Seeking the Signal


an eigenblog

Semiotics in an Age of Artificial Intelligence

Falling just short of a science for lacking standardized hypothesis testing, semiotics, which is the ‘‘formal doctrine of signs,’’ seldom arises in modern conversation, thrice buried in time: firstly by the trading of metaphysics for science; again by nature of its intimidating perspective; once more (and amusingly so) for an abundance of intradisciplinary communication issues. However, as we coquet with realizing of the dream of engineering something like consciousness, understanding meaning and the making thereof is more critical than ever before. Astronomer Carl Sagan once said that we humans are ‘‘the custodians of life’s meaning.’’ Meanwhile, most of us struggle to explain what meaning means, what it is, or whether it even fundamentally exists.
The present rate at which signals propagate across humanity grants thoughtful observers insight: the method by which meaning is produced; the dispersions and aggregations around frequented or weighty ideas, gently at times (e.g. banking) and violently at others (e.g. viral content). Some are maintained for millennia (e.g. buddhism), while others are fleeting (e.g. flappy bird). As we grow older, we see that some ideas exist that didn’t used to (e.g. cryptocurrency), and likewise inversely (e.g. that words are somehow connected to that which they signify). We find that meaning can be engineered (e.g. morality), and does not always adhere to reality (e.g. idealization).
The ability to generate a generator (e.g. to give birth to a child; to make a machine) marks a change in the rate of the shift of the meaning of the interpretants of that which it generates. Consider the long and diverse history of the book, which crosses countless mediums and methods, sacred instrument and anathema. Far from the Middle Ages when the cost of a book (made from animal skins) could easily reach the relative value of a modern automobile, methods of generating documents eventually improved sufficiently to permit concepts like the pulps, print news, and the romance novel. Present self-publishing methods permit just about anyone to generate some form of book. My point is that the rate of change of the normal eigen-book is related to the rate at which books can be produced.
The reason for this is quite simple: interpretants necessarily exist on the manifold of the comprehensible, while the hyperreal propagate by representamen, which exist on the manifold of physical reality. From this perspective, the rate of change of the meaning of a thing is related to the rate of change of the meaning of the ability to produce such a thing. After all, intentional production represents permutation on the manifold of the comprehensible (i.e. ideation), which is existentially bound by physical reality, and thereby subject to general relativity, which is to say that no one thing is twice produced. In other words, given the necessarily singular nature of each production, the rate of production operates on the lower limit of meaning variance (the bulk of the influence thereafter, I suspect, remaining subject to human motivations).
The present state of artificial intelligence presents the capability to rapidly generate content, meanwhile measuring performance and improving according to whatever relationship is to be extremized. Even if a production rate merely influences the lower limit on meaning variance, one such as we now find ourselves able to command will necessarily drive variance to perceptible magnitudes. Indeed, in many ways it already has (e.g. shopping, entertainment, news media). We may have to become more accustomed to more frequent and rapid shifts in meanings (I suspect this effect is already measurable, although I have not yet pursued such data). However, since variance in AI generated content depends in many cases on metrics, all manner of confundities are invited; relationships shrouded by increasingly bizzare and unexpected presentations. The rate of precession of simulacra is bound to increase wildly, and in irrational directions.
Now suppose that we are neither the sole producers nor the sole consumers. Despite the disparity between our inseperable involvement in the process of generating meaning and our understanding thereof, we find ourselves rapidly attempting to engineer our own observing engineers of meaning. This is the passing of the reigns from one species to the next, reducing our relative contribution to hyperreality. Semiotics asserts no moral position, but can elucidate the fundamental action by which humanity’s dominion over hyperreality might be threatened. Perhaps we might be quite pleased; more likely confused into extinction.


Open Information: Regional Databases

While working on a recent analysis, I wanted to collect some regional information, and happened upon the King County Open Data. It’s an impressive collection of well organized data pertaining to the county. It turns out, most high-population cities, counties, and states also have fantastic open-access databases. I went looking, and collected this list of county databases.


deciphering reality

Astronomy is generally regarded as the oldest field of natural science. Human interest in the affairs of heavenly bodies began in time immemorial, and the hereto earliest discovered astronomical records date back some 5,000 years. In the present age, we’ve become comfortable with notions like supernovae, satellite galaxies, time dilation, gravitational lensing; a great many concepts that would have seemed outlandish and scarcely intelligible just a few centuries ago. The theoretical and technological advancements that brought us to this degree of understanding of the universe beyond our clouds span countless disciplines and scales. In our comfort with these many notions, it’s easy to forget that their validation (and otherwise) has all depended on information carried across the universe by (until only recently) one tiny messenger: the photon. Even on the darkest nights, we’re showered with light from countless sources, often after it has passed through various mediums and bounced between many materials. Our own bodies emit light in the infrared and gamma spectra. The noise is cacophonous. Nevertheless, we’ve engineered tools and techniques to extract ample information; sift the signals from oceans of noise, sufficient to confirm the extraordinary theories to which we’ve become accustomed.