Exploration gene

Scientists from the University of Basel in Switzerland studied the exploratory behavior of different species of cichlid fish from Africa’s Lake Tanganyika. The 240 species found in the lake evolved from the same ancestral species but are extremely diverse in shapes, sizes and behaviors. The study measured the exploratory behavior of the different species in the lab and correlated it to their genetics, with one Single Nucleotide Polymorphism (SNP) in one gene accounting for major differences in the propensity to explore, something we could call "curiosity". Physical differences in the wild correlate: the most daring explorers tend to be larger.

They were able to change the exploratory behavior of fish by mutating the target nucleotide. The confirmation that a single nucleotide is responsible for a drastic change in exploratory propensity, and probably lead to the extraordinary diversity of observed phenotypes, is mind blowing.

An additional interesting connection: the human analog of the gene (cacng5b) that is closest to the SNP is associated with psychiatric disorders such as bipolar disorder and schizophrenia. In the fish, "The gene proved very active in the habenula, a region of the cichlid’s brain that registers fear and instigates movement."

Ansel Adams, a generative HI (human intelligence)

The estate of Ansel Adams complained to Adobe on Threads about images generated by invoking an "Ansel Adams-style" landscape. Using hashtag#genai to produce content that is "in the style" of an artist goes against Adobe's Generative AI content policy. I am neither impressed by Adobe's enforcement of its own policy (or its ability to enforce it) nor am I enthused with the rather kitschy output the prompt yielded.

But that gives me a chance to point out one 1931 photograph by Ansel Adams, not of a landscape but of a pair of scissors and thread, aptly titled "Scissors and Thread". While Adams' style is recognizable, what is fascinating is the generative process: he dropped the thread dozens of times until he was satisfied with the pattern. In other words, he generated dozens of possible "solutions" to his problem (creating an interesting pattern) until he found what he was looking for. I don't know whether he knew ahead of time what configuration he wanted but he probably sought a fluid and natural pose of the thread, and I don't know if he was able to modify his dropping method to reinforce the micro-patterns he liked, but we can assume that it was a fairly inefficient process. In the end, it is the eye of the artist intersecting with the generative output that lead to this photograph. In this case, it turns out the generative process is a human exploiting the physics of a falling thread: it is an example of hashtag#genHI, generative human intelligence.

In the same vein, we can think of other generative agents: physical, chemical, algorithmic. But it is the eye of the {artist, engineer, human} that gives meaning to the generative output.