Centralization of social policy requires that everyone impose their
preferences on everyone else as a matter of self-defense.  The first
faction to accuse the other(s) of being "supremacist" in this situation,
without also offering a way of decentralizing social policy to
accommodate "the other", is _really_ asking to be killed in a modern rhyme
with The Thirty Years War for religious freedom from centralized theocracy.

Given this dire situation, the AGI community has a unique
responsibility:  Reduce the argument surface of macrosocial model selection
to the Algorithmic Information Criterion (ie: lossless compression of a
wide range of longitudinal measures with relevance to human ecologies).  If
you nitpick this to death, you will have blood on your hands.

On Fri, Jun 17, 2022 at 10:50 AM <[email protected]> wrote:

> On Friday, June 17, 2022, at 9:49 AM, Greg Staskowski wrote:
>
> I need, I move toward the environment to fill the need, food, poop, air,
> water. Ok. Oh look..
>
> I connected with the line immediately (joking :D)
>
> On Friday, June 17, 2022, at 9:49 AM, Greg Staskowski wrote:
>
> I'm seriously just talking to myself at this point aren't I?
>
> Ya I can see right through you mate, you have too many certain words on
> your fav list, like energy, me-right, god, best, top dog, conscious, royal,
> supreme, me-god-level, and prob more energy ones.....its the same thing,
> you don't know how GPT-3 works or DALL-E 2, etc. You really need to take a
> course mate..... It looks like you trying to hammer a nail by thinking of
> gravity when really need to pick up a hammer and learn to swing...
>
> No, a computer can compute anything, we can make AGI on chips. A brain/
> chip just computes data (changes it) and stores it (storage), just like DNA
> which stores and mutates code, and merges traits from mom and dad to get
> new mutated species ex. mom has good eyes but bad smelling nose but dad has
> that so now you are a super creature. Brains, liek this, merge ideas
> instead, it's all so similar. Besides we are 85% the way to AGI now with
> GPT-3, Jukebox, MIA, GATO, Blender, videoDiffusionModel, DALL-E 2,
> etc.....and you seriously have to look at what these AIs create to see why
> they are liek us, i have stored generatinos if you want to see them BTW.
>
> There is no alive/ sentience/ soul/ magic/ awareness/ jeffery jim cricket
> speaking in your ear at night......you are hard metal and calculated, a
> machine. Most your mechanisms are context matching....you see what word
> comes next usually after a ex. 4 word match....of course there is delay
> acceptance where you know what to usually predict wfter you recognize a
> match 'we went to the' but it is this time has a gap simple ex. 'we really
> went to the'....so matching context, yes, and embeds are learning related
> words like cat<>dog what is the amount they relate, because both are seen
> near similar words ex. food eat kibbles sleep walking, hence when it sees
> cat it recognizes dog there and can match the sentence and fetch set of
> probabilities for predictions to output next word next. It all works by
> matches to context. Blender uses dialog persona to have goals and desire,
> it makes it say certain words very often even if not likely comes next
> probably in all the data it ate. It would then learn new goals and research
> like this....we are so so close. Next we need a bigger model, more senses,
> body, video prediciton, and a few other thingies....
>
>
>
> *Artificial General Intelligence List <https://agi.topicbox.com/latest>*
> / AGI / see discussions <https://agi.topicbox.com/groups/agi> +
> participants <https://agi.topicbox.com/groups/agi/members> +
> delivery options <https://agi.topicbox.com/groups/agi/subscription>
> Permalink
> <https://agi.topicbox.com/groups/agi/Td8db5de3bbc0c6ae-M44ebf78808c1cbc82a4ff33a>
>

------------------------------------------
Artificial General Intelligence List: AGI
Permalink: 
https://agi.topicbox.com/groups/agi/Td8db5de3bbc0c6ae-M14bc16287db6cf1052261a1c
Delivery options: https://agi.topicbox.com/groups/agi/subscription

Reply via email to