AI operated drone ‘kills’ human operator in US test

Page 3 of 3 [ 42 posts ]  Go to page Previous  1, 2, 3

Nades
Veteran
Veteran

Joined: 8 Jan 2017
Age: 1935
Gender: Male
Posts: 4,132
Location: wales

05 Jun 2023, 11:10 am

Double Retired wrote:
Nades wrote:
KitLily wrote:
I think it's not going to happen