subject:"NYTimes.com\: We Need to Talk About How Good A.I. Is Getting"

Re: NYTimes.com: We Need to Talk About How Good A.I. Is Getting

2022-08-25 Thread Brent Meeker

On 8/25/2022 1:00 PM, smitra wrote:
If intelligent life that's way more intelligent than insects gets
replaced by AI systems with insect level intelligence together with
all other life except microbes on a planet, then they are back where
biology was in the Cambrian era.

The fundamental issue is that intelligence creatures will create tools
that will make the work they need to do easier. The tools become ever
more sophisticated, so that a lot more work can be done. At some point
we get machines and then we get machines with some level of
intelligence and then we get to a point where the machines do all the
work themselves including the work needed to repair and build themselves.

And then Darwinian evolution machines can take off. Presumably we will
suppress this, just as evolved bacteria have suppressed in new
origination of life on Earth. But it's not a sure thing.

Brent

But this is going to to be reached when the typical intelligence of
the machines is way less than that of the intelligent creatures that
gave rise to the machines.

Saibal

On 25-08-2022 12:12, Telmo Menezes wrote:

I suspect AI is already past insect-level for some tasks, e.g. image
recognition and language understanding. The number of parameters in a
state-of-the-art huge language model or in something like DALL-E 2
means that these are probably already more complex than an insect
nervous system. I might be wrong.

In any case, what I don't understand about the supposed solution to
the "Fermi paradox" is: why would they expect technological-life
evolution to stagnate? In fact I think this reinforces the Fermi
paradox. Why aren't we being visited by alien AIs? Of course the usual
hypothesis apply: the universe is too young, life is to rare, etc etc

Telmo

Am Mi, 24. Aug 2022, um 19:28, schrieb smitra:

Thanks for sharing! My comment on this article:

https://nyti.ms/3dQoxqU#permid=120043436

"It is inevitable that AI systems will end up becoming good enough to
run the economy, repair and reproduce themselves. Biology shows us that
this does not require highly intelligent systems. As things stand now,
even insects outperform our best AI systems, but then we may not even
need insect-level intelligence to fully automatize our economy.

This development is then driven by economic growth, it's not something
that's easy to regulate. Companies will use whatever technology is
available to reduce costs and to get to higher profits. The current
climate crisis shows just how hard it is to regulate the rather simple
process of our use of energy to reduce CO2 emissions.

When in the future the economy is run by autonomous machines that
maintain and copy each other while producing all the stuff we consume,
there will exists a new machine biology besides the original biology.
It's then inevitable that the machine biology will not be fully
compatible with the original biology. Toxic compounds are likely to be
produced.

The problem we'll then face is that we'll have even less power to
mitigate such problems than we have now when dealing with our CO2
emissions. It's then likely that the new machine biology will destroy
most of the original biology.

All intelligent life in the universe likely ends in this way. The
takeover by machines with insect-level intelligence or less, then
explains why the galaxy hasn't already been colonized (the so-called
Fermi Paradox)."

Saibal

On 24-08-2022 14:39, John Clark wrote:

Check out this article from The New York Times. Because I'm a
subscriber, you can read it through this gift link without a
subscription.

We Need to Talk About How Good A.I. Is Getting

We’re in a golden age of progress in artificial intelligence. It’s
time to start taking its potential and risks seriously.

https://www.nytimes.com/2022/08/24/technology/ai-technology-progress.html?unlocked_article_code=CEIPuomT1JKd6J17Vw1cRCfTTMQmqxCdw_PIxftm3iWka3DFDmweiPgYCIiG_EPKarskbtp2wzmQRNlGNLggVblq1OhQJUF2UE-ovp6A0twjEhkClLiSDCkwzo6fGvcx6yPrZW20b710ybPitBzZdWLoUKLA1XV2IRI1qJpmaV372SYKlazAReYl3cJsnqt0XuAMTjgFbCCLv_TjGk8-bI3ANkeAn1FwD-JJWjjTnsqe4qYAdWhRClHHRXB44wUs-Y8WeYNXbOukcUlWKIepiq4RC2doMI6iG5YwIoDUnL9gurLMwgeevnYkS2GsPvx_F8Tqd-ALMQ=em-share

--
You received this message because you are subscribed to the Google
Groups "Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send
an email to everything-list+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/everything-list/CAJPayv21vyY7ZgNzkcVQ1wYr_6dxYq5LyveYMZZrvkig%2B_oV6A%40mail.gmail.com

[1].

