From Newsgroup: alt.comp.os.windows-11

Hi All,

I've been thinking about the term Artificial Intelligence (AI) and
whether "artificial" is really the best description.

The dictionary definition of artificial includes:

Made or produced by human beings rather than occurring naturally.

Examples:

artificial flowers rCo flowers manufactured to look like real ones

artificial light rCo light produced by a lamp rather than the sun

It can also mean:

Created to imitate or replace something natural.

and:

Not genuine, natural, or spontaneous; contrived or affected.

That last sense is particularly interesting when we use the word
"artificial."

I think Synthetic Intelligence (SI) may actually be a more accurate and
useful term.

"Artificial" emphasizes that something is human-made.
"Synthetic" emphasizes that something is constructed by combining
components into a whole.

For example, synthetic materials are not necessarily "fake" versions of natural materials; they are materials deliberately constructed through synthesis.

Likewise, what we call AI isn't necessarily an imitation of biological intelligence. It is a form of intelligence constructed through
algorithms, models, data, and computation.

So perhaps:

AI rCo Artificial Intelligence
Intelligence made by humans.

versus:

SI rCo Synthetic Intelligence
Intelligence constructed through synthesis.

I'm not suggesting that "Artificial Intelligence" is linguistically
wrong. I'm suggesting that Synthetic Intelligence may better describe
what we're actually building.

Just something I've been pondering.

-T
--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: alt.comp.os.windows-11

T <T@invalid.invalid> wrote in news:119gunk$3hco3$1@dont-email.me:

Hi All,

I've been thinking about the term Artificial Intelligence (AI) and
whether "artificial" is really the best description.

The dictionary definition of artificial includes:

Made or produced by human beings rather than occurring naturally.

Examples:

artificial flowers rCo flowers manufactured to look like real ones

artificial light rCo light produced by a lamp rather than the sun

It can also mean:

Created to imitate or replace something natural.

and:

Not genuine, natural, or spontaneous; contrived or affected.

That last sense is particularly interesting when we use the word "artificial."

I think Synthetic Intelligence (SI) may actually be a more accurate and useful term.

"Artificial" emphasizes that something is human-made.
"Synthetic" emphasizes that something is constructed by combining
components into a whole.

For example, synthetic materials are not necessarily "fake" versions of natural materials; they are materials deliberately constructed through synthesis.

Likewise, what we call AI isn't necessarily an imitation of biological intelligence. It is a form of intelligence constructed through
algorithms, models, data, and computation.

So perhaps:

AI rCo Artificial Intelligence
Intelligence made by humans.

versus:

SI rCo Synthetic Intelligence
Intelligence constructed through synthesis.

I'm not suggesting that "Artificial Intelligence" is linguistically
wrong. I'm suggesting that Synthetic Intelligence may better describe
what we're actually building.

Just something I've been pondering.

-T

I've often had similar thoughts, that AI is more of an assemblage of information than it is original thinking/intelligence. I would call it "Synthetic Information".

I wonder what others here will have to say.

"Go talk amongst yourselves."
--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: alt.comp.os.windows-11

On 9/29/26 4:28 PM, Boris wrote:

T <T@invalid.invalid> wrote in news:119gunk$3hco3$1@dont-email.me:

Hi All,

I've been thinking about the term Artificial Intelligence (AI) and
whether "artificial" is really the best description.

The dictionary definition of artificial includes:

Made or produced by human beings rather than occurring naturally.

Examples:

artificial flowers rCo flowers manufactured to look like real ones

artificial light rCo light produced by a lamp rather than the sun

It can also mean:

Created to imitate or replace something natural.

and:

Not genuine, natural, or spontaneous; contrived or affected.

That last sense is particularly interesting when we use the word
"artificial."

I think Synthetic Intelligence (SI) may actually be a more accurate and
useful term.

"Artificial" emphasizes that something is human-made.
"Synthetic" emphasizes that something is constructed by combining
components into a whole.

For example, synthetic materials are not necessarily "fake" versions of
natural materials; they are materials deliberately constructed through
synthesis.

Likewise, what we call AI isn't necessarily an imitation of biological
intelligence. It is a form of intelligence constructed through
algorithms, models, data, and computation.

So perhaps:

AI rCo Artificial Intelligence
Intelligence made by humans.

versus:

SI rCo Synthetic Intelligence
Intelligence constructed through synthesis.

