Biologic Institute Announces First Self-Replicating Motor Vehicle — April 1st, 2009 by Douglas Axe
Researchers at Biologic Institute have stunned the scientific community with the announcement today of a fully functioning automobile capable of replicating itself. Although simple autocatalytic versions of self-replication have previously been demonstrated, the complexity of the system described today—complete with GPS navigation, DVD player, and onboard WiFi—has taken everyone by surprise. In the minds of many, this discovery has forever altered the once fundamental distinction between life and non-life.
Reactions from the automotive industry have, understandably, been less philosophical. One executive, who wished to remain unnamed, characterized the development as “altogether unhelpful.” An assembly plant worker was less restrained: “This is unreal… I just hope they quarantine the [expletive] things before this gets out of hand.”
According to lead scientist Otto Cloner, “In the right kind of environment the process of self-replication just takes off. I still get goose bumps watching it.” The prototype self-replicator is a slightly modified version of the popular Jeep Wrangler—unmanned. When just one of these self-propelled prototypes is placed in an appropriate environment (one lacking any other self-propelled vehicles) magic happens. Or so it seems. Dr. Cloner himself takes the more modest view that “the replicative mechanism is really quite simple when properly understood”.
[1]
Describing the process as being “sensitive to environmental conditions”, Cloner explained that the most critical requirement seems to be a generous supply of the right kind of precursors. “We’ve done a lot of fiddling with the conditions, and one very consistent pattern is that the precursors need to be pretty much fully loaded Wranglers perched on narrow wooden blocks, with engines running in low gear. We looked at more stripped-down precursors, but the DVD and GPS were lost very early in the replication process under those conditions. [We are] still trying to pinpoint exactly what’s going on there.”
Forgive us. It’s the first of April.
Believe it or not there’s a serious point here having to do with scientific studies of self-replication and the origin of life. Consider the recent Science paper by Tracey Lincoln and Gerald Joyce. [2] The paper describes RNA chains about 70 nucleotides long that produce copies of themselves when placed in the right kind of mixture. The authors use the term “cross-replication” to describe this because they found that it works best with two distinct RNA chains, each of which catalyzes formation of the other one from supplied precursors. But since either of these RNAs could potentially kick the process off (by forming the other), much of the commentary on this widely publicized study refers to it as an example of self-replication.
The study itself is a helpful contribution to our understanding of catalytic RNA, but the hype accompanying its publicity is much less helpful. For example, under the heading “A never-ending dance of RNA”, Erika Check Hayden writes that:
Joyce’s group had already made [RNA] enzymes capable of catalyzing their own replication, but they could only reproduce themselves a limited number of times. The new enzymes can reproduce themselves indefinitely. “This is the first time outside of biology where you have immortalized molecular information.” [3]
Stirring language indeed, but is it justified? Technically speaking, of course, we could apply the language of immortality to our tongue-in-cheek Jeep example. The sounds of “replication”— the thump of bumper contact followed by the chirp of tread meeting pavement—could keep on going indefinitely. The only limitation is the supply of precursors. Right?
Well… yes but therein lies a formidable problem. To fully appreciate it, we need to recall the gold standard of self-replication—life. Oak trees make more oak trees out of air, sunlight, water, and minerals. No one knows exactly how they do it, but the amazing and undeniable fact is that they, like all life, assemble things of stunning complexity from things of sheer simplicity. The complexity of the finished products is itself remarkable, but when we consider replication specifically it is this contrast in complexity that is most striking. Life consistently delivers more than it demands—far more.
The RNA demonstration, like the Jeep one, falls well short of this. Both show how a spontaneous process can produce a finished product, but they only do so by relying on precursors that are every bit as unlikely as the products themselves. In other words, what is being presented as a step toward solving the origins problem is really just a displacement of that problem. The humble truth is that the catalytic RNAs simply join two pre-made halves together by making a single new chemical bond. [2] What’s more, the molecular structure for accomplishing this joining is built into the precursors in such a way that 1) wrong ends cannot be joined, and 2) the energy for the correct joining is pre-supplied.
How reasonable is it to call something so carefully set up “self-replication”?
It all boils down to the question of what is doing the work of making the replicas. If it isn’t really being done by the things being replicated, then it isn’t really self-replication. Living systems have fully earned that title by actually doing the work. But in the RNA demonstration, as in the Jeep one, the real work occurred behind the scenes. For the Jeeps it happened at a sophisticated assembly plant in Ohio. For the RNAs it happened on a sophisticated oligonucleotide synthesizer, supplied with the requisite reagents and programmed by people who went to great lengths to identify RNA base sequences that would work.
To get an idea of how little was actually being accomplished (comparatively speaking) by the RNAs themselves, we should see how the total number of chemical bonds in the complete RNAs compares to the number made (one) during “self-replication”. Ignoring hydrogen atoms, which don’t join atoms up into large molecules, each complete RNA molecule had over 1,600 specific chemical bonds. Except for the final one, all of these bonds were pre-made in the process of making the precursors.
So, advertising this as “self-replication” is a bit like advertising something as “free” when the actual deal is 1 free for every 1,600 purchased. It’s even worse, though, because you need lots of the pre-made precursors in cozy proximity to a finished RNA in order to kick the process off. That makes the real deal more like n free for every 1,600 n purchased, with the caveats that n must be a very large number and that full payment must be made in advance.
The Fallacy of Divide and Conquer sans Intelligence
Dasgupta, Papadimitriou, and Vazirani have described the ‘divide and conquer’ approach to computational problem solving, whereby large problems are broken into successively smaller sub-problems of the same type, until “the sub-problems are so small that they are solved outright.” [4] Scientists attempting to explain biological origins without recourse to intelligence always seem to be envisioning that kind of solution. By their way of thinking, Lincoln and Joyce have solved part of a big problem. It may be a small part, but if everyone pitches in and does likewise, soon we’ll have the whole solution. That’s the sentiment expressed in the subtitle of the never-ending dance article, which reads: The recreation of life’s origins comes a self-catalysing step closer. [3] And that’s what sparks the celebration surrounding the announcement of each little step. We’re celebrating the fact that today we know something to be possible which yesterday we only guessed was possible.
But here’s the rub. The work of ‘divide and conquer’ isn’t completed simply by solving a bunch of small problems, but rather by “appropriately combining their answers.” [4] And if each of the small problems is as hard as the one Joyce’s team has labored over, what does that say about the odds of combining their solutions?
There is a domain of endeavor where odds like these are beaten (daily) by converting possibilities into reliable certainties. It’s the domain where big problems are routinely broken into smaller ones, and the smaller ones are studied until they’re solved, and those small solutions are then pieced together into big solutions. It’s the domain that gave us things like automobiles and ‘divide and conquer’ algorithms.
Whether origin-of-life research will ever give us something of comparable value remains to be seen. Judging by what gets celebrated in that field, you have to wonder whether anyone honestly thinks it will.
[1] Jeep is a registered trademark of Chrysler LLC.
[2] http://www.ncbi.nlm.nih.gov/pubmed/19131595
[3] http://www.nature.com/news/2009/090108/
[4] http://www.cs.berkeley.edu/~vazirani/algorithms/chap2.pdf