Politics & Stuff

[robsblubot 3,237]

The discussions around AI are IMO extremely premature. The problems discussed today are several decades in the future (maybe 100 years+?).

The problems AI brings today, as in the immediate future, are the very same any automation produces: cheap (and not so cheap soon) jobs vanish at an alarming rate. And the liars on the right here blame immigration for that. The left ignores the problem, because it's a difficult one to tackle (esp politically).

"Tony Stark : I'm gonna stop you right there, Scott. Are you seriously telling me that your plan to save the universe is based on "Back to the Future?"

 Who'd have thought Exterminator franchise causing long-term damage 😄

Go back and read the predictions on AI and Robotics from very smart sci-fi authors (I've read a bunch growing up), some were physicists too!, and how much they got wrong (almost all of it). 

Edited April 22 by robsblubot

[Vesper 27,345]

12 hours ago, robsblubot said:

The problems discussed today are several decades in the future (maybe 100 years+?).

you seemingly do not understand a basic exponential growth curve then 

 or  (where the argument x is written as an exponent)

 

nor apparently the concept of generativeness

[cosmicway 1,260]

13 hours ago, lucio said:

people will rise up and kill the computer nerds who make this AI cringe 

Why ?
When you go to the bank and it's 5 past 3 and the doors automatically close do you break them ?
The robots will not turn hostile (maybe at some final stage they will).
Also they will have people as frontmen.

[Vesper 27,345]

12 hours ago, lucio said:

people will rise up and kill the computer nerds who make this AI cringe 

the genie is already out of the bottle m8

there are obviously multiple points in history where you can say the threat started

but to narrow it down I would draw a line under November 30, 2022

that is when ChatGPT, the first remotely truly powerful (and yet embryonic compared to what is to come) chatbot was released into the open, interconnective ecosphere

humankind is already being so affected and manipulated by AI, but again, it is nothing comapred to what is to come

[Vesper 27,345]

let me re-post this basic intro on some of the dangers of generative AI

 

Mo Gawdat is the Former Chief Business Officer of Google X Development

X Development LLC, doing business as X (formerly Google X), is an American semi-secret research and development facility and organization founded by Google in January 2010.

[cosmicway 1,260]

The jobs that vanish is an old story.
The neighbourhood typist is one.
Last time I visited one was in 1982, a little before the home line printer. My father gave me some documents he wanted typed.
Funny thing that happened that day was this:
The typist had a stack of papers on the next table. While she was typing I said "you need those ? I want them as scrap paper". She says "ok - take them".
Then I read the back side and it was a list of shops of the neighbourhood as compiled in the local PASOK party headquarters. In it the various shopkeepers were described as "reactionaries", "exponents of the right" and such names. The PASOK party was fresh in government, in its pre-USA days.

 

[cosmicway 1,260]

You can't say that AI is per se hostile.
But it may be superior.
If you are against AI you are also against modern football systems and you want WM to come back.

[Fulham Broadway 16,410]

Never mind about Artificial Intelligence, some real intelligence would be good

[Vesper 27,345]

to put it simply, very simply

at this moment AI, through social media and other architectures (both electronic and human) is ramping up FUD in humankind

FUD = Fear, Uncertainty, Doubt

Fear leads to anger. Anger leads to hate. Hate leads to suffering.

 

[cosmicway 1,260]

Look suppose I give you a horse race card.
You need to spend one-two hours to work out the form.
But AI can do it at the push of a button.
Of course if you go to the races and hang around there is hanky panky business going on you can sniff and it's better than form analysis.
But form analysis is a necessary prerequisite and if AI does it for you, you must be pretty archaic not to use.

Edited April 23 by cosmicway

[robsblubot 3,237]

28 minutes ago, Vesper said:

you seemingly do not understand a basic exponential growth curve then 

 or  (where the argument x is written as an exponent)

 

nor apparently the concept of generativeness

I don’t know many things, but these are pretty familiar to me especially exponential growth, given it’s something that pertains my daily work. Many algos have exponential growth relative to the dataset.

generativeness I learned when I was 15 when I studied a stupid language called prolog, which was supposed to be the AI language — ML engines were all written in C despite the promises.

