Monday, May 23, 2011

Belated Prospect

I realise that I meant to put up earlier my May column from Prospect. Almost time for the June column now, but here goes.
________________________________________________________

The notion that God has an inordinate fondness for beetles, credited to the biologist J. B. S. Haldane, retains a whiff of solipsism. For beetles are not so unlike us: multicellular, big enough to see, and legged. But God surely favours single-celled organisms far more. Beetles and humans occupy two nearby tips on the tree of life, while single-celled life forms have two of the three fundamental branches all to themselves: bacteria and archaea, so alike that it was only in the 1970s that the latter were awarded their own branch. Archaea have a different biochemistry to bacteria – their metabolism usually produces methane – and they are found everywhere, including the human gut.

Our place on the ‘tree of life’ now looks like it may be even more insignificant, for a team at the University of California, working with genomics pioneer Craig Venter, claims to have found hints of a fourth major branch in the tree, again populated only by single-celled organisms. These branches, called domains, are the most basic divisions in the Linnaean system of biological classification. We share our domain, the eukaryotes (distinguished by the way their cells are structured), with plants, fungi and yet more monocellular species.

Like most things Venter is involved in, the work is controversial. But perhaps not half so controversial as Venter’s belief, expressed in a panel debate titled ‘What is life?’ in Arizona in February, that all life on Earth might not even have a common origin. “I think the tree of life is an artefact of some early scientific studies, which are not really holding up”, he said, to the alarm of fellow panellist Richard Dawkins. His suggestion that there may be merely a “bush of life” only made matters worse.

Drop in the ocean

Despite the glee of creationists, there was nothing in Venter’s speculative remark that need undermine the case for Darwinian evolution. The claim of a fourth domain is backed by a little more evidence, but remains highly tentative. The data were gathered on a now famous round-the-world cruise that Venter undertook between 2003 and 2007 on his yacht to gather genomic information about the host of unknown microorganisms in the oceans. The rapid gene-analysing techniques that he helped to develop allow the genes of different organisms to be rapidly compared in order to identify evolutionary relationships between them. By looking at the same group of genes in two different organisms, one can deduce where in the tree of life they shared a common ancestor.

Using Venter’s data, Jonathan Eisen in California discovered that two families of genes in these marine microbes each seem to show a branch that doesn’t fit on the conventional tree of life. It’s possible that these genes might have been acquired from some unknown forms of virus (viruses are excluded from the tree altogether). The more exciting alternative is that they flag up a new domain. If so, its inhabitants would seem so far to be quite rare – a minor anomaly, like the Basque language, that has persisted quietly for billions of years. But since we are ignorant about perhaps 99 per cent of species on the planet, who knows?

Thinking big

The European Union is looking for big ideas. Really big ones. Its Flagship programme offers to fund two scientific projects to the tune of €1 bn over the next ten years. These must be “ambitious large-scale, science-driven, visionary research initiatives that aim to achieve a scientific breakthrough, provid[ing] a strong and broad basis for future technological innovation and economic exploitation in a variety of areas, as well as novel benefits for society.” In other words, they’ve got to achieve a heck of a lot, and will have truckloads of money to do so.

Six of the applications – all of them highly collaborative, international and interdisciplinary – have now been selected for a year of pilot funding, starting in May. They range from the highly technical to the borders of science fiction.

One promises to develop graphene, the carbon material that won last year’s physics Nobel prize, into a practical fabric for information technologies. Another proposes to truly figure out how the brain works; a third will integrate information technology with medicine to realise the much-advertised ‘personalized medicine’. But these things will all be pursued regardless of the Flagship scheme. More extraordinary, and therefore both more enticing and more risky, are two proposals to develop intelligent, sensitive artificial agents – characterized here as Guardian Angels or Robot Companions – that will help us individually throughout our lives. The sixth proposal (which received the highest rating) is to develop massive computer-simulation systems to model the entire ‘living Earth’, offering a ‘crisis observatory’ that will forecast global problems ranging from wars to economic meltdowns to natural disasters – the latter now all too vivid. The two initiatives to receive full funding will be selected in mid-2012 for launch in 2013.

