inquiry.

While Britain built much of modern science, today it neglects it, lagging behind other comparable nations in funding, and instead prioritising a financial industry prone to blowing up. Consider that we spent more money bailing out the banks in a single year than we have on science in the entirety of history.

We scarcely pause to consider the difference in return on investment. Rather than prop up old industries, we should invest in world-leading research institutes with a specific emphasis on high-risk, high-payoff research.

Those who say this is not government’s role fail the test of history. Much great science has come from government investment in times of crisis. Without Nasa, there would be no SpaceX. These government investments were used to provide a long-term, transformative vision on a scale that cannot be achieved through private investment alone – especially where there is a high risk of failure but high reward in success. The payoff of previous investments was enormous, so why not replicate the defence funding agencies that led to them with peacetime civilian equivalents?

In order to be the nation where new discoveries are made, we must take decisive steps to make the UK a magnet for talented young scientists.

However, a recent report on ensuring a successful UK research endeavour scarcely mentioned young scientists at all. An increased focus on this goal, alongside simple steps like long-term funding and guaranteed work visas for their spouses, would go a long way. In short, we should be to scientific innovation what we are to finance: a highly connected nerve centre for the global economy.

The political candidate that can leverage a pro-science platform to combine economic stimulus with the reality of economic pragmatism will transform the UK. We should lead the future by creating it.

James Phillips is a PhD student in neuroscience at the HHMI Janelia Research Campus in the US and the University of Cambridge. 
Matthew Phillips is a PhD student in neuroscience at the Sainsbury Wellcome Centre, University College London


UPDATE

Alan Kay, the brilliant researcher I mentioned above, happened to read this blog and posted this comment which I will also paste below here…

[From Alan Kay]

Good advice! However, I’m afraid that currently in the US there is nothing like the fabled Bell Labs or ARPA-PARC funding, at least in computing where I’m most aware of what is and is not happening (I’m the “Alan Kay” of the famous quote).

It is possible that things were still better a few years ago in the US than in the UK (I live in London half the year and in Los Angeles the other half). But I have some reasons to doubt. Since the new “president”, the US does not even have a science advisor, nor is there any sign of desire for one.

A visit to the classic Bell Labs of its heyday would reveal many things. One of the simplest was a sign posted randomly around: “Either do something very useful, or very beautiful”. Funders today won’t fund the second at all, and are afraid to fund at the risk level needed for the first.

It is difficult to sum up ARPA-PARC, but one interesting perspective on this kind of funding was that it was both long range and stratospherically visionary, and part of the vision was that good results included “better problems” (i.e. “problem finding” was highly valued and funded well) and good results included “good people” (i.e. long range funding should also create the next generations of researchers). in fact, virtually all of the researchers at Xerox PARC had their degrees funded by ARPA, they were “research results” who were able to get better research results.

Since the “D” was put on ARPA in the early 70s, it was then not able to do what it did in the 60s. NSF in the US never did this kind of funding. I spent quite a lot of time on some of the NSF Advisory Boards and it was pretty much impossible to bridge the gap between what was actually needed and the difficulties the Foundation has with congressional oversight (and some of the stipulations of their mission).

Bob Noyce (one of the founders of Intel) used to say “Wealth is created by Scientists, Engineers and Artists, everyone else just moves it around”.

Einstein said “We cannot solve important problems of the world using the same level of thinking we used to create them”.

A nice phrase by Vi Hart is “We must insure human wisdom exceeds human power”.

To make it to the 22nd century at all, and especially in better shape than we are now, we need to heed all three of these sayings, and support them as the civilization we are sometimes trying to become. It’s the only context in which “The best way to predict the future is to invent it” makes any useful sense.

15 thoughts on “On the referendum #26: How to change science funding post-Brexit [updated with comment by Alan Kay]

  1. Good advice! However, I’m afraid that currently in the US there is nothing like the fabled Bell Labs or ARPA-PARC funding, at least in computing where I’m most aware of what is and is not happening (I’m the “Alan Kay” of the famous quote).

    It is possible that things were still better a few years ago in the US than in the UK (I live in London half the year and in Los Angeles the other half). But I have some reasons to doubt. Since the new “president”, the US does not even have a science advisor, nor is there any sign of desire for one.