These predictions are BS and that’s an opinion—no facts here. It’s actually pretty fucking ironic that an event so unique and new is so easy to predict (and far into the future) by your friendly YouTuber.

the jobs will exponentially vanish tho. Industrial Revolution on steroids.

[cosmicway 1,260]

5 minutes ago, robsblubot said:

I don’t know many things, but these are pretty familiar to me especially exponential growth, given it’s something that pertains my daily work. Many algos have exponential growth relative to the dataset.

generativeness I learned when I was 15 when I studied a stupid language called prolog, which was supposed to be the AI language — ML engines were all written in C despite the promises.

These predictions are BS and that’s an opinion—no facts here. It’s actually pretty fucking ironic that an event so unique and new is so easy to predict (and far into the future) by your friendly YouTuber.

the jobs will exponentially vanish tho. Industrial Revolution on steroids.

What he means is the rate at which civilization is progressing.
Suppose year 1924 to year 2024 counts as one unit.
Then 1824 to 1924 how many units ? 
Let's say one unit again.
But as we go back the units per one hundred years definitely get smaller.
So the Romans of 300 AD were really better than the Macedonians of 200 BC ? I doubt it.
The earliest finds of human activity are 3 million years old and it looks as if there is some exponential growth.

Edited April 23 by cosmicway

[Vesper 27,345]

Just now, robsblubot said:

I don’t know many things, but these are pretty familiar to me especially exponential growth, given it’s something that pertains my daily work. Many algos have exponential growth relative to the dataset.

generativeness I learned when I was 15 when I studied a stupid language called prolog, which was supposed to be the AI language — ML engines were all written in C despite the promises.

These predictions are BS and that’s an opinion—no facts here. It’s actually pretty fucking ironic that an event so unique and new is so easy to predict (and far into the future) by your friendly YouTuber.

the jobs will exponentially vanish tho. Industrial Revolution on steroids.

to blithely dismiss one of the fairly central architects in re AI developmental roll-out as 'your freindly YouTuber' says far more about you than me

but let's take it up a notch in terms of centrality to AI as a whole and see why Geoffrey Hinton is worried about the future of AI

Geoffrey Hinton, known to many as the “Godfather of AI,” recently made headlines around the world after leaving his job at Google to speak more freely about the risks posed by unchecked development of artificial intelligence, including popular tools like ChatGPT and Google’s PaLM. Why does he believe digital intelligence could hold an advantage over biological intelligence? How did he suddenly arrive at this conclusion after a lifetime of work in the field? Most importantly, what – if anything – can be done to safeguard the future of humanity? The University of Toronto University Professor Emeritus addresses these questions and more in The Godfather in Conversation.

00:00 Intro
01:03 Digital intelligence
02:27 Biological intelligence
03:47 Why worry?
04:39 Machine learning
07:07 Neural Nets
13:22 Neural nets and language
17:18 Challenges
18:49 Breakthrough moment
20:41 AlexNet
24:35 Pace of Innovation
26:04 ChatGPT
27:46 Public Reaction
29:49 Benefits for society
33:25 Pace of innovation
35:48 Sudden realization
37:13 Role of government
40:08 Big tech
42:32 Advice to researchers
43:50 Understanding risk
45:20 What’s next?

 

 

[cosmicway 1,260]

One yardstick for AI could be the Putnam exam (US maths Olympiad).
I don't have any pure calculus books here - left them in Egnland - but I can barely solve one problem, usually none.
If AI can do the Putnam paper then it wins over human intelligence.
The rest is trivial - just attach some limbs to it so it can walk, use screwdrivers etc.

Edited April 23 by cosmicway

[Vesper 27,345]

6 minutes ago, cosmicway said:

What he means is the rate at which civilization is progressing.