3 comments:

JimmyGiro said...

"The sixth proposal (which received the highest rating) is to develop massive computer-simulation systems to model the entire ‘living Earth’, offering a ‘crisis observatory’ that will forecast global problems ranging from wars to economic meltdowns to natural disasters"

I would like to hear Adam Curtis' view on such a project, as he's just finished his documentary "All Watched Over By Machines Of Loving Grace".

You've mentioned this project in an earlier post, and I recall my disdain of such a venture. I think it would be a great tool for governmental abuse; and considering that government would be a Euro-bureaucracy, I'm sure there will be much gnashing of teeth and throwing of bombs, by those who want a more independent life.

If it is as reliable as the BBC's weather forecasts, there will by riots. If it turns out to be kosher, there will be panic and riots, as people will anticipate disasters that have not yet happened, and react disproportionately due to no present feed-back.

Imagine if the machine indicated on Tuesday that a market crash was due for Thursday; what will people do? And what if a search for global stability, ended in the machine concluding that humanity should by arranged according to Pareto's law, and that 20% should enslave 80% of mankind? Do we act like Stalin, and ignore the bits we don't like, and heed those we do?

Philip Ball said...

Good questions Jim. What if reliable models of society (if they exist) suggest politically unpalatable conclusions or solutions? That's a possibility that has to be faced - some find it is the case already with game theory and the Prisoner's Dilemma. It's not at all clear that some won't try to edit the results or tailor them to their own advantage. But there again, that happens already with economic modelling.

On abuse of models: sure, that could happen, and sadly it probably will, if history is any guide. See above (it's hard to imagine that the abuses could be much worse than those perpetrated already by free marketeers).

On Adam Curtis: I like his documentaries, but that's not to say I agree with all they say. Last night's eposide was especially contentious. As it happens, I have a comment on this in my latest draft column for Prospect: here is what I say:
All Watched Over By Machines of Loving Grace, the latest documentary by Adam Curtis (BBC 2), offered up Curtis’s trademark polemic, and as ever it was a dish of thought-provoking and contentious fare with a sprinkling of the plain misleading. Over-reliance on and excessive faith in computers is surely a problem worth airing, but in suggesting that this lay behind the current financial crisis, Curtis dangerously misidentified the culprit. “The machines”, he assured us, “created mathematical models that parcelled out the loans and then hedged and balanced them so that there was no risk.” Needless to say, the machines did nothing of the sort. The models were created by financial engineers (quantitative analysts or quants) from outmoded theories that survive through academic inertia and ideological bias. All Watched Over purported to indict the financiers and economic advisers, but they must in fact have been delighted to hear that their pseudo-scientific models could be blamed on the machines used for the number-crunching. Computers had nothing to do with the failures that led to the crash—but they might well be needed to build a better alternative.

JimmyGiro said...

I wonder if Iain McGilchrist likes the Adam Curtis treatment? As it surely pipes directly to the 'right hemisphere' of the mind.

As such, your 'left hemisphere' critique is rational and true, but may miss the broader message of the whole.

True that the machines are just the tools, or the 'messengers'; therefore why blame them for the crimes of the controllers? Similarly, why blame guns when it's people that kill?

What I take from the documentary so far, is that the global influence of the machines, including the world of networks, has allowed the few to control the singularity of the 'world hub'. The efficiency of the machines has enticed the world to put every thing it values into one basket.

By being the best horse, it becomes the only horse, therefore its evolution has ended. And in a one horse race, it beats itself, regardless of what Jockey we elect to ride it.

Alan Greenspan (Adam plays the music for Rand's romantic tragedy, like a Wagnerian leitmotif) is judged the most powerful man in the world; the 'Google' of economics. Therefore the cooperative stability indicated in the pong game, is no longer a true analogy; the better analogy would be to replace the two paddles with a single paddle, that played against itself.

Which brings me back to the original problem with the 'Crisis Observatory' computer: it too will become a super hub, effectively making original thought a near redundancy, thereby allowing a synthetic orthodoxy to form unchallenged.

Electronic fascism anybody?