Links:
--
[1]
https://groups.google.com/d/msgid/everything-list/CAJPayv21vyY7ZgNzkcVQ1wYr_6dxYq5LyveYMZZrvkig%2B_oV6A%40mail.gmail.com?utm_medium=email_source=footer

--
You received this message because you are subscribed to the Google
Groups "Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send
an email to

Re: NYTimes.com: We Need to Talk About How Good A.I. Is Getting

2022-08-25 Thread smitra

If intelligent life that's way more intelligent than insects gets
replaced by AI systems with insect level intelligence together with all
other life except microbes on a planet, then they are back where biology
was in the Cambrian era.

The fundamental issue is that intelligence creatures will create tools
that will make the work they need to do easier. The tools become ever
more sophisticated, so that a lot more work can be done. At some point
we get machines and then we get machines with some level of intelligence
and then we get to a point where the machines do all the work
themselves including the work needed to repair and build themselves. But
this is going to to be reached when the typical intelligence of the
machines is way less than that of the intelligent creatures that gave
rise to the machines.

Saibal

On 25-08-2022 12:12, Telmo Menezes wrote:

Telmo

Am Mi, 24. Aug 2022, um 19:28, schrieb smitra:

Thanks for sharing! My comment on this article:

https://nyti.ms/3dQoxqU#permid=120043436

"It is inevitable that AI systems will end up becoming good enough to
run the economy, repair and reproduce themselves. Biology shows us
that

this does not require highly intelligent systems. As things stand now,
even insects outperform our best AI systems, but then we may not even
need insect-level intelligence to fully automatize our economy.

Saibal

On 24-08-2022 14:39, John Clark wrote:

Check out this article from The New York Times. Because I'm a
subscriber, you can read it through this gift link without a
subscription.

We Need to Talk About How Good A.I. Is Getting

We’re in a golden age of progress in artificial intelligence. It’s
time to start taking its potential and risks seriously.

--
You received this message because you are subscribed to the Google
Groups "Everything List" group.
To unsubscribe from this group and stop receiving emails from it,
send

an email to everything-list+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/everything-list/CAJPayv21vyY7ZgNzkcVQ1wYr_6dxYq5LyveYMZZrvkig%2B_oV6A%40mail.gmail.com
[1].

Links:
--
[1]
https://groups.google.com/d/msgid/everything-list/CAJPayv21vyY7ZgNzkcVQ1wYr_6dxYq5LyveYMZZrvkig%2B_oV6A%40mail.gmail.com?utm_medium=email_source=footer

--
You received this message because you are subscribed to the Google

Re: NYTimes.com: We Need to Talk About How Good A.I. Is Getting

2022-08-25 Thread smitra

On 24-08-2022 19:41, John Clark wrote:

On Wed, Aug 24, 2022 at 1:28 PM smitra wrote:

All intelligent life in the universe likely ends in this way. The

takeover by machines with insect-level intelligence or less, then
explains why the galaxy hasn't already been colonized (the so-called

Fermi Paradox)."

If human beings go extinct because they are outsmarted by something
"_with insect-level intelligence or less_" then they deserve to go
extinct; but I haven't found many insects that can play Chess and GO
at a superhuman level, or make original paintings that are far far
better than anything I can do if I were just told to paint "infinite
joy". This is what a computer thinks it looks like:

We're currently struggling getting climate change under control by
fixing the way we use energy. What's the intelligence in the system
responsible for getting fossil fuels out of the ground to generate power
for our economy? It's far simpler than the metabolic processes in a
cell.

