[ExI] Google Just Achieved Mathematical AGI

[Jason Resch]

On Tue, Mar 3, 2026, 5:57 AM John Clark via extropy-chat <
extropy-chat at lists.extropy.org> wrote:

>
> On Mon, Mar 2, 2026 at 9:47 PM <spike at rainier66.com> wrote:
>
> *> I do freely admit AI is advancing a lot faster than I expected even a
>> few months ago.  They are nailing milestones like pickets on a country
>> fence.  *
>
>
> *I agree. Just when I think I understand the rate of AI development I am
> proven wrong.  *
>

Exponential my models don't fit the data. This is clear from looking at how
poorly they fit the progress of computing or the rate of historical change.
Hyperbolic models fit much better.

In a hyperbolic model the amount of time for a fixed amount of change keeps
halving. Eventually as this time approaches 0, the rate of change shoots to
infinity.

We are now in an era where this halving time is on the order of months.
Soon it will be weeks, then days.

When you map out the rates of historic change (even using data sets that
are decades old) they show a singularity point at the end of 2026/start of
2027.

See this for reference:
https://docs.google.com/presentation/d/18jn51f6DXMykCAL6gjZilK27TXAZielm5djcnHuh-7k/edit?usp=drivesdk
(Best to look at on a full desktop computer so you can see the slide notes)

>
> *> How can we objectively define when the singularity occurred?*
>
>
> *When predictions consistently prove to be wrong is the normal definition
> of a singularity. Back in January I predicted that 2025 would be the last
> normal year ... so now it looks like my prediction is on track for being
> proven correct... so my prediction was proven to be wrong... so my
> prediction was right ... so I'm very confused.... so I guess the
> Singularity is happening.   *
>

Population models made in the 1960s pointed to November of 2026 as the time
when human population would shoot to infinity. I think the only wrong
assumption was that since the 60s we've been offloading more and more of
our decision making to machines. And historic rates of change haven't
slowed, and also point to 2027 as the singularity point. So it's not
unreasonable that the population of "decision makers" (humans+AI agents)
will still (effectively) shoot to infinity by that time.


Jason



>
> *John K Clark*
>
>
>
>
>
>
>
>
>>
>> *Google Just Achieved Mathematical AGI*
>> <https://www.youtube.com/watch?v=N_piE0I34gc&t=10s>
>>
>>
>>
>>
>>
>>
> _______________________________________________
> extropy-chat mailing list
> extropy-chat at lists.extropy.org
> http://lists.extropy.org/mailman/listinfo.cgi/extropy-chat
>
-------------- next part --------------
An HTML attachment was scrubbed...
URL: <http://lists.extropy.org/pipermail/extropy-chat/attachments/20260303/836785b1/attachment.htm>