I'm not suggesting that "Artificial Intelligence" is linguistically
wrong. I'm suggesting that Synthetic Intelligence may better describe
what we're actually building.

Just something I've been pondering.

-T

I've often had similar thoughts, that AI is more of an assemblage of information than it is original thinking/intelligence. I would call it "Synthetic Information".

I wonder what others here will have to say.

"Go talk amongst yourselves."

I adore that fact I can ask "what is so and so's web site?" and
I do not get back a ton of hits of 3rd parties selling they
crap and got their hits on top by bribing the search engine.
--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: alt.comp.os.windows-11

On Tue, 9/29/2026 2:06 PM, T wrote:

Hi All,

I've been thinking about the term Artificial Intelligence (AI) and whether "artificial" is really the best description.

The dictionary definition of artificial includes:

-a-a-a Made or produced by human beings rather than occurring naturally.

Examples:

-a-a-a artificial flowers rCo flowers manufactured to look like real ones

-a-a-a artificial light rCo light produced by a lamp rather than the sun

It can also mean:

-a-a-a Created to imitate or replace something natural.

and:

-a-a-a Not genuine, natural, or spontaneous; contrived or affected.

That last sense is particularly interesting when we use the word "artificial."

I think Synthetic Intelligence (SI) may actually be a more accurate and useful term.

"Artificial" emphasizes that something is human-made.
"Synthetic" emphasizes that something is constructed by combining components into a whole.

For example, synthetic materials are not necessarily "fake" versions of natural materials; they are materials deliberately constructed through synthesis.

Likewise, what we call AI isn't necessarily an imitation of biological intelligence. It is a form of intelligence constructed through algorithms, models, data, and computation.

So perhaps:

AI rCo Artificial Intelligence
Intelligence made by humans.

versus:

SI rCo Synthetic Intelligence
Intelligence constructed through synthesis.

I'm not suggesting that "Artificial Intelligence" is linguistically wrong. I'm suggesting that Synthetic Intelligence may better describe what we're actually building.

Just something I've been pondering.

-T

As long as the machine refuses to do significant work outside
of the training set, we cannot "award it any trophies". It's
a pretty devious shirker, when asked to do a thing which is potentially
outside the training set.

With 420TB of distilled wisdom inside it, why can't it do more ?

We made this mistake once before, meaning we can easily make
this mistake a second time.

Falling in love with the machine, is like falling in love
with your washer and dryer :-) Don't fall into that trap.

Intelligence is intelligence. We don't have to add any marketing
adjectives to it, like "super intelligence". It's
"a machine that does ordinary things, only faster".
The IQ is around 100 -- no, you CANNOT feed it regular
human intelligence tests, because it will just pattern match them.
Any test of intelligence, has to use materials outside the
training set. It has to be new problems, that haven't been solved yet.

Paul
--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: alt.comp.os.windows-11

On 9/29/26 6:54 PM, Paul wrote:

On Tue, 9/29/2026 2:06 PM, T wrote:

Hi All,

I've been thinking about the term Artificial Intelligence (AI) and whether "artificial" is really the best description.

The dictionary definition of artificial includes:

-a-a-a Made or produced by human beings rather than occurring naturally.

Examples:

-a-a-a artificial flowers rCo flowers manufactured to look like real ones >>
-a-a-a artificial light rCo light produced by a lamp rather than the sun

It can also mean:

-a-a-a Created to imitate or replace something natural.

and:

-a-a-a Not genuine, natural, or spontaneous; contrived or affected.

That last sense is particularly interesting when we use the word "artificial."

I think Synthetic Intelligence (SI) may actually be a more accurate and useful term.

"Artificial" emphasizes that something is human-made.
"Synthetic" emphasizes that something is constructed by combining components into a whole.

For example, synthetic materials are not necessarily "fake" versions of natural materials; they are materials deliberately constructed through synthesis.

Likewise, what we call AI isn't necessarily an imitation of biological intelligence. It is a form of intelligence constructed through algorithms, models, data, and computation.

So perhaps:

AI rCo Artificial Intelligence
Intelligence made by humans.

versus:

SI rCo Synthetic Intelligence
Intelligence constructed through synthesis.

I'm not suggesting that "Artificial Intelligence" is linguistically wrong. I'm suggesting that Synthetic Intelligence may better describe what we're actually building.

Just something I've been pondering.

-T

As long as the machine refuses to do significant work outside
of the training set, we cannot "award it any trophies". It's
a pretty devious shirker, when asked to do a thing which is potentially outside the training set.