It's true that we can do amazing things with AI. Of course, an insect
cannot play chess, but we need to consider here that we're training an
AI system for some specific task like playing chess, while an insect has
to do many different tasks that have to do with staying alive. It has to
find food, stay away from predators etc. etc.

To compare an AI to an insect you must see how well they perform when
their entire machinery is devoted to the same task. Since we can't hack
an insect's brain to optimize it for playing chess, we then need to let
an AI simulate a virtual insect to see how well it performs compared to
a real insect. There then isn't a big difference in performance anymore.
We know that real brains function differently from AI systems, AI
systems use a lot of brute force that real brains are able to avoid:

https://www.technologyreview.com/2018/02/19/145532/why-even-a-moths-brain-is-smarter-than-an-ai/

But this is actually good news, as it means that a lot of progress can
still be made.

There is no contradiction with an AI being able to do many things better
than we can. Your brain can also do many things that you can't do well
consciously. For example, controlling the muscles when you walk is done
unconsciously. If you had to consciously control all the muscles needed
for walking including keeping balance, you would be unable to walk.
Walking would seem to be a task that is way beyond human control.

Saibal

--
You received this message because you are subscribed to the Google Groups
"Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to everything-list+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/everything-list/789bf05e653c6844f4ac0e24b1c7fdc4%40zonnet.nl.

Re: NYTimes.com: We Need to Talk About How Good A.I. Is Getting

2022-08-25 Thread John Clark

On Thu, Aug 25, 2022 at 6:13 AM Telmo Menezes 
wrote:

*> Why aren't we being visited by alien AIs? *


Yes, that is the big question. The most obvious answer is because we are
the first, and sometimes the most obvious answer turns out to be correct.

John K ClarkSee what's on my new list at  Extropolis


1mz

-- 
You received this message because you are subscribed to the Google Groups 
"Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to everything-list+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/everything-list/CAJPayv3WRLJjkunxsd0Yfv75_tWxS%3D9-k8M8mZjKGE0-K%3DM6jA%40mail.gmail.com.

Re: NYTimes.com: We Need to Talk About How Good A.I. Is Getting

2022-08-25 Thread Telmo Menezes

I suspect AI is already past insect-level for some tasks, e.g. image 
recognition and language understanding. The number of parameters in a 
state-of-the-art huge language model or in something like DALL-E 2 means that 
these are probably already more complex than an insect nervous system. I might 
be wrong.

In any case, what I don't understand about the supposed solution to the "Fermi 
paradox" is: why would they expect technological-life evolution to stagnate? In 
fact I think this reinforces the Fermi paradox. Why aren't we being visited by 
alien AIs? Of course the usual hypothesis apply: the universe is too young, 
life is to rare, etc etc

