An Apology

To Whom in May Concern,

Yesterday I made a mistake. I sent an email to the University of Washington run email list that reaches all faculty, students, and apparently some former students associated with the EE department. The email contained a link to a petition on starting a conversation about the control of automatic weapons, and also a suggestion that people sign the petition and forward the link. Some people welcomed the message, some viewed it as spam, and some people suggested that I broke the rules.

To clarify, my intention was to offer the people in my community another way to cope with the tragedy in Connecticut by suggesting that they take some sort of action. I felt that doing so was department business, because the stress of these events has affected us all so much - even in the workplace. Being somewhat politically naive, I did not view the petition is taking a strong stance. It asks for a conversation about possible legislation, but does not say in any detail what that legislation ought to be.

I realize now that some people perceived my action as taking a strong political stance and that I may have been using university resources and my position as a faculty member to further that stance. That was absolutely not my intention. I simply wanted to help relieve the pain that many of us are feeling. I will be much more careful in the future to identify which of my own thoughts are opinions and which are not, and I will endeavor to keep any of my ideas that could be perceived as political outside of university life and university resources.

Sincerely,

Eric Klavins
Associate Professor
Electrical Engineering
University of Washington
Seattle, WA

Off to Another Galaxy!

Finally! Quantum communication!

  http://www.nature.com/nature/journal/vaop/ncurrent/full/nature11472.html

Now when I upload my brain into a computer, make a copy, and send it via space probe to another galaxy, I can communicate with the copy of myself instantly! This paper solves so many problems.

In more detail, I should be able to do the following: When brain-machine interfaces are ready (and people are working on this, so I am optimistic), I upload my brain into a computer and make a copy of it. Send the copy in a robot, along with a bunch of entangled bits, to another solar system or galaxy. Once there, link the two copies of my brain back together with the entangled bits and have instantaneous access to what's happening in the other solar system.

Perhaps some alien civilization is already done this (and SETI should really be looking for quantum communicators, not listening to temporary technologies like radio) -- but the timescales are perhaps such that it would take 14B years before beings began to figure this out and another 14B years to implement it on a grand scale. Quantum communication is just too fragile to implement easily. In any case, eventually everyone in the universe would figure this out and you could do multi-hop communications with anyone anywhere. At that point, maybe you could communicate with someone outside of what we now consider the visible universe, since you would just need a bunch of short (million-year travel time to set up the entanglements) links in a row. Once the entire universe is entangled thusly, it is effectively smaller in terms of information propagation, shrinking all the time as new links are set up.

Then we start using the information gained instantaneously from all over the universe to steer the universe by changing the positions of various masses available in each location -- effectively closing the loop on the multi-body control problem. Using this new power, we arrange the universe into a more pleasing, safe, and reliable form and enjoy it in comfort until all the big energy runs out, and after that we live off black-hole radiation for the rest of time. Of course, at some point someone's computer might get a bug and the whole computation core-dumps, exploding the universe and we get a new big bang and start all over.

Ideas about ideas

My students and I recently had a conversation of lunch about ideas and how to find good ones. In academia we are, for better or worse, very concerned with the ownership of ideas, although in a way that is different from industry with patenting and copyrighting. Essentially, if you are an academic, you want your name to be associated with as many good ideas as possible. The way to get your name associated with an idea is to publish a paper on the idea before anyone else does. If you're lucky, others will even name the idea after you. For example: RSA encryption, which is a really great idea, is named after Ron Rivest, Adi Shamir, and Leonard Adleman; Turing machines are named after Alan Turing; in synthetic biology, we speak of Gibson cloning to refer to a method of cloning popularized by Daniel Gibson. Gibson himself didn't think of that name (I don't think), but it is just easier for everyone else to call it Gibson cloning, so we do.

One of the worst things that can happen to an academic is to get scooped: Someone else publishes on an idea that you are working on, but haven't yet been able to finish off. It is incredibly irritating to have spent so long working on an idea, caring for it, experimenting with it, understanding its context, teaching it to your students and colleagues, getting funding for it, writing about it, and dreaming about it only to see it next to someone else's name in the table of contents of some journal. Even worse, some distant collegue or student forwards it to you in email and says "hey, isn't this similar to what you are working on?".

Here's the problem for academics though: Ideas have lives of their own. (Lots of people call them memes -- although I'd rather not do that here. I think of a meme as more general. The context in which you call something "awesome" or the way you twirl your pencil are memes, but you are not going to publish on them). The point is that ideas come from somewhere: you read about them or they pop up in conversation and then they fester in your head, mutating and mixing with your other ideas. For me, new ideas often come from not correctly understanding other ideas. That is, I'll think I understand something and I'll really figure it out well, and then I'll realize that (a) I got it all wrong but (b) what I have is actually new and interesting anyway (Jung would call me "mutable").