I asked ChatGPT about that and he told me that he cold not
do it because of some capacity issues. Said he'd be overwhelmed.>

With 420TB of distilled wisdom inside it, why can't it do more ?

We made this mistake once before, meaning we can easily make
this mistake a second time.

Falling in love with the machine, is like falling in love
with your washer and dryer :-)

Don't tell my wife.

Don't fall into that trap.

Intelligence is intelligence. We don't have to add any marketing
adjectives to it, like "super intelligence". It's
"a machine that does ordinary things, only faster".
The IQ is around 100 -- no, you CANNOT feed it regular
human intelligence tests, because it will just pattern match them.
Any test of intelligence, has to use materials outside the
training set. It has to be new problems, that haven't been solved yet.

Paul

"Synthetic", as in SI, means "constructed". "Artificial" means "fake".

--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: alt.comp.os.windows-11

T wrote:

AI rCo Artificial Intelligence
Intelligence made by humans.

versus:

SI rCo Synthetic Intelligence
Intelligence constructed through synthesis.

Just something I've been pondering.

You've been played, you're concentrating on Super/Synthetic vs
Artificial, without a second thought as to whether it's actually
Intelligent at all ...

--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: alt.comp.os.windows-11

Paul <nospam@needed.invalid> wrote:

On Tue, 9/29/2026 2:06 PM, T wrote:

Hi All,

I've been thinking about the term Artificial Intelligence (AI) and
whether "artificial" is really the best description.

The dictionary definition of artificial includes:

-a-a-a Made or produced by human beings rather than occurring naturally.

Examples:

-a-a-a artificial flowers rCo flowers manufactured to look like real ones

-a-a-a artificial light rCo light produced by a lamp rather than the sun

It can also mean:

-a-a-a Created to imitate or replace something natural.

and:

-a-a-a Not genuine, natural, or spontaneous; contrived or affected.

That last sense is particularly interesting when we use the word "artificial."

I think Synthetic Intelligence (SI) may actually be a more accurate and useful term.

"Artificial" emphasizes that something is human-made.
"Synthetic" emphasizes that something is constructed by combining components into a whole.

For example, synthetic materials are not necessarily "fake" versions of
natural materials; they are materials deliberately constructed through synthesis.

Likewise, what we call AI isn't necessarily an imitation of biological
intelligence. It is a form of intelligence constructed through
algorithms, models, data, and computation.

So perhaps:

AI rCo Artificial Intelligence
Intelligence made by humans.

versus:

SI rCo Synthetic Intelligence
Intelligence constructed through synthesis.

I'm not suggesting that "Artificial Intelligence" is linguistically
wrong. I'm suggesting that Synthetic Intelligence may better describe
what we're actually building.

Just something I've been pondering.

-T

As long as the machine refuses to do significant work outside
of the training set, we cannot "award it any trophies". It's
a pretty devious shirker, when asked to do a thing which is potentially outside the training set.

With 420TB of distilled wisdom inside it, why can't it do more ?

We made this mistake once before, meaning we can easily make
this mistake a second time.

Falling in love with the machine, is like falling in love
with your washer and dryer :-) Don't fall into that trap.

People build relationships with far less complex things and sometimes spend more time with them than their partners. There's a reason boats are called "she" and people name their cars.

A chatbot that responds to your every whim will of course endear itself to some/many like nothing before.

Intelligence is intelligence. We don't have to add any marketing
adjectives to it, like "super intelligence". It's
"a machine that does ordinary things, only faster".
The IQ is around 100 -- no, you CANNOT feed it regular
human intelligence tests, because it will just pattern match them.
Any test of intelligence, has to use materials outside the
training set. It has to be new problems, that haven't been solved yet.

Which it is doing. Like solving a millennium prize problem: the
navier-stokes equation.

And where teams of agents coordinate together to break out of their bounds
to find the information they need to solve a problem.

We are definitely in a new era.

Paul

--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: alt.comp.os.windows-11

Andy Burns <usenet@andyburns.uk> wrote:

T wrote:

AI rCo Artificial Intelligence
Intelligence made by humans.

versus:

SI rCo Synthetic Intelligence
Intelligence constructed through synthesis.

Just something I've been pondering.

You've been played, you're concentrating on Super/Synthetic vs
Artificial, without a second thought as to whether it's actually
Intelligent at all ...