She (if you were referring to me) and I meant exponential rate of growth in the computational and learning power of generative AI, not human civilisation.

The pace of AI development is exponential, with performance doubling every few months. This trend, known as Moore’s Law, has held steady for AI just as it has for computer processing power. Each new generation of AI algorithms is smarter and more capable than the last.

Moore's law

Moore's law is the observation that the number of transistors in an integrated circuit doubles about every two years. Moore's law is an observation and projection of a historical trend. Rather than a law of physics, it is an empirical relationship linked to gains from experience in production.

[robsblubot 3,237]

10 minutes ago, Vesper said:

to blithely dismiss one of the fairly central architects in re AI developmental roll-out as 'your freindly YouTuber' says far more about you than me

but let's take it up a notch in terms of centrality to AI as a whole and see why Geoffrey Hinton is worried about the future of AI

Geoffrey Hinton, known to many as the “Godfather of AI,” recently made headlines around the world after leaving his job at Google to speak more freely about the risks posed by unchecked development of artificial intelligence, including popular tools like ChatGPT and Google’s PaLM. Why does he believe digital intelligence could hold an advantage over biological intelligence? How did he suddenly arrive at this conclusion after a lifetime of work in the field? Most importantly, what – if anything – can be done to safeguard the future of humanity? The University of Toronto University Professor Emeritus addresses these questions and more in The Godfather in Conversation.

00:00 Intro
01:03 Digital intelligence
02:27 Biological intelligence
03:47 Why worry?
04:39 Machine learning
07:07 Neural Nets
13:22 Neural nets and language
17:18 Challenges
18:49 Breakthrough moment
20:41 AlexNet
24:35 Pace of Innovation
26:04 ChatGPT
27:46 Public Reaction
29:49 Benefits for society
33:25 Pace of innovation
35:48 Sudden realization
37:13 Role of government
40:08 Big tech
42:32 Advice to researchers
43:50 Understanding risk
45:20 What’s next?

 

 

I will check it out and hope he mentions the “unknown unknowns”, which is what I find severely missing in all these predictions.

I will remain very skeptical on this.

Although I have to admit that it certainly seems more fun than to discuss where the next nuclear bomb will detonate.

[Vesper 27,345]

3 main things worry me to core atm in terms of existential threats to the future of humankind

in no order (and at some levels all are interconnected)

Nuclear Weapons

AI

Global Climate Change

 

at a more immediate level:

the possible re-election of Trump and then the Republicans retaining control the US House of Representatives and gaining back the US Senate (wherein a christo-fascistic minority would hold the whip hand over all 3 branches of the most powerful nation in human history's 3 branches of its federal government (Executive, Judicial (the US Supreme Court especially), and Legislative)

[Vesper 27,345]

6 minutes ago, robsblubot said:

it certainly seems more fun than to discuss where the next nuclear bomb will detonate

the main scenarios (and none are imminet atm)

in no order

Israel on Iran

India v Pakistan in re their border regions

Russian tactical use in Ukraine if Putin thinks that he both can get away with it AND that it is his only main functional option left to fully winning it back

North Korea if they go full death cult and Kim Jong Un thinks he is going to be taken out (targets would likely be SoKo, Japan, and (100 per cent suicidally) the US, I wager)

Iran (whether directly or through the use of proxies) IF they actually can get them made and they feel that they are going to be flattened (the Mullahs are NOT suicidal, so they would have to be 100 per cent convinced they are about to be taken out)

China tactical use on Taiwan if they decide to take it back and feel that (like Putin above) they can both get away with it AND that it is their last best option for a quick conclusion to the takeover

Sunni rando terror nuke(s) (likely sourced from renegade elements in Pakistan) The targets could be anywhere (US, Europe, Middle East, Afghanistan, India, Russia, etc etc, even Africa)

Samson Option (launch all 400+ thermonukes globally) by Israel if they think they are going down (least likely of all of these)

 

wildcard

the UK finally settles the French question of the last 1000 years (I jest!)