    A visit to the classic Bell Labs of its heyday would reveal many things. One of the simplest was a sign posted randomly around: “Either do something very useful, or very beautiful”. Funders today won’t fund the second at all, and are afraid to fund at the risk level needed for the first.

    It is difficult to sum up ARPA-PARC, but one interesting perspective on this kind of funding was that it was both long range and stratospherically visionary, and part of the vision was that good results included “better problems” (i.e. “problem finding” was highly valued and funded well) and good results included “good people” (i.e. long range funding should also create the next generations of researchers). in fact, virtually all of the researchers at Xerox PARC had their degrees funded by ARPA, they were “research results” who were able to get better research results.

    Since the “D” was put on ARPA in the early 70s, it was then not able to do what it did in the 60s. NSF in the US never did this kind of funding. I spent quite a lot of time on some of the NSF Advisory Boards and it was pretty much impossible to bridge the gap between what was actually needed and the difficulties the Foundation has with congressional oversight (and some of the stipulations of their mission).

    Bob Noyce (one of the founders of Intel) used to say “Wealth is created by Scientists, Engineers and Artists, everyone else just moves it around”.

    Einstein said “We cannot solve important problems of the world using the same level of thinking we used to create them”.

    A nice phrase by Vi Hart is “We must insure human wisdom exceeds human power”.

    To make it to the 22nd century at all, and especially in better shape than we are now, we need to heed all three of these sayings, and support them as the civilization we are sometimes trying to become. It’s the only context in which “The best way to predict the future is to invent it” makes any useful sense.

    Liked by 2 people

    • Dear Alan
      Thanks very much for your comment.
      You may be interested that in my blog a year ago on the anniversary of the referendum, I wrote about ARPA and referred to your writings a lot.
      https://dominiccummings.com/2017/06/23/on-the-referendum-23-a-year-after-victory-a-change-of-perspective-is-worth-80-iq-points-how-to-capture-the-heavens/
      Best wishes — and I hope people with power start listening to what you’ve been saying for so long about research
      Dominic

      Liked by 1 person

      • Of the many things not understood about the ARPA-PARC process, the hardest to explain are the levels and kinds of cooperation combined with the willingness and ability to think about “systems” rather than “technologies”. It was the great ability of Licklder not to try to “be in control” (and realizing that as things scale you can’t be in control, so other ways have to be devised to make systems that behave closely enough to what is desired). We can look around at “people and cultures” today in the world who are trying to get back control, but they are missing the system dynamics, and really contributing to the deep problems in which we are embedded.

        Liked by 1 person

        • Agreed. In my blog a year ago I referred to your comments on ‘out of control’ processes.
          But I think the problem is even worse than you describe in this comment, though I have also watched enough of your talks to know you know this point extremely well and much better than me — it is not just the ‘peoples and cultures’ but the LEADERS, including leading civil servants, who do not understand systems thinking. (Of course there are odd exceptions but the number is tiny and far below the critical threshold necessary.)
          I have tried and tried and tried to get people in power to think about this with close to zero success.
          I wrote on the interesting way in which Mueller developed ‘systems management’ on Apollo and have written other things on systems thinking but I cannot get people to change how they operate. It is interesting that things like the ObamaCare rollout smashed up and people in Washington say ‘gee, this is a disaster, how do we deal with these systems problems’ and are almost completely unaware that people like you and others have been thinking/writing/doing on ‘systems’ for decades.
          A big part of the problem obviously is screwed incentives.
          Also ironically, I implemented some of the ideas about systems I’ve seen in talks by you and others in the referendum campaign. Unsurprisingly they had dramatically good effects and were part of the reason why about 10-15 people with almost no money managed to beat the Establishment despite it having practically every structural advantage.
          But of course everybody watching has focused on other things and almost nobody pays any attention to the *nature of the organisation and its processes*, and particularly systems thinking, that did it.
          This has strengthened my view that 1) if one could capture the apex of power in a state like the UK one could have dramatic positive effects by applying these ideas, but 2) the nature of bureaucracies is they will treat such attempts as mortal threats to be crushed mercilessly, and 3) the warped incentives make it very hard to persuade people in power to take the risk on such things. There is fruit simultaneously very low hanging and the hardest thing in the world to grab.
          So they keep doing things the same way and failing in all the same ways…
          Best
          d