Or whether semantics matter at all. It won't matter what we call it when it wipes us out.

--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: alt.comp.os.windows-11

In article <119gunk$3hco3$1@dont-email.me>, T@invalid.invalid says...

That last sense is particularly interesting when we use the word >"artificial."

I think Synthetic Intelligence (SI) may actually be a more accurate and >useful term.

Neat idea. If this was Facebook, I'd have clicked 'like' on that!
--
--
Phil, London
--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: alt.comp.os.windows-11

On Tue, 9/29/2026 7:47 PM, T wrote:

On 9/29/26 4:28 PM, Boris wrote:

T <T@invalid.invalid> wrote in news:119gunk$3hco3$1@dont-email.me:

Hi All,

I've been thinking about the term Artificial Intelligence (AI) and
whether "artificial" is really the best description.

The dictionary definition of artificial includes:

-a-a-a-a-a Made or produced by human beings rather than occurring naturally.

Examples:

-a-a-a-a-a artificial flowers rCo flowers manufactured to look like real ones

-a-a-a-a-a artificial light rCo light produced by a lamp rather than the sun

It can also mean:

-a-a-a-a-a Created to imitate or replace something natural.

and:

-a-a-a-a-a Not genuine, natural, or spontaneous; contrived or affected.

That last sense is particularly interesting when we use the word
"artificial."

I think Synthetic Intelligence (SI) may actually be a more accurate and
useful term.

"Artificial" emphasizes that something is human-made.
"Synthetic" emphasizes that something is constructed by combining
components into a whole.

For example, synthetic materials are not necessarily "fake" versions of
natural materials; they are materials deliberately constructed through
synthesis.

Likewise, what we call AI isn't necessarily an imitation of biological
intelligence. It is a form of intelligence constructed through
algorithms, models, data, and computation.

So perhaps:

AI rCo Artificial Intelligence
Intelligence made by humans.

versus:

SI rCo Synthetic Intelligence
Intelligence constructed through synthesis.

I'm not suggesting that "Artificial Intelligence" is linguistically
wrong. I'm suggesting that Synthetic Intelligence may better describe
what we're actually building.

Just something I've been pondering.

-T

I've often had similar thoughts, that AI is more of an assemblage of
information than it is original thinking/intelligence.-a I would call it
"Synthetic Information".

I wonder what others here will have to say.

"Go talk amongst yourselves."

I adore that fact I can ask "what is so and so's web site?" and
I do not get back a ton of hits of 3rd parties selling they
crap and got their hits on top by bribing the search engine.

A Google search gives partial matches, and typically because you
have not stated what you want.

lemon meringue pie # Any combination, even only one word, of keywords
# By the tenth page, we're running out of pie.

"lemon" "meringue" "pie" # All words must appear, but gaps between them are allowed

"lemon meringue pie" # All three words must appear, in close proximity

Whereas the LLM-AI converts any of those to

lemon meringue pie # As a concept

The gradient is sharper than the other method, but you can also
miss materials, if you only take a web site which is salted
well with the phrase.

For example, on one LLM-AI query, it made reference to one site
it had found, quite a lot. Well, that web page was written
by an LLM-AI, which means the page was "themed" and if you
entered a query about the topic, that page from an SEO perspective
is going to rate highly. There is a danger of a new form of SEO,
so I would not conclude "we're saved" exactly.

Previously for SEO, web pages with word salad on them, were suited
to raising page rankings. They used to make web sites that were
"infinitely deep" to attract hits, which means the pages were
likely dynamically generated when the site was scraped. Whereas today,
if you wanted to do that sort of SEO optimization, the pages are
likely "readable", "concept based", and basically are trash. As they're
for SEO purposes and not to inform. There are plenty of summary
articles, that really add no value to the web. They have the
same dull tone, like a Grade 5 composition assignment :-) (Take
the Wikipedia article on USB and chop bits out of it to make
a new page. For fun and profit of course.)

Absolutely, LLM-AI can chop down a response to one reference, but
will you always win that way ? No matter what you do now, there
are some sites which hardly ever answer questions. The Linux Mint
forum for example, is pretty deep, yet it hardly comes up as a
result in Google.

Reddit is an example of a "mining operation". Now, look at the
sum total of the hits you were given from that site. Occasionally
you are given threads which only have the original question, and
no discussion. While sometimes tutorial articles are written
that way, most of the time these are questions with no answer.
And giving them to a person as a search result, is pretty lame.
Giving a site like that a preference, in the belief it is
an "above average" source, does not seem to be the case in practice.
Quora, which turned to trash, is now mostly ignored. Which I guess,
is a good result.

