This is conceived as an informal and spontaneous annex to my more extensive blog, Grand Strategy: The View from Oregon.

Subscribe to the Grand Strategy Newsletter for regular updates on work in progress.

Discord Invitation

10th March 2018

Post with 1 note

Addendum on Biological Bias

image

Missing from my discussion on biological bias was a recognition that what we usually think of as bias is manifested in conscious judgments that are explicitly formulated and communicated to others by way of language. Kant made a related point in this way:

“…if the world only consisted of lifeless beings, or even consisted partly of living, but yet non-rational beings, the existence of such a world would have no worth whatever, because there would exist in it no being with the least conception of what worth is.” (Immanuel Kant, Critique of Judgment, sec. 87)

Kant’s formulation is in terms of value, but it obtains, mutatis mutandis, for meaning, for knowledge, and ultimately also for existence. At its most general, we might formulate this as conscious agent bias, because an agent must be conscious in order to form judgments that can be biased. If, minimally, there is no such agent, then there is no such bias. Indeed, there is no bias at all.

I have expressed this very imperfectly. It is difficult to formulate this idea precisely, because there are so many conditions and qualifications that must be made that the original idea is in danger of being submerged under them so that we lose sight of the point we were making.

For example, phototropic plants manifest a bias by growing in the direction of the sun. A given species of plant will grow well in one area, under one set of conditions, among a given set of other plants, and will not grow well in another area among a different set of other plants, under another set of conditions. These could be taken as instances of manifesting a biological bias, but they do not involve any conscious agency. We could simply formulate them without reference to biological bias, as when we use the term “phototropism,” but then we miss the opportunity to identify the deep roots of biological bias in the biosphere. Phototropism is not the action of a conscious agent, but it can be taken as a precursor to what will be biological bias in later forms of life that do possess conscious agency.

We get into an even more difficult gray area when we consider other animal species that are almost certainly conscious, but which cannot explicitly formulate judgments. Dolphins and chimpanzees seem to be conscious, but they can’t express their biological bias, their mammalian bias, or their conscious agent bias because it seems that, while conscious, they don’t possess a peer cognitive capacity with human beings. They have experiences, they feel, they may even think, but it seems unlikely that they say to themselves cogito, ergo sum.

Dolphins may be talking about their conscious judgments to each other in a language that we cannot understand, and in the long multi-million year history of primates there may have been some Descartes among chimpanzees that came to explicit self-consciousness and formulated this self-consciousness to itself. If any of this is going on, or has been going on, we don’t know it to be the case.

With dolphins and chimpanzees we share our biological bias and our mammalian bias and our terrestrial bias, but not our anthropic bias (except in so far anthropic bias is constituted by these aforementioned biases). With an artificial consciousness we would share a material or physical bias and perhaps also a common terrestrial bias if the artificial consciousness originated on Earth, but we would not share biological bias or mammalian bias. It is interesting to speculate whether a machine intelligence constructed by human beings would retain some trace of anthropic bias as a result of using ourselves as a model for intelligence.

We could, as a kind of experiment, attempt to construct a form of conscious as different as possible from human consciousness, and one way to define this would be in terms of the fewest biases shared with human consciousness. For every bias that characterizes human consciousness, there is the possibility of the absence of that bias or the antithesis of that bias. If we could build a machine to build intelligent machines, over several generations of machines building machines rather than human beings building machines, any remaining trace of anthropic bias would eventually be transcended. One might argue that this is what is often called the “technological singularity,” but more interesting yet is whether the resulting intelligence would still be recognizable to human beings as being intelligent and as being conscious.    

Since we are not even fully aware of all the biases to which we are subject, if we were building an artificial consciousness to balance and correct our own biases, or, at least to suggest to us alternative ways of viewing the world, we would probably do better to attempt to constitute a set of artificial intelligences specifically engineered to counter known human biases. As with early precision mechanical clocks, made of parts that would counter each other – expanding while another contracted, for example – the precision engineering of human reason needs to build into its structure counters to our natural biases that follow from the kind of beings that we are.

Better human reasoning, then, is a function of an adequate ontology, because we must be able to say what it means to be a biological being, what it means to be a material being, what it means to be a conscious being, and so on for every kind of being that we are, in order to counter the biases that follow from being the kind of beings that we are. I would say that our present-day ontology is perhaps comparable to the development of physics in Newton’s time, and as we bring an improved scientific focus on traditional philosophical questions we are likely to be in for revolutions that overturn classical ontology even as its remains, within its scope, a roughly accurate approximation.

image

Tagged: intelligencebiological biasconsciousnesscogito ergo sumanthropic bias

  1. geopolicraticus posted this