Telmo

Am Mi, 24. Aug 2022, um 19:28, schrieb smitra:
> Thanks for sharing! My comment on this article:
>
> https://nyti.ms/3dQoxqU#permid=120043436
>
> "It is inevitable that AI systems will end up becoming good enough to 
> run the economy, repair and reproduce themselves. Biology shows us that 
> this does not require highly intelligent systems. As things stand now,  
> even insects outperform our best AI systems, but then we may not even 
> need insect-level intelligence to fully automatize our economy.
>
> This development is then driven by economic growth, it's not something 
> that's easy to regulate. Companies will use whatever technology is 
> available to reduce costs and to get to higher profits. The current 
> climate crisis shows just how hard it is to regulate the rather simple 
> process of our use of energy to reduce CO2 emissions.
>
> When in the future the economy is run by autonomous machines that 
> maintain and copy each other while producing all the stuff we consume, 
> there will exists a new machine biology besides the original biology. 
> It's then inevitable that the machine biology will not be fully 
> compatible with the original biology. Toxic compounds are likely to be 
> produced.
>
> The problem we'll then face is that we'll have even less power to 
> mitigate such problems than we have now when dealing with our CO2 
> emissions. It's then likely that the new machine biology will destroy 
> most of the original biology.
>
> All intelligent life in the universe likely ends in this way. The 
> takeover by machines with insect-level intelligence or less, then 
> explains why the galaxy hasn't already been colonized (the so-called  
> Fermi Paradox)."
>
> Saibal
>
> On 24-08-2022 14:39, John Clark wrote:
>> Check out this article from The New York Times. Because I'm a
>> subscriber, you can read it through this gift link without a
>> subscription.
>> 
>> We Need to Talk About How Good A.I. Is Getting
>> 
>> We’re in a golden age of progress in artificial intelligence. It’s
>> time to start taking its potential and risks seriously.
>> 
>> https://www.nytimes.com/2022/08/24/technology/ai-technology-progress.html?unlocked_article_code=CEIPuomT1JKd6J17Vw1cRCfTTMQmqxCdw_PIxftm3iWka3DFDmweiPgYCIiG_EPKarskbtp2wzmQRNlGNLggVblq1OhQJUF2UE-ovp6A0twjEhkClLiSDCkwzo6fGvcx6yPrZW20b710ybPitBzZdWLoUKLA1XV2IRI1qJpmaV372SYKlazAReYl3cJsnqt0XuAMTjgFbCCLv_TjGk8-bI3ANkeAn1FwD-JJWjjTnsqe4qYAdWhRClHHRXB44wUs-Y8WeYNXbOukcUlWKIepiq4RC2doMI6iG5YwIoDUnL9gurLMwgeevnYkS2GsPvx_F8Tqd-ALMQ=em-share
>> 
>> 
>>  --
>> You received this message because you are subscribed to the Google
>> Groups "Everything List" group.
>> To unsubscribe from this group and stop receiving emails from it, send
>> an email to everything-list+unsubscr...@googlegroups.com.
>> To view this discussion on the web visit
>> https://groups.google.com/d/msgid/everything-list/CAJPayv21vyY7ZgNzkcVQ1wYr_6dxYq5LyveYMZZrvkig%2B_oV6A%40mail.gmail.com
>> [1].
>> 
>> 
>> Links:
>> --
>> [1]
>> https://groups.google.com/d/msgid/everything-list/CAJPayv21vyY7ZgNzkcVQ1wYr_6dxYq5LyveYMZZrvkig%2B_oV6A%40mail.gmail.com?utm_medium=email_source=footer
>
> -- 
> You received this message because you are subscribed to the Google 
> Groups "Everything List" group.
> To unsubscribe from this group and stop receiving emails from it, send 
> an email to everything-list+unsubscr...@googlegroups.com.
> To view this discussion on the web visit 
> https://groups.google.com/d/msgid/everything-list/468ea47509d9246593730ed39b099f9c%40zonnet.nl.

-- 
You received this message because you are subscribed to the Google Groups 
"Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to everything-list+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/everything-list/fe88eebb-d75d-4f3a-9fa0-657f218c2802%40www.fastmail.com.

Re: NYTimes.com: We Need to Talk About How Good A.I. Is Getting

2022-08-24 Thread Samiya Illias

…
30:30 So direct your face toward the religion, inclining to truth. [Adhere to] 
the fitrah of Allah upon which He has created [all] people. No change should 
there be in the creation of Allah. That is the correct religion, but most of 
the people do not know. 
…
The Same Religion (الدِّينِ) 
https://signsandscience.blogspot.com/2018/02/the-same-religion.html 