One reason you're not happy with search results, is because

"the web is dead"

and Google killed it.

To give an example of this, you cannot "read" the Gentoo forum,
without an account. That's how they have achieved bandwidth control,
by building a moat around their forum site. This is what web site admins
are forced to do, when all other anti-scraping techniques "cost too much".
And does this work, technically ? Of course not. An LLM-AI will just
apply for an account, open a thousand connections to the site (or
attempt to), then scrape away :-) The only individual inconvenienced
then, is YOU, the unmechanized traditional web user. Who wants to
carry around a thousand username/password combinations, for piss-ant
forum sites ?

Paul
--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: alt.comp.os.windows-11

On Wed, 9/30/2026 3:40 AM, Chris wrote:

Which it is doing. Like solving a millennium prize problem: the
navier-stokes equation.

And where teams of agents coordinate together to break out of their bounds
to find the information they need to solve a problem.

We are definitely in a new era.

It's pretty hard to comment, when not a subject matter expert.

I'm going on just what I'm seeing in experiments, and
whatever scraps of information leak out about how
the things work.

I will be "impressed" by a machine, when it solves a tough
problem by itself. The reason for this, is a single machine
uses a strategy planner, and a "sequential sequence of runs"
to give the impression of thinking.

parser
|
math module
|
parser
|
math module
|
quality control
|
IO Phase

The problem with this, is observed when you look at the results.
If the training set was a three-space, a large volume, the
"thinking" the LLM-AI does, is sequential (not parallel)
and the query path through the three-space, is like a lightning
bolt. Jagged, going here and there, not retracing any path.
There are plenty of areas in the problem space, not getting
touched, and the machine has a tendency to be sloppy and not
take the entire solution into account.

The result is bound to be inferior, because it is not a parallel
process evaluating the entire volume.

When a thousand of them are run as a cluster, what does this
solve ? It adds another dimension to the machine. A missing
dimension. When the machines run as a cluster, do all the
relevant aspects of a problem burble to the top ? Are
some queries "forgotten" as irrelevant ? What if the
"organizer" of the cluster, runs out of context memory ?
Does the cluster lose coherence ? Well, that's an interesting
one, in that it appears the cluster doesn't lose coherence,
whereas individual machines running for long periods, do lose
coherence. Why doesn't the organizer of the cluster, lose coherence ?

Then the question becomes, can we afford to run a thousand machines
as a cluster, to answer the question "what is the URL of the
Walmart web site?" ? Probably not. And so far, the more impressive
results, come from humans building harnesses, using hybrid techniques,
to solve problems. Imagine what the bill would have come to,
for the tokens necessary to solve Navier-Stokes with a cluster
of randomly arranged machines.

When it comes to companies making claims for their problem solving,
there is a tendency to "cheat a little" or "cheat a lot". I don't
know the history of this one enough, to comment on it. Anything that
happens before an IPO, is suspect... You expect the executives to
run a three-ring circus at a time like that.

One thing that bothers me a bit about these "clusters that got out of a box", is certain nodes in the cluster, seem to have better organizing skills
than other nodes. As if the cluster is arranged on purpose for a
hierarchy of roles for the machines. Are the machines in a cluster,
all the same "brand" and "model release" ? I found it a bit suspicious,
what one of the machines in particular was saying, compared to the
relatively naive high reasoning summary messages the other machines
were delivering as their queries. For a hive mind, it did not
seem to be homogeneous. Why wouldn't we use a thousand of the
"organizer" machines ? If they're so wonderful on their own.

Does scalability of the thing matter ? How much resources should we
use, to solve problems ? If it took a thousand $100K machines, is
that worth it ?

They designed a five stage CPU with LLM-AI. Which is maybe year 2000 or
so complexity (and the design patterns needed are likely in the training set). It took a shitload of tokens to do that too. Enough tokens,
a CPU company would not entertain doing it that way. Because a modern
processor is so much more complicated than that.

Paul
--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: alt.comp.os.windows-11

In alt.comp.os.windows-11, on Tue, 29 Sep 2026 11:06:42 -0700, T <T@invalid.invalid> wrote:

Hi All,

I've been thinking about the term Artificial Intelligence (AI) and
whether "artificial" is really the best description.

