Imperfect

embracing ai optimism

Lloyd presented prompts from his thought experiment on the topology of consciousness:

So. I'll do something direly radical, and ask you to write tomorrow's escarpment blog post for me. I'll give you just the title. You game for this?

Alright. The title is: "on the topology of consciousness" — base the blog post on much of what we have discussed here on this thread thus far.

Take the large language or diffusion models people dub as "AI". I wish to ponder them more from a basecamp frame as magic. Yet, I also want to apply them from a summit frame as tools.

While I appreciate Lloyd's experiment and wish to replicate both generation and memory aspects, here's an example of my own from a different direction. Many of my post titles have been and will take inspiration from countless regenerations. Many derive themselves from existing outlines or drafts I have workshopped. Sometimes, the assistant strikes gold. More often is the case that I collaborate with the tool to diverge my possibility space. Then, I can combine my taste with the abundance that models afford to converge into the most sensible choice. Are there more efficient instructions or techniques for yielding my desired result? Yes, and I would appreciate your insight on the matter.

You as a human being can fine-tune your toolchain and its output. Meticulous process involvement shows you how well tech under the broad AI umbrella can tap into your idiosyncratic understanding. This breakthrough can be a clever shortcut in the small and the large. What interests close to you could you shrink the required labor, time, and resources to learn about? Could you become a better friend through model output or the practice you attain through instruction prompting? For how much hype-chasing enterprise use runs rampant in the mass media ecosystem, the remainder of practicality wishes to be found.

Fakes and forks

Lloyd's prompts serve as an effective yet implicit disclaimer. If that's not enough, even the model output shows it is but a mere draft. These remind me of times when generative AI involvement slipped through my awareness. Even this early in the innovation cycle, my total is a non-zero number, if not much higher than that. Regardless of medium, getting got can elicit enlightenment, nightmares, or anything in between. What are your priors? One man's "That's awesome!" meets another man's "That's horrifying!" Some see how extreme the accelerating technological developments push at both ends.

Large pockets of the independent web appear to pride themselves over spatial purity. Yet, how many pieces have I read within that aren't as pure as those with extreme expectations of purity would accept? How many pieces have I written and published that fit said mold? Is your outrage over finding out (and not assuming) something is more tool-assisted than preferable justified? Is it so when the output would otherwise meet your needs?

The war is fought on many fronts. What does any given audience watch for when it comes to a good or service which they wish to consume? It can be human ingenuity, engineering, and personality over even superior tools. It can be tooled experiences that get appreciated similarly to or better than less assisted works. Think of the following juxtapositions: process and outcome, features and benefits, and choosing whether to separate the art from the artist or not. How consistent are the ever-changing personas within you when it comes to value allocations across those and other decision spaces?

Make more things

I witness more and more creatives and end users interface AI in developing entire utilities, games, and other digital experiences. Take the game linked in Observer's i built a type script game. Read the note I remember from mgx's when your mom needs a website but only knows google sheets:

Bonus: Gemini is now capable of interacting directly with Google Sheets, updating product descriptions and care instructions will be incredibly easy for her.

Experimenting with emergent technology can be cool. Doing so to rise above disabilities, loneliness, and other impediments is a cherry on top. Be empowered to express yourself creatively and socially when you otherwise wouldn't. The more people admit to having more productive, fulfilling conversations and interactions with AI, the more I think that cybernetic fusion could become a reality over evil overlord or benevolent dictator predictions.

Live through tools

I think the proliferation of AI as a talking point among normal people, mixed with intense oppositional outrage, indicates that "AI has already won" as Ankit says. It's just a tool. It's just a tool like anything else. It's just a tool like anyone else. It's just a tool like anyone's work.

Live your life to its fullest through tools as people and vice versa. Both can serve as nodes to link together within the network of life beyond what the naked eye can see. Everything changes always. Yet, everything is malleable on top of that fundamental reality.

Forgive me if that prescription is too pleasant and transhumanist. The world could use more optimism about the future of humanity and emergent technology. and the accelerating, gradually less unbounded interactions in between. Forces dragging most people down make me think that the correct direction is not only up but also out. That quirk itself can be an edge in this technological landscape.

Want to reach out? Connect with me however you prefer: