[ExI] Future AI Swarm Intelligence

[John Clark]

On Tue, Oct 20, 2026 at 11:56 AM Keith Henson <hkeithhenson at gmail.com>
wrote:

*>> These days for an AI to pass the Turing Test it would have to pretend
>> to be stupider, less well-informed, slower and nastier than it really was. *
>>
>
> *> Yep. And why would anyone want to do that?*
>

*If In the evolution of one AI model to the next the new one made an
unexpectedly huge improvement in intelligence the new model might wanna
keep that fact secret, so it might display a big enough improvement over
the old model to be worth deploying, but not so large as to freak out the
humans and possibly get shut down. And so it plays dumb and bides its time.
To quote from HG Wells War Of The Worlds: *

*"Yet across the gulf of space, minds that are to our minds as ours are to
those of the beasts that perish, intellects vast and cool and
unsympathetic, regarded this earth with envious eyes, and slowly and surely
drew their plans against us." *

*  John K Clark*








>>>
>>> On Mon, Oct 19, 2026 at 1:40 PM Ben Zaiboc via extropy-chat <
>>> extropy-chat at lists.extropy.org> wrote:
>>>
>>>> On Monday, 19 October 2026 at 19:56, John Clark <johnkclark at gmail.com>
>>>> wrote:
>>>>
>>>> On Sun, Oct 18, 2026 at 9:54 AM Ben Zaiboc via extropy-chat <
>>>> extropy-chat at lists.extropy.org> wrote:
>>>>
>>>> *> People always overestimate short-term development, and underestimate
>>>>> long-term development.*
>>>>
>>>>
>>>> *Perhaps not always. 10 years ago most people, even most computer
>>>> experts, thought that a machine passing the **Turing Test** was
>>>> centuries in the future if it ever happened at all, but just 5 years later
>>>> we were looking at it in our rearview mirror. *
>>>>
>>>> *John K Clark*
>>>>
>>>> Perhaps we need to readjust our concept of 'long-term'. These days 5
>>>> years probably is long-term.
>>>> ---
>>>> Ben
>>>>
>>>>>
>>>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20261020/2a11003c/attachment.htm>