Friday, October 18, 2019

Yikes! I’ve got a lot goin’ on! – Literature, AI, progress, talent search, Human Swarm, camping trip [Ramble 14]

And that’s a problem, because I only have so much time in a day and I can’t do all these things at once, nor can I really advance on each project each day. But how do I sort things out?

Anyhow.... Here’s what I’m working on, in the order that they occur to me.

Heart of Darkness

I’ve thinking about a post that provisionally ‘solves’ the problem of ring composition in that text. The idea is to have the ring-form be an ‘emergent’ [what a high-class weasel word that is!] of purely local decisions, in this case about narrative strategy. If Conrad wants 1) to keep his readers in the dark about Kurtz for almost half of the tale, and 2) wants them also to know most of his story before they meet him, how does he do this? Aren’t these contradictory aims? He does it by the tricky use of double narration. But why, you ask, would he want to do that? One thing at a time, Grasshopper, one thing at a time.

I’m also thinking to submit a book proposal to the Dead Letter Office at Punctum Books.

Intelligence

What is it? I think about this from time to time. The notion that “intelligence is what intelligence tests say it is” is not as empty as it seems–I’ve been through this before on the five factor personality model. But it doesn’t help us to understand AI, because those machines wouldn’t do very well on such a test. What about playing Go better than the best humans? Or the manual dexterity of OpenAI’s current robotic hand? The informal “know how” seems more useful, for it doesn’t carry the baggage that intelligence carries. Much of that baggage, of course, is coupled to the notion of IQ and all that’s associated with it.

Talent Search

I’ve got a bunch of posts on searching for talent that I want to gather into a working paper. I suppose I’ve got two notions:
ONE: The search process has three phases, 1) identify your a pool of prospects, 2) do a crude search of the pool to narrow the field so that you can 3) conclude with a more careful examination of that narrowed field.

TWO: Just what “talent” is can vary. Perhaps we have a continuum from “self-contained” to “context-sensitive” talent. The most self-contained talent is subject to objective measure, e.g. how fast can you sprint? Context sensitive talent is much more difficult to evaluate because you are trying to estimate how a person’s abilities will interact with and perhaps even change their cultural umwelt.
Progress, painful but real

Here I just want to provide a setting for two papers by David Hays. “On the Painfulness of Progress” is in memory of his friend Raoul Naroll, who argued “that progress causes ... pain during periods of transition but alleviates this and other kinds of pain over the long term.” In “Relativism and Progress” Hays examines worldwide statistics on health, wealth, and social life and includes an appendix with 32 scatter plots depicting the data.

The Human Swarm

I’ve got to keep reading and blogging my way through Mark Moffett’s The Human Swarm, which takes a broad view of the nature of society. The idea is to write a formal review (for 3 Quarks Daily) when I’m done and then package the whole set as a working paper.

Visualizing Form and Evolution in Literary History: A response to Nan Z. Da and to Moretti and Sobchuk

That’s the working title for a longish article I’m working on. It will be of a methodological and theoretical nature. I’ve posted a working paper of draft material for the second part of the article, On the direction of literary history: How should we interpret that 3300 node graph in Macroanalysis, Version 2. I’ve sent that to a few folks and gotten useful feedback I’ve also set up a feedback session at Academia.edu, but I haven’t gotten much feedback. Da dismissed that graph in one long paragraph. I explain why she was wrong to do so, suggesting we view that graph as a trace of the activity of a dynamical system.

The first part of the article will about the cultural analog to phylogeny in biological evolution, which Moretti and Sobchuk touched on. The crucial distinction is between a tree-like structure and a network structure. Tree structures are interesting because: If you have a collection of examples that you know are related, you can infer a phylogeny by careful comparison of your examples. You can’t do that with network structures. What does that tell you about the nature historical process in certain highly structured objects (such as life forms, or cultural artifacts and processes)?

My next task is to draft the first half of the article. When I’ve done that I’ll know whether I want to release it in draft form for comment or just write the whole article.

Kisangani 2150

My ongoing project to write a text depicting Central Africa in 2150. This is conceived as a kind of sequel and reply to [dialectical turn on?] Kim Stanley Robinson's New York 2140. It would center on Kisangani, which used to be known as Stanleyville and played the role of Central Station in Heart of Darkness. At the moment it's on (deep) back burner.
BAPC Camping Trip

I took well over 600 photos on last weekend’s camping trip with the board of the Bergen Arches Preservation Coalition. I posted two tree photos on Sunday and a handful of fire photos on Monday. I posted a bunch more photos on Tuesday and, for some reason, decided to frame them with a story. And that meant I had to finish the story.

Why? Because that meant that posting the daily photo or photos became a lot more complicated. Now I had to pick them so they fit a story I had to dream up. Why’d I decide to do this? Because it’s fun. Work, but fun.

I intend to compose and post the final episode after I’ve posted this.

* * * * *

I suppose that article with the long title is the really big item on my agenda. It looks like it’s shaping to be a major statement. As such it will supplement and complement my 2006 paper, Literary Morphology: Nine Propositions in a Naturalist Theory of Form, PsyArt: An Online Journal for the Psychological Study of the Arts, August 2006, Article 060608. 


More later.

No comments:

Post a Comment