> On 24-Aug-2022, at 10:28 PM, smitra  wrote:
> 
> Thanks for sharing! My comment on this article:
> 
> https://nyti.ms/3dQoxqU#permid=120043436
> 
> "It is inevitable that AI systems will end up becoming good enough to run the 
> economy, repair and reproduce themselves. Biology shows us that this does not 
> require highly intelligent systems. As things stand now,  even insects 
> outperform our best AI systems, but then we may not even need insect-level 
> intelligence to fully automatize our economy.
> 
> This development is then driven by economic growth, it's not something that's 
> easy to regulate. Companies will use whatever technology is available to 
> reduce costs and to get to higher profits. The current climate crisis shows 
> just how hard it is to regulate the rather simple process of our use of 
> energy to reduce CO2 emissions.
> 
> When in the future the economy is run by autonomous machines that maintain 
> and copy each other while producing all the stuff we consume, there will 
> exists a new machine biology besides the original biology. It's then 
> inevitable that the machine biology will not be fully compatible with the 
> original biology. Toxic compounds are likely to be produced.
> 
> The problem we'll then face is that we'll have even less power to mitigate 
> such problems than we have now when dealing with our CO2 emissions. It's then 
> likely that the new machine biology will destroy most of the original biology.
> 
> All intelligent life in the universe likely ends in this way. The takeover by 
> machines with insect-level intelligence or less, then explains why the galaxy 
> hasn't already been colonized (the so-called  Fermi Paradox)."
> 
> Saibal
> 
>> On 24-08-2022 14:39, John Clark wrote:
>> Check out this article from The New York Times. Because I'm a
>> subscriber, you can read it through this gift link without a
>> subscription.
>> We Need to Talk About How Good A.I. Is Getting
>> We’re in a golden age of progress in artificial intelligence. It’s
>> time to start taking its potential and risks seriously.
>> https://www.nytimes.com/2022/08/24/technology/ai-technology-progress.html?unlocked_article_code=CEIPuomT1JKd6J17Vw1cRCfTTMQmqxCdw_PIxftm3iWka3DFDmweiPgYCIiG_EPKarskbtp2wzmQRNlGNLggVblq1OhQJUF2UE-ovp6A0twjEhkClLiSDCkwzo6fGvcx6yPrZW20b710ybPitBzZdWLoUKLA1XV2IRI1qJpmaV372SYKlazAReYl3cJsnqt0XuAMTjgFbCCLv_TjGk8-bI3ANkeAn1FwD-JJWjjTnsqe4qYAdWhRClHHRXB44wUs-Y8WeYNXbOukcUlWKIepiq4RC2doMI6iG5YwIoDUnL9gurLMwgeevnYkS2GsPvx_F8Tqd-ALMQ=em-share
>> --
>> You received this message because you are subscribed to the Google
>> Groups "Everything List" group.
>> To unsubscribe from this group and stop receiving emails from it, send
>> an email to everything-list+unsubscr...@googlegroups.com.
>> To view this discussion on the web visit
>> https://groups.google.com/d/msgid/everything-list/CAJPayv21vyY7ZgNzkcVQ1wYr_6dxYq5LyveYMZZrvkig%2B_oV6A%40mail.gmail.com
>> [1].
>> Links:
>> --
>> [1]
>> https://groups.google.com/d/msgid/everything-list/CAJPayv21vyY7ZgNzkcVQ1wYr_6dxYq5LyveYMZZrvkig%2B_oV6A%40mail.gmail.com?utm_medium=email_source=footer
> 
> -- 
> You received this message because you are subscribed to the Google Groups 
> "Everything List" group.
> To unsubscribe from this group and stop receiving emails from it, send an 
> email to everything-list+unsubscr...@googlegroups.com.
> To view this discussion on the web visit 
> https://groups.google.com/d/msgid/everything-list/468ea47509d9246593730ed39b099f9c%40zonnet.nl.

-- 
You received this message because you are subscribed to the Google Groups 
"Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to everything-list+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/everything-list/596E31DA-4415-46EC-95B7-A03DCF988B93%40gmail.com.

Re: NYTimes.com: We Need to Talk About How Good A.I. Is Getting

2022-08-24 Thread Brent Meeker

The funny thing is that it's not really that AI is reaching high levels
of intelligence; it's that humans aren't nearly as smart as they think
they are. I just read about the Rosenhan Experiment, which I had not
heard of before although I was good friends with a psychologist at the
local state mental asylum at the time.
https://en.wikipedia.org/wiki/Rosenhan_experiment People are easily
misled into seeing what they want to see or what they fear seeing. I
think it's probably easier for an AI to write good propaganda than to
write good poetry.