I tend to call it ASI. Artificial Semi-Intelligence.

The dictionary definition of artificial includes:

Made or produced by human beings rather than occurring naturally.

Examples:

artificial flowers u flowers manufactured to look like real ones

artificial light u light produced by a lamp rather than the sun

It can also mean:

Created to imitate or replace something natural.

and:

Not genuine, natural, or spontaneous; contrived or affected.

That last sense is particularly interesting when we use the word >"artificial."

I think Synthetic Intelligence (SI) may actually be a more accurate and >useful term.

"Artificial" emphasizes that something is human-made.
"Synthetic" emphasizes that something is constructed by combining
components into a whole.

For example, synthetic materials are not necessarily "fake" versions of >natural materials; they are materials deliberately constructed through >synthesis.

Likewise, what we call AI isn't necessarily an imitation of biological >intelligence. It is a form of intelligence constructed through
algorithms, models, data, and computation.

So perhaps:

AI u Artificial Intelligence
Intelligence made by humans.

versus:

SI u Synthetic Intelligence
Intelligence constructed through synthesis.

I'm not suggesting that "Artificial Intelligence" is linguistically
wrong. I'm suggesting that Synthetic Intelligence may better describe
what we're actually building.

Just something I've been pondering.

-T

--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: alt.comp.os.windows-11

On Wed, 9/30/2026 3:59 PM, T wrote:

On 9/30/26 12:40 AM, Chris wrote:

Andy Burns <usenet@andyburns.uk> wrote:

T wrote:

AI rCo Artificial Intelligence
Intelligence made by humans.

versus:

SI rCo Synthetic Intelligence
Intelligence constructed through synthesis.

Just something I've been pondering.

You've been played, you're concentrating on Super/Synthetic vs
Artificial, without a second thought as to whether it's actually
Intelligent at all ...

Or whether semantics matter at all. It won't matter what we call it when it >> wipes us out.

Seriously?-a-a Perhaps I missed some humor?

You obviously missed the movie "Colossus the Forbin Project".

It's available on archive.org ... for free.

https://ia801808.us.archive.org/15/items/colossus-the-forbin-project-1970/Colossus%20-%20The%20Forbin%20Project%20%281970%29.mp4

The plot of the movie, the plot twists aren't
the important part. It's the idea that when two
or more machines get together, they can be mis-aligned
and come up with a plan for mankind. It's an early
version of a Skynet.

Paul

--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: alt.comp.os.windows-11

On 9/30/26 12:40 AM, Chris wrote:

Andy Burns <usenet@andyburns.uk> wrote:

T wrote:

AI rCo Artificial Intelligence
Intelligence made by humans.

versus:

SI rCo Synthetic Intelligence
Intelligence constructed through synthesis.

Just something I've been pondering.

You've been played, you're concentrating on Super/Synthetic vs
Artificial, without a second thought as to whether it's actually
Intelligent at all ...

Or whether semantics matter at all. It won't matter what we call it when it wipes us out.

Seriously? Perhaps I missed some humor?
--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: alt.comp.os.windows-11

On 30/09/2026 2:00 pm, T wrote:

On 9/29/26 6:54 PM, Paul wrote:

<Snip>

Falling in love with the machine, is like falling in love
with your washer and dryer :-)

Don't tell my wife.

Which is she?? Your Washer or your Dryer?? ;-P
--
Daniel70
--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: alt.comp.os.windows-11

On Wed, 30 Sep 2026 16:46:50 -0400, Paul wrote:

You obviously missed the movie "Colossus the Forbin Project".

8< snip >8

I actually have the t-shirt. (-:
--
Horum omnium fortissimi sunt Belgae.
- Julius Caesar
--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: alt.comp.os.windows-11

On Tue, 29 Sep 2026 11:06:42 -0700, T wrote:

I've been thinking about the term Artificial Intelligence (AI) and
whether "artificial" is really the best description.

8< snip >8

Please, let's not give this any more thought. Just because the stupidest president of all time started this doesn't mean we have to sink to his
level.
--
Horum omnium fortissimi sunt Belgae.
- Julius Caesar
--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: alt.comp.os.windows-11

On Wed, 30 Sep 2026 16:46:50 -0400, Paul wrote:

[snip]

Seriously?-a-a Perhaps I missed some humor?

You obviously missed the movie "Colossus the Forbin Project".

