This post was rejected for the following reason(s):

  • LessWrong has a particularly high bar for content from new users and this contribution doesn't quite meet the bar. (We have a somewhat higher bar for approving a user's first post or comment than we expect of subsequent contributions.)

  • Clearer Introduction. It was hard for me to assess whether your submission was a good fit for the site due to its length and that the opening didn’t seem to explain the overall goal of your submission.  Your first couple paragraphs should make it obvious what the main point of your post is, and ideally gesture at the strongest argument for that point. It's helpful to explain why your post is relevant to the LessWrong audience. 

    (For new users, we require people to state the strongest single argument in the post within the introduction, to make it easier to evaluate at a glance whether it's a good fit for LessWrong)

I propose that human amplification using ML tools was, is and will be our future

  • answer to the question everyone started asking back in 2012: "what's the killer app?"
  • my argument here is basic: I see it happening, at least in this particular way
  • I'm seeing a painting and being asked to prove it is there... well, here it is

This amplification I envision along a year-long 2-way training process with "crane" tools

  • that explore, refine & represent the fractalic landscape of the sense plane to us
  • whose controls can be perfectly tailored to the individual operator
  • that will be lifting unimaginable weights under solid human control

Current situation, imho, is ML research plays in a local minima and doesn't look up

  • not connecting back to the human, to create this tool with which you can train
  • it's just targeting a few of those useful flowers in the sense plane and leave it
    • agreed bringing progress, showing experiments, being very interesting
  • but while doors are presented, pursuit lacks courage to go outside at a higher level
  • showing you can throw a rock don't explain how to build a rocket engine
    • even though it's clear you can polish a rock so well
  • myopic investment is not beneficial and we'll gasp back at the hole we dug

So I'm saying forget momentarily about current hype, "AI" and such definitions which I argue constrain actual explorative depth from grasping our future. Take it form a software engineer pondering this very simple idea for years, I want to

  • point out that human amplification is possible, it is our future, we have the tools
  • find some ears open to it so we make it an "obviously, duh" one
  • jolt you to pursue it more ably, as I'm the last of my class if this community were one
  • solve AI doom

When you're amplified you'll be in the main seat of understanding

  • your understanding changes and that will demand/result (in) better tools
  • now there's a more able you which will then repeat.
  • solid human control, in a team of 3 since going alone is never a good idea

AI doom is solved since in the chicken->von Neumann->AI chart we move the goalpost

  • in all "tools that <verb> (to) us" we will be changing the "us" via amplification

 

I'm like a noob traveling to 1900s to tell how I've seen soccer being played today

  • can't do much other than inspire change in open-minded professionals
  • so the "OK, take the ball and show me how it's played" argument has no usefulness

Or in terms of lifting weights I'm optimistic a 63yr old fat dude can beat The Mountain

  • there's many Mountains disagreeing right now :)

0

New Comment