Brent

On 8/24/2022 5:39 AM, John Clark wrote:
Check out this article from The New York Times. Because I'm a
subscriber, you can read it through this gift link without a subscription.

We Need to Talk About How Good A.I. Is Getting

We’re in a golden age of progress in artificial intelligence. It’s
time to start taking its potential and risks seriously.

--
You received this message because you are subscribed to the Google Groups
"Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to everything-list+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/everything-list/10cf817b-5ddf-5719-f381-2beb204cd27a%40gmail.com.

Re: NYTimes.com: We Need to Talk About How Good A.I. Is Getting

2022-08-24 Thread smitra

Thanks for sharing! My comment on this article:

https://nyti.ms/3dQoxqU#permid=120043436

Saibal

On 24-08-2022 14:39, John Clark wrote:

Check out this article from The New York Times. Because I'm a
subscriber, you can read it through this gift link without a
subscription.

We Need to Talk About How Good A.I. Is Getting

We’re in a golden age of progress in artificial intelligence. It’s
time to start taking its potential and risks seriously.

Links:
--
[1]
https://groups.google.com/d/msgid/everything-list/CAJPayv21vyY7ZgNzkcVQ1wYr_6dxYq5LyveYMZZrvkig%2B_oV6A%40mail.gmail.com?utm_medium=email_source=footer

--
You received this message because you are subscribed to the Google Groups
"Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to everything-list+unsubscr...@googlegroups.com.
To view this discussion on the web visit
https://groups.google.com/d/msgid/everything-list/468ea47509d9246593730ed39b099f9c%40zonnet.nl.

NYTimes.com: We Need to Talk About How Good A.I. Is Getting

2022-08-24 Thread John Clark

Check out this article from The New York Times. Because I'm a subscriber,
you can read it through this gift link without a subscription.

We Need to Talk About How Good A.I. Is Getting

We’re in a golden age of progress in artificial intelligence. It’s time to
start taking its potential and risks seriously.

https://www.nytimes.com/2022/08/24/technology/ai-technology-progress.html?unlocked_article_code=CEIPuomT1JKd6J17Vw1cRCfTTMQmqxCdw_PIxftm3iWka3DFDmweiPgYCIiG_EPKarskbtp2wzmQRNlGNLggVblq1OhQJUF2UE-ovp6A0twjEhkClLiSDCkwzo6fGvcx6yPrZW20b710ybPitBzZdWLoUKLA1XV2IRI1qJpmaV372SYKlazAReYl3cJsnqt0XuAMTjgFbCCLv_TjGk8-bI3ANkeAn1FwD-JJWjjTnsqe4qYAdWhRClHHRXB44wUs-Y8WeYNXbOukcUlWKIepiq4RC2doMI6iG5YwIoDUnL9gurLMwgeevnYkS2GsPvx_F8Tqd-ALMQ=em-share

-- 
You received this message because you are subscribed to the Google Groups 
"Everything List" group.
To unsubscribe from this group and stop receiving emails from it, send an email 
to everything-list+unsubscr...@googlegroups.com.
To view this discussion on the web visit 
https://groups.google.com/d/msgid/everything-list/CAJPayv21vyY7ZgNzkcVQ1wYr_6dxYq5LyveYMZZrvkig%2B_oV6A%40mail.gmail.com.

Re: NYTimes.com: We Need to Talk About How Good A.I. Is Getting

Re: NYTimes.com: We Need to Talk About How Good A.I. Is Getting

Re: NYTimes.com: We Need to Talk About How Good A.I. Is Getting

Re: NYTimes.com: We Need to Talk About How Good A.I. Is Getting

Re: NYTimes.com: We Need to Talk About How Good A.I. Is Getting

Re: NYTimes.com: We Need to Talk About How Good A.I. Is Getting

Re: NYTimes.com: We Need to Talk About How Good A.I. Is Getting

Re: NYTimes.com: We Need to Talk About How Good A.I. Is Getting

NYTimes.com: We Need to Talk About How Good A.I. Is Getting

9 matches

Site Navigation

Mail list logo

Footer information