Furthermore, ideas travel like wildfire. I tell my students: if you are talking with a bunch of academics at dinner and a great idea comes up: stay way. That idea is probably stuck in the heads of hundreds of people, and unless your lab is enormous and fantastically funded or you already have a paper on the subject 75% done, you run a high risk of being scooped. In fact, most of research is dominated by herd behavior with academics following trendy ideas to get funding. In fact, it often seems that the ideas control us, and not the other way around. This depressing behavior of ideas might suggest to you that you might want to keep your ideas to yourself -- or among trusted colleagues -- while they are young. But who wants to live not talking about the very ideas that get us up in the morning? Besides, being secretive is also not good for the ideas: you never know when you might talk to someone who knows something about that can help you make progress. Sometimes good ideas need all the help they can get.

In general, academics wanting to own ideas and ideas wanting to be free are really at odds with each other. Probably the best situation for an academic is to be working in a field that no one cares about, or to invent an entirely new field, and then to discover something really cool in that field that has huge impact. Of course, that later creates all sorts of stressful competition (RSA encryption, for example, rendered the formerly sleepy field of number theory somewhat exciting for a while, much to the chagrin of some mathematicians). It is really an uphill battle to work on obscure problems. Nobody knows if what you are doing has any potential use at all. Even you may not know. Funding proposals and papers will get rejected, etc. The work may never pay off: you just have to hope to get lucky and have a "black swan" event. If fame is what you want, skip academia. If care and feeding of interesting and obscure ideas is what you want, research is for you.

So what to do? Be an individual. Avoid herds. Think about your brain is the substrate on which ideas live. Help the ideas and seek out the strangest ones. Don't try to own ideas, try to shepherd them.

USAID

Here's a great idea: Start a new research funding program that focuses on solving the dire problems of hunger, lack of clean water, disease, poverty, etc. If researchers in academia and industry focused on these problems -- and actually had funding to work on them -- we might actually make some progress. This is exactly what USAID wants to do. See here. It is only $100M (compared to DARPA's $5B for example), but it is a wonderful beginning and could have huge impact!

Green explosives, again

Rob Carlson and Daniel Grushkin weigh in on synthetic biology for manufacturing explosives. I still can't believe that someone would think of making explosives by tweaking the metabolic pathways of bacteria or yeast. The idea satisfies a lot of constraints: Synthetic biology, check. Potentially useful for the military, check. Not a bioweapon, sort of check. Anyway, it will be interesting to see where this thread goes.

Diagnosis and Moderation

I have been thinking about my favorite potential application of synthetic biology lately: diagnosis. I am inspired by the immune system, which can diagnose almost any disease you have already had within hours and develop an entirely new diagnoser for a novel disease within days.

Diagnosis is incredibly important. And it must be done very inexpensively and quickly. For now PCR and ELISA type assays are just too expensive and so we are more or less in the dark about the dynamics of infectious diseases. Here is a challenge: build a diagnostic kit that is easy for anyone to use that costs 10 cents and that can tell you which variety of TB your patient has: TB, MDR TB, XDR TB, or the frightening TDR TB. At this price, you can screen everyone in a community and apply the correct treatment to the people who need it, slowing the spread and development of drug resistant strains. What's the cheapest way to deploy ultrasensitive detectors? Put all the computation into a self-replicating machine: a cell. I think that's the way to go.

In looking into who is on the front lines of research into diagnosis and who might fund my crazy ideas, I realized that military doctors and researchers are doing a fair amount of this work. The ARO and ONR fund work here in addition to the NIH and in particular NAID. Most of this work is not on synthetic biology, it is on ultra-sensitive or ultra cheap ELISAs for example. On the other hand, some of the work on recreating 1918 flu, which is arguably synthetic, was funded by the army and the NIH. In any case, I really can't complain about anyone wanting to fund bio-defense related research especially on diagnostics. Such work really could save the world. I'll still go to NAID and the Gates foundation first, however.

As Rob Carlson has pointed out to me in the many recent conversations he and I have been having about military funding of biotechnology, and as he described in his book, the issues are subtle. A peacenik like me is going to be suspicious about any military involvement in anything, but I certainly don't want to stand in the way of important work that must be done to protect ourselves from one of our greatest threats: emerging infections diseases.

Detecting and responding?

Just a bit of cognitive dissonance for you. The US signed the BWC which, among other things,

... bans the development, production, stockpiling, acquisition and retention of microbial or other biological agents or toxins, in types and in quantities that have no justification for prophylactic, protective or other peaceful purposes.

On the other hand, the Office of Navel Research in their Synthetic Biology Program that

Living organisms may be modified using synthetic biology to manufacture materials, create stealthy and distributed ‘sentinels’ that detect and respond to threats...

I guess responding to threats is "protective" - but that sure sounds close to the edge to me.