          Liked by 1 person

          • Ps. The system dynamics of the EU was the main reason I thought it would be better for the UK to get out. The quantum computer pioneer David Deutsch put it very well when he said that the EU system is very poor at (actually in many ways hostile to) *error-correction*, which is so fundamental to a healthy economic/political system. Of course, reasonable people can disagree on whether leaving or trying to reform it was a better path…

            Pps. I am going to post a long piece on Licklider/PARC etc in the next few weeks, having read The Dream Machine and a few other things including more of your work/talks. I will keep trying to get people to fund research like this here.

            Liked by 1 person

          • A reply to your reply below. The first question I would ask about the referendum on the EU is “Did anyone on either side of the question think to create a simulation model of the important relationships and flows, and then run it to try to understand some of the ramifications?” This is quite possible to do, and computing is now up to the task of carrying out both the modeling and the simulation. This is now done all the time for systems like bridges, automobiles, planes, biological interactions, epidemics, the climate, etc. Basically, you don’t want to experiment on people if you can do most of the testing in a good simulator and the ramifications can be vital to the point of life and death.

            Liked by 1 person

  2. Also I just got an email from someone pointing out to me in email re your emphasis on treating grad students as scientists without PhDs to the effect — Kay is right, this is important, and it’s a big mistake Janelia has made even though Janelia tried explicitly to learn from PARC.

    Like

    • There is some truth to this. There is a very strong lab head/everyone else divide at Janelia, and its difficult for non lab heads to be independent. They just completely got rid of their independent researcher position. What this means is that talented young scientists wanting to pursue ideas outside of a single lab cannot do so. At universities, young scientists can get their own fellowships + funding. At Janelia, everything comes from all powerful lab heads, and it can create unfortunate power dynamics. As an example, if two people in their 20s have a great idea they want to pursue together, it is impossible in Janelia to do that without a lab head giving up a large fraction of their overall resources to essentially charitably allow them to do this, which in my view is a serious flaw in the model.

      However, it is a more general problem in academia these days: even the postdoctoral researcher position, ie, people typically in their 30s, is a ‘training position’, which is a ridiculous euphemism that lets them get away with paying scientists effectively minimum wage. This is the age that most great scientists had already done/were doing their best work (Einstein, Darwin, Newton, Feynman, Crick, Brenner, Watson, Franklin, Dirac, Marr, Hodgkin, Huxley……). Janelia has been very good at giving lab head positions to very young people, and it invariably has worked out. Other institutes stubbornly fail to learn from this.

      Perhaps a better criticism would be to say that Janelia places an unfortunately excessive divide on lab heads/everybody else, preventing the kind of bottom up, spontaneous collaboration that is the hallmark of much great science. By focussing so much on lab heads, they essentially narrow their bets. Further, as I said, concentration of power can lead to real issues….

      Like

  3. I’d be interested in some examples of where these algorithms and models can be shown to be really, really working. I’m not a cynic particularly but don’t have enough knowledge to see for myself. Though I’ve followed the climate change debate for a number of years and the best arguments the sceptics or lukewarmers have is that the predictive models have a dreadful record for accuracy – they run much too hot and climate alarmists have no answer, apparently.

    Anyway, this was on radio 4 recently on the topic of algorithms in recruitment, criminal justice and so on. Interesting.
    https://www.bbc.co.uk/programmes/b0b4zxcn

    Like

    • I just downloaded and skimmed this paper (I’ll try to be more substantive after I read it more thoroughly). The most glaring problem is that — though they mention that things changed dramatically in the midst of the Viet Nam war (the Mansfield Amendment put the “D” on) — they persist in trying to somehow treat similarly what were two or three separate very different “ARPA” processes. I think this is a real mistake given that the big successes were mostly done in ARPA without the D and with the initial ARPA processes. These processes are the ones that need to be most studied and understood.

      Like

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

w

Connecting to %s