Philip Sugg | Blog

Sometime around 1993 in North America–and a little later, maybe 1995, for much of the rest of the world–the number of people who used the internet began to accelerate. More people put more stuff (today this stuff is often called “content”) online: at first mostly text, then images, finally video and all the rest of media. That makes about 30 years in which large numbers of people have been creating for the internet: information that is now used to train huge textual datasets like the language model behind ChatGPT.

I had a vision, really more of a sci-fi premise, about where this goes in another 30 years. It seems reasonable to assume that the next 30 years of the popular internet will involve more AI-generated media than the first. In the time between now and 2053, bots generate so much content that the media from the first, human era of the internet becomes impossibly obscure, far more rare than the handful of known cave paintings from the earliest Homo sapien pre-history. The bot era will a be derivative byproduct of the internet’s human era, but the bots have–for a while–kicked the ladder out from underneath them. The old human internet rots away under piles of bot trash. Humans still “produce content” (maybe–because it seems there is more at stake–they go back to calling it “writing” again), but even the most advanced search engines will be too overwhelmed to find it. The human data source for the bot era still exists, somewhere, online, but it goes darker than anything on the dark web today.

There is so much talk lately of machine intelligence, and what it has or will accomplish. I am impressed and even a little surprised by some of what I have seen, but the biggest effect that the current ChatGPT boom has had on me is that it has led me to seek out more trusted opinions by people who really know how to think. I’ve spent more time than normal seeking out both familiar and novel commentary about what is going on. The cliche has proven true for now: handing more work over to automation has cast me further onto deeper and more essential human capacities: to understand, to analyze, and to reconfigure the world into a new synthesis.

See also

A few days ago I had to take my son somewhere during the day, and I walked by this excavator tearing down an old residential building near the university’s medical center. I don’t know anything about the history of this block, but the buildings were the only buildings in a wide area that appeared to have a residential purpose. They were adjacent to an empty lot, like they’d once been part of a residential area of which they were now the last holdouts. Next time I go by, I expect they’ll be gone.

I’d noticed these buildings before, and thought they looked out of place, like their walls were unclothed, meant to be hidden behind a jumble of similar buildings, zebra-like. Instead, in the last phase of life, they’ve been singled out, home to no one.

Most municipalities in the U.S. have land records about what stood on a piece of property. It would be difficult for any individual to review them on a large scale, and to understand the trends behind the factual archive. For the most part, the building is the archive. Once it is gone, even the people who lived there would have trouble locating the spot anymore. A memory of a physically uprooted place is an orphaned memory–a person can never go back and refresh it.

David Cole points out that originalism, the interpretive doctrine most associated with the U.S. Supreme Court and Constitution, has attracted so much attention in recent eras that his readers need a reminder of the alternatives. Cole writes in the New York Review of Books:

But there is a familiar and sensible alternative—familiar because every judge and justice in the history of the United States has employed it most of the time, even the few who profess on occasion to be originalists. This method starts with the text of the Constitution but recognizes that the ways its broad and open-ended provisions apply will be elucidated gradually over time as judges confront particular cases and seek to make sense of what has gone before, analogize from precedent to contemporary circumstances, and explain their reasoning to provide guidance for the future. The University of Chicago law professor David Strauss has called this method “common law constitutional interpretation.”

Cole would likely agree that a “common law interpretation” of the constitution fails to qualify as a distinct method–and he might argue that this is a good thing.

Cole’s article made me think that a distinguishing quality of other modern traditions of Constitutional interpetation is their willingness to accept the improvisational quality of all interpretative activity.

Maybe originalism’s most defining feature is not a fidelity to any foundational set of figures, sources, or meanings; instead what sets it apart is a striving for interpretive formulas, and the (supposed) elimination of judicial discretion. Cole does not say this, but his description gave me the sense that originalism had more in common with trends like quantification in the behavioral sciences, scientism against cultural explanation, Darwinian materialism–and of course biblical fundamentalism. In a speculative future, we could imagine other “originalisms:” maybe one based on fidelity to the radio communications of an alien civilization, the runic stones of an ancient archaeological site, or the esoteric patterns of falling leaves in a pond. Those could also be “originalist” in the same spirit as constitutional originalism, if the point is to limit interpretative possibility, to make it seem methodical, bound by a foundational source.

Seen recently: the migrations of several hundred Canada Geese above my apartment.

They had likely staged at nearby Washington Park. I saw the same thing within a few days of this date last year.

The migrations of large birds, like the geese seen here, can be hard to ignore. When I stepped outside because of the noise, the birds appeared like they had forgotten how to organize. They cross each other’s path in a roiling mass, like ants on an uncovered mound. They were loud, it was an unsettling sight, as if I was witnessing a disturbance in the natural order:

They begin to fly roughly to the south:

The size of the migrating flocks may have something to do with their proximity to the lake. Kenn Kaufman writes in A Season on the Wind about the spring migration:

All over eastern North America at this moment, or at least all along the advancing edge of daylight, migrating birds are dropping out of the sky and looking for cover. In most places they are so widely dispersed that they’ll go unnoticed. But here, up against the barrier of the lake, numbers build as more and more birds pause and then come down before the water’s edge. In the trees, in the thickets, in the marsh edges, the arriving birds pile up. In the immediate vicinity of the lakeshore, so many small birds will concentrate that they’ll be impossible to ignore.

Visible migrations like this are rare. It’s a reminder of the thin but collectively huge layer of birds that streams across Chicago both fall and spring.