Pat Gunn (dachte) wrote,
Pat Gunn

Unfair as a Judge

More futurist daydreams:

Notion that at some point we may create two "gods" (of a different type of the numerous mythological entities we've made so far) - an advanced expert-system (or maybe theoretical intelligence with no neuroplausibility), which we use as a substitute for human efforts in the sciences (because we don't live long enough, because we don't have time to learn disciplines deeply enough to utilise the crossdisciplinary ties that we're already starting to miss out on, etc), and a reconstructed human.

Our theoretical intelligence gives birth to the reconstructed human - by analysing enough experiments on the brain, by pulling together all the data from single cell recording, MRI, MRS, and other imaging techniques, and by applying all the deductive patterns we've programmed into it, it deduces the regularities and brings our understanding of humans and physics together with an almost audible thunk. We take that and drop it into a simulator, and we have a neuroplausible-but-immortal human living in an environment we create, and it delves into the sciences, philosophy, and other fields of endeavour as broadly as our theoretical intelligence does. A time of two gods.

What would we do if they developed a sophisticated disagreement on something? Would we declare ourselves to be the judge? I suspect most people would make terrible judges of matters like this, particularly if the raw content of the advances they could make would be vast. Perhaps we'd see them attempt to debug the core intelligences of each other...

I, personally, am committed to the idea that logic and maths are pragmatic creations of empiricism, that there is no deep truth in it, and that anything we care to call knowledge is bootstrapped from pragmatism and a commitment to be data-driven. As such, I believe that that debugging each other would in fact be an attempt to reconcile learned knowledge (including knowledge about appropriate pragmatic paths, such as the process of science) with the foundations of the counterpart's mind.

I suspect that this may be a task humanity will be up to only once, and even then only partially. The nature of intelligence in the physical brain and the nature of intelligence in the abstract are both very high-hanging fruit, and I think that once we get close enough to either of them to implement intelligent beings, that will be the high-water mark of our species (if we even make it that far). I've come to be unsure about the "classic" transhumanist singularity - physics may limit future beings' ability to have the pace of growth people have speculated. It's still an interesting idea - already with automated theorem provers (limited as they can be), among many other fields, CPU-intensive approaches are replacing sustained intellectual effort by humans.

I wonder how people will take to having the human brain being understood someday - we already know that we're all "dumb beasts" on many levels, but it'd really drive the point home to have broad models of the mind available. It'd be cute (and worrying) if we had undiscovered bugs exploitable by knowing more of that source - there are some wonderful videos about attentional blindness, but I imagine when we know enough (if it's human parsable within our lifetimes), we'd see newer cousins to all the classic scams as our flashlight points towards higher hanging fruit.

Not exactly on that note, while looking around the bus on the way home, I wondered again about face perception - that people notice the subtle differences better for the human races they see most often (sheltered people think all Chinese look the same). When I see someone particularly ugly of another ethnicity, and someone of that same ethnicity that's attractive, can I be sure that were I to see more of that ethnicity, the ordering there would be preserved? Complication: beauty is partly cultural and partly biological. I imagine that just as there's presumably some general intelligence "g", there's presumably a general attractiveness "a", and cultures shape that "a" by their specific inclinations. Our preference that our friends, business partners, and other company be attractive being an evolutionary preference for good health, are those of other ethnicities/cultures distinct enough that we lose sight of how to judge their health, or is it just that what once was simple and adaptive has lost a lot of that feature in the transition to modern times? Is it just noise in the process when I find someone attractive who isn't "classically beautiful"? I wonder if the emotional/intellectual fulfillment aspect of intelligence has roots in the EEA or if it's just a mechanism that's become sidetracked. A strong negative reaction to the mentally off is probably evolutionarily productive, but has the converse been generally true?

This EvPsych stuff - I'm pretty sure it's on pretty solid grounds (if its toned down a bit - I've commented before that people in the EvPsych movement tend to take some claims that, despite my very limited qualifications to say anything about the field, strike me as far too strong), but it's also kinda freaky.

Finally, amused to think at all the books I actually have on my laptop that I've been slack at reading, yet I still insist that I'd love to have every book I own in a good PDF (non-DRM) format (ideally annotatable too) so I wouldn't ever need to lug books around again. Issue: apart from the obvious function of providing content, books also remind us *to* read - they're like advertisements for things we'd like to do. Maybe having a devoted book-device in addition to a laptop is important to keep us reading? Or maybe a prominent huge icon would do the job?

As a stodgy Unix geek who uses WindowMaker as his WM, I only keep a few things docked because dock icons are HUGE. Right now, there's just a terminal program, the WindowMaker prefs program, 2 clocks (always running, one set at localtime, the other at GMT), and the Gnome filebrowser, plus the clip (a per-workspace dock) I have:

  • Workspace 1: Pidgin
  • Workspace 5: Firefox
  • Workspace 7: Amarok, K3b
  • Workspace 9: GIMP, Transmission
Everything else I access through the terminal or (more rarely) through the menus in gnome-panel.

Maybe I'd be reminded to read books by making one of those huuuge icons on the dock? I wonder. Reading on a computer is fine, it just needs inner advertisements. (I think this is the only legitimate use of advertisements - those that some advertiser uses to get you to buy their product: mental pollution)

It would be interesting-but-sad if we couldn't take advantage of the theoretical lower cost of having access to lots of information (books, etc) to lower the cost of a decent life for most people - do we really need a separate "book" device to replace books when our computers are just fine (if unwieldy) for the role?

I'm kind of bothered by Pink Floyd being fussy about trying to define an appropriate context for their music to be heard. It's not that I particularly care for EMI, but cultural content, in my view, is like a child - once you release it into the world, it will do its own thing and the artist's wishes no longer matter. We'll listen to music in any order we like, in whatever context we like, we'll remix it, we'll edit intros out, we'll make covers, we'll share it with our friends, just like any bit of culture. The proprietary attitude PF has has no place.

Nice weather: pleasant weight loss as good-looking fruits become available and replace sweets

EFF on export rule reform as a mechanism that also affects rates of societal change - just so.


  • Still alive

    Been feeling a bit nostalgic. Not about to return to LiveJournal - their new ownership is unfortunate, but I wanted to briefly note what's been up…

  • Unplugging LJ

    It's about time I pulled the plug on the LJ version of my blog: 1) I'm much more active on G+ than I am with general blogging. I post many times a…

  • Mutual Trust

    I don't know which should be considered more remarkable: That a cat should trust a member of a far larger and stronger species that it can't…

  • Post a new comment


    Anonymous comments are disabled in this journal

    default userpic

    Your reply will be screened

    Your IP address will be recorded 

  • 1 comment