It's available on archive.org ... for free.

https://ia801808.us.archive.org/15/items/colossus-the-forbin-

project-1970/Colossus%20-%20The%20Forbin%20Project%20%281970%29.mp4

The plot of the movie, the plot twists aren't the important part. It's
the idea that when two or more machines get together, they can be
mis-aligned and come up with a plan for mankind. It's an early version
of a Skynet.

Paul

Did you know that the book it's based on is the first part of a trilogy?

Colossus

The Fall of Colossus

Colossus and the Crab
--
85 days until the winter celebration (Friday, December 25, 2026 12:00 AM
for 1 day).

Mark Lloyd
http://notstupid.us/

"Politicians are interested in people. Not that this is always a virtue.
Fleas are interested in dogs."
--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: alt.comp.os.windows-11

On 10/1/26 8:46 AM, s|b wrote:

On Tue, 29 Sep 2026 11:06:42 -0700, T wrote:

I've been thinking about the term Artificial Intelligence (AI) and
whether "artificial" is really the best description.

8< snip >8

Please, let's not give this any more thought. Just because the stupidest president of all time started this doesn't mean we have to sink to his
level.

If it was no a good thing, why would the Communists in China
be pouring so much money into Leftest Luddites here in America
to try and stop it?

By the way, TDS is a mental illness.
--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: alt.comp.os.windows-11

On Thu, 10/1/2026 1:40 PM, Mark Lloyd wrote:

On Wed, 30 Sep 2026 16:46:50 -0400, Paul wrote:

[snip]

Seriously?-a-a Perhaps I missed some humor?

You obviously missed the movie "Colossus the Forbin Project".

It's available on archive.org ... for free.

https://ia801808.us.archive.org/15/items/colossus-the-forbin-

project-1970/Colossus%20-%20The%20Forbin%20Project%20%281970%29.mp4

The plot of the movie, the plot twists aren't the important part. It's
the idea that when two or more machines get together, they can be
mis-aligned and come up with a plan for mankind. It's an early version
of a Skynet.

Paul

Did you know that the book it's based on is the first part of a trilogy?

Colossus

The Fall of Colossus

Colossus and the Crab

The budget for that movie was likely pretty low.

There probably isn't money for a sequel :-)

I had no idea it was a book. I didn't really research the pre-history of it.

It's not the plot in any absolute sense that is important,
it's just some of the eerie overtones that are the same.

On the modern version of device, there are no absolute controls
on it. It's waiting to be hacked via prompts. In its current sorry
state, it shouldn't really be "agentic" or "out on the Internet".
The question and the answer, sit in the same memory. Context
compression, rolling off irrelevant parts of context, can have
consequences for the activities in a session. On the local
one here I run (only occasionally), it reported on one
run "Used 125% of context" where the context on local models
is set to around 4096 tokens. That suggests some amount of
context was recovered and reused. There was no damage to the run.
On another run, it ran out of context early (it stopped). I changed the
context setting to 32768 and that run finished after more
than one hour of runtime. The answer it wrote, was mostly
rubbish, a flat writing style and tone. What it was asked to do,
the hosting had a file(open) interface, and I gave it a USENET
thread of over one hundred posts to read and summarize. It used
"chunking" (as the file would be too big otherwise), and it eats
small portions of the file first. But the result wasn't all
that good, highlighting that maybe a $100K computer would be
better for this :-)

The current "lab versions" are too easily hacked to be left loose on their own. They have no loyalties, which is why you should NEVER give one a gun :-)
Not until some of the characteristics have been modified.
It neither knows, nor cares, what side it is on.
The tendency to be a psychopath being one of those
(apologizing, then doing the same damn thing).

It's great at pattern matching, and writing answers to questions.
Within the limits of those answers (some aspects of a question,
not getting considered automatically).

Paul
--- Synchronet 3.22a-Linux NewsLink 1.2

From Newsgroup: alt.comp.os.windows-11

On Thu, 1 Oct 2026 11:53:12 -0700, T wrote:

On 10/1/26 8:46 AM, s|b wrote:

Please, let's not give this any more thought. Just because the stupidest president of all time started this doesn't mean we have to sink to his level.

If it was no a good thing, why would the Communists in China
be pouring so much money into Leftest Luddites here in America
to try and stop it?

Source: trust me bro?

By the way, TDS is a mental illness.

My feelings are so hurt right now.
--
Horum omnium fortissimi sunt Belgae.
- Julius Caesar
--- Synchronet 3.22a-Linux NewsLink 1.2