Tagged: ai

Evolutionary Optimization and Environmental Coupling

Red QueensCarl Schulman and Nick Bostrom argue about anthropic principles in “How Hard is Artificial Intelligence? Evolutionary Arguments and Selection Effects” (Journal of Consciousness Studies, 2012, 19:7-8), focusing on specific models for how the assumption of human-level intelligence should be easy to automate are built upon a foundation of assumptions of what easy means because of observational bias (we assume we are intelligent, so the observation of intelligence seems likely).

Yet the analysis of this presumption is blocked by a prior consideration: given that we are intelligent, we should be able to achieve artificial, simulated intelligence. If this is not, in fact, true, then the utility of determining whether the assumption of our own intelligence being highly probable is warranted becomes irrelevant because we may not be able to demonstrate that artificial intelligence is achievable anyway. About this, the authors are dismissive concerning any requirement for simulating the environment that is a prerequisite for organismal and species optimization against that environment:

In the limiting case, if complete microphysical accuracy were insisted upon, the computational requirements would balloon to utterly infeasible proportions. However, such extreme pessimism seems unlikely to be well founded; it seems unlikely that the best environment for evolving intelligence is one that mimics nature as closely as possible. It is, on the contrary, plausible that it would be more efficient to use an artificial selection environment, one quite unlike that of our ancestors, an environment specifically designed to promote adaptations that increase the type of intelligence we are seeking to evolve (say, abstract reasoning and general problem-solving skills as opposed to maximally fast instinctual reactions or a highly optimized visual system).

Why is this “unlikely”? The argument is that there are classes of mental function that can be compartmentalized away from the broader, known evolutionary provocateurs. For instance, the Red Queen argument concerning sexual optimization in the face of significant parasitism is dismissed as merely a distraction to real intelligence:

And as mentioned above, evolution scatters much of its selection power on traits that are unrelated to intelligence, such as Red Queen’s races of co-evolution between immune systems and parasites. Evolution will continue to waste resources producing mutations that have been reliably lethal, and will fail to make use of statistical similarities in the effects of different mutations. All these represent inefficiencies in natural selection (when viewed as a means of evolving intelligence) that it would be relatively easy for a human engineer to avoid while using evolutionary algorithms to develop intelligent software.

Inefficiencies? Really? We know that sexual dimorphism and competition are essential to the evolution of advanced species. Even the growth of brain size and creative capabilities are likely tied to sexual competition, so why should we think that they can be uncoupled? Instead, we are left with a blocker to the core argument that states instead that simulated evolution may, in fact, not be capable of producing sufficient complexity to produce intelligence as we know it without, in turn, a sufficiently complex simulated fitness function to evolve against. Observational effects, aside, if we don’t get this right, we need not worry about the problem of whether there are 10 or ten billion planets suitable for life out there.

Singularity and its Discontents

Kimmel botIf a machine-based process can outperform a human being is it significant? That weighty question hung in the background as I reviewed Jürgen Schmidhuber’s work on traffic sign classification. Similar results have emerged from IBM’s Watson competition and even on the TOEFL test. In each case, machines beat people.

But is that fact significant? There are a couple of ways we can look at these kinds of comparisons. First, we can draw analogies to other capabilities that were not accessible by mechanical aid and show that the fact that they outperformed humans was not overly profound. The wheel quickly outperformed human legs for moving heavy objects. The cup outperformed the hands for drinking water. This then invites the realization that the extension of these physical comparisons leads to extraordinary juxtapositions: the airline really outperformed human legs for transport, etc. And this, in turn, justifies the claim that since we are now just outperforming human mental processes, we can only expect exponential improvements moving forward.

