[ExI] Observation

Keith Henson hkeithhenson at gmail.com
Wed May 3 17:00:01 UTC 2023


On Tue, May 2, 2023 at 9:16 PM spike jones via extropy-chat
<extropy-chat at lists.extropy.org> wrote:
>
> -----Original Message-----
> From: Keith Henson <hkeithhenson at gmail.com>
> ubject: Re: [ExI] Observation
>
> On Tue, May 2, 2023 at 3:37 PM spike jones via extropy-chat <extropy-chat at lists.extropy.org> wrote:
>
> snip
>
> >>...  I am inherently optimistic, one of those who believes AI will not kill us all.  It might kill us some, and that part will be bad.  But I don’t think humanity will become extinct.
>
> >...I don't think you are using enough imagination.  How about nobody dies and humans become biologically extinct anyway.
>
> Keith
>
> Possible but unlikely Keith: humans have managed to spread all over the globe and occupy every available ecosystem by modifying our immediate surroundings.

That's true.  Cave bears went extinct largely because humans are
smarter than the average bear.  But now we are talking about AI which
is about as smart as the average human and expected to get much
smarter.  AIs currently lack consciousness and volition but given the
progress in recent years, I would not bet against AIs having them
shortly.

The point is that it is an entirely different game.  I don't think the
past provides much guidance for the future.  Mind you, I have thought
about it a lot, but I can't predict what it will be like on the other
side of the AI singularity.

It will not be long before we find out.

Keith

I think humans may have radiated more thoroughly than any other
species, although I can't really say with respect to rats and
mosquitoes.  We are at least in their class, if we haven't beaten both
those species.
>
> As I wrote that previous comment, it occurred to me that I was feeling good about competing in the same class as mosquitoes and rats.
>
> In any case, I would be so happy if nobody dies, however it very well might already be too late for that.  I experimented with ChatGPT as a health coach.  It does a good job if you don't expect it to act as a doctor.  It answers accurately enough on medical conditions, but the problem with humans is that we are bad to imagine we have conditions that we really don't have.  The internet has made this problem an order of magnitude worse, for it makes it far too easy for people to Google around and find some obscure cool-sounding condition, then somehow convince themselves they have it, because they have experienced five of the top eleven top correlated symptoms.
>
> OK, think of GPT as a super-Google that makes people do still more silly things with respect to their own health.  Then think about how our diet principles work in our times: entire sections of bookstores dedicated to diet principles, when we all know in our hearts what the damn problem really is: we evolved under conditions of scarcity, but suddenly technology gave us any type of food we want in any quantity.  Well hell, of COURSE most people are going to be over their ideal weight and of COURSE most people will eat too much and really really enjoy themselves doing it, but sheesh of COURSE it has negative health consequences.  Well then, given that, most people want to blame anything else for their health problems.  So... we scurry to the internet to find something to blame, when really it is that we are living in a toxic food environment: it is just too good and too plentiful, never mind all the unknown stuff that is added for flavor, color and shelf life.
>
> OK so GPT offers true answers, but the user might manage to delude themself into doing the wrong thing with theys health.  I don't know if GPT has been around long enough to cause anyone to slay themself, but I wouldn't be a bit surprised if someone did by now.  I am confident Google has inadvertently caused people to do the wrong thing and cause harm or fatalities.  It has surely done good of course, probably saving more than it has indirectly slain.  I see GPT as a kind of super-Google.
>
> I probably could have presented that idea with fewer words.  I need to work that problem.  Summary of my breezy commentary: I don't think GPT or any other AI platform will lead to the extinction of humankind, but it may result in some casualties.  There may be suicides from despondency over jobs being replaced, even if we end up with more and better jobs than before, for instance.  Some GPT users will do stupid things.  The internet doesn't necessarily make us smarter, only much better informed.
>
> spike
>
>
> _______________________________________________
> extropy-chat mailing list
> extropy-chat at lists.extropy.org
> http://lists.extropy.org/mailman/listinfo.cgi/extropy-chat



More information about the extropy-chat mailing list