[ExI] Communications a Moore's Law Phenomenon?

Kelly Anderson postmowoods at gmail.com
Mon Mar 18 13:33:32 UTC 2024


On Wed, Mar 13, 2024 at 12:11 PM efc--- via extropy-chat
<extropy-chat at lists.extropy.org> wrote:
> On Wed, 13 Mar 2024, Mike Dougherty via extropy-chat wrote:
> > On Wed, Mar 13, 2024, 12:50 PM Kelly Anderson via extropy-chat <extropy-chat at lists.extropy.org> wrote:
> >
> My take on it is probably twofold. First how much do I care about the
> person writing, and second, how much do I care about the topic of
> discussion.
>
> The less I care about the person, the less attention the message will
> get in what ever channel it might come through. And likewise, the less I
> care about the subject, the less attention the message gets.
>
> This is then applied within three circles, my private world, business
> world and the rest.

This circle of concerns model that begins at intimates, then
acquaintances and finally strangers is pretty familiar. Perhaps our AI
agents will bring our direct attention selectively to those we most
care for or have responsibility for.

> In terms of a potential AI communicating for us, I see it as a firewall
> blocking/dealing with annoying things such as salesmen, booking
> meetings, booking travel, reservations, getting through to support via
> phone/chats etc.

And yet, talking to salespeople can be helpful to your endeavors, if
they turn out to have something we actually need. An AI agent could
have that conversation for us, assumedly with the AIs working for the
salesmen. I kind of like that because it means that companies that
produce good products, but aren't good at more traditional marketing
will be more likely to get my attention. If the best products end up
getting my business instead of the companies that are best at
marketing, then that seems like it would lubricate the economy like a
higher quality oil would lubricate an engine better.

> Likewise I see these recipients to have their own AI:s, so we'd have
> AI:s negotiating/booking times with teach other, and we would only end
> up seeing the result.

Yes, just so.

> When it comes to the amount of messages I have two strategies for
> dealing with the quantity:
>
> 1. Carefully choosing my channels and not exposing my email address
> where it is not supposed to be.
>
> 2. Automated sorting and replying to some emails.
>
> I'm no super user, but my methods were able to successfully deal with
> 250-300 emails per day during my peak in the global IT industry.

So back around 2003, I created a company called Habit 3 software
(after the third Habit of Steven Covey's famous 7, namely, "Put first
things first.") that intended to create an email client that sorted
email in order of importance. I got it working fairly well using what
passed for AI at the time... multiple heuristics... but eventually
abandoned the project because gmail looked like it might go in that
direction. It never did, and I regret abandoning the project now.
Anyway, within that program, I also had a concept of "Roles" so that
if you were in the "Role" of "scoutmaster", scout's emails would
bubble to the top, but if you were in the Role of "Dad" then your
kid's emails would bubble to the top. The idea was to set a specific
amount of time on each role each day, and address the most important
things to that role first, and keep track of what kind of time was
spent on each role, so that you could properly prioritize your time.
Having an AI automatically answer low priority emails in such a system
would be a nice addition these days. I was talked out of marketing
this product to work on another project that was even cooler. That
particular project didn't go terribly well, even though it would still
be cool today. The government ruins almost as many things as religion.

-Kelly



More information about the extropy-chat mailing list