But this may be a category mistake in more than the obvious differentiator of the mental and the physical. Instead, the category mismatch is between levels of complexity. The number of parts in a Boeing 747 is 6 million versus one moving human as the baseline (we could enumerate the cells and organelles, etc., but then we would need to enumerate the crystal lattices of the aircraft steel, so that level of granularity is a wash). The number of memory addresses in a big server computer is 64 x 10^9 or higher, with disk storage in the TBs (10^12). Meanwhile, the human brain has 100 x 10^9 neurons and 10^14 connections. So, with just 2 orders of magnitude between computers and brains versus 6 between humans and planes, we find ourselves approaching Kurzweil’s argument that we have to wait until 2040. I’m more pessimistic and figure 2080, but then no one expects the Inquisition, either, to quote the esteemed philosophers, Monty Python.

We might move that back even further, though, because we still lack a theory of the large scale control of the collected software modules needed to operate on that massive neural simulation. At least Schmidhuber’s work used an artifical neural network. The others were looser about any affiliation to actual human information processing, though the LSI work is mathematically similar to some kinds of ANNs in terms of outcomes.

So if analogies only serve to support a mild kind of techno-optimism, we still can think about the problem in other ways by inverting the comparisons or emphasizing the risk of superintelligent machines. Thus is born the existential risk school of technological singularities. But such concerns and planning doesn’t really address the question of whether superintelligent machines are actually possible, or whether current achievements are significant.

And that brings us to the third perspective: the focus on competitive outcomes in AI research leads to only mild advances in the state-of-the-art, but does lead to important social outcomes. These are Apollo moon shots, in other words. Regardless of significant scientific advances, they stir the mind and the soul. It may transform the mild techno-optimism into moderate techo-optimism. And that’s OK, because the alternative is stationary fear.

Evolution, Rationality, and Artificial Intelligence

We now know that our cognitive facilities are not perfectly rational. Indeed, our cultural memory has regularly reflected that fact. But we often thought we might be getting a handle on what it means to be rational by developing models for what good thinking might be like and using it in political, philosophical, and scientific discourse. The models were based on nascent ideas like the logical coherence of arguments, internal consistency, few tautologies, and the consistency with empirical data.

But an interesting and quite basic question is why should we be able to formulate logical rules and create increasingly impressive systems of theory and observations given a complex evolutionary history. We have big brains, sure, but they evolved to manage social relationships and find resources–not to understand the algebraic topology of prime numbers or the statistical oddities of quantum mechanics–yet they seem well suited for these newer and more abstract tasks.

Alvin Plantinga, a theist and modern philosopher whose work has touched everything from epistemology to philosophy of religion, formulated his Evolutionary Argument Against Naturalism (EANN) as a kind of complaint that the likelihood of rationality arising from evolutionary processes is very low (really he is most concerned with the probability of “reliability,” by which means that most conclusions and observations are true, but I am substituting rationality for this with an additional Bayesian overlay).

Plantinga mostly wants to advocate that maybe our faculties are rational because God made them rather than a natural process. The response to this from an evolutionary perspective is fairly simple: evolution is an adaptive process and adaptation to a series of niche signals involves not getting those signals wrong. There are technical issues that arise here concerning how specific adaptation can result in more general rational facilities but we can, at least in principle, imagine (and investigate) bridge rules that extend out from complex socialization to encompass the deep complexities of modern morality and the Leviathan state, and the extension of optimizing spear throwing to shooting rockets into orbit.

I’ve always held that Good Old Fashioned AI that tries to use decision trees created by specification is falling into a similar trap as Plantinga. By expecting the procedures of mind to be largely rational they result in a brittle response to the world that is as impotent as Plantinga’s “hyperbolic doubt” about naturalism. If so, though, it leads to the possibility that the only path to the kind of behavioral plasticity and careful balance of rationality and irrationality that we see as uniquely human is through simulating a significant portion of our entire evolutionary history. This might be formulated as an Evolutionary Argument Against AI (EAAAI), but I don’t think of it as a defeater like that, but as something more like an Evolutionary Argument for the Complexity of AI (and I’ll stop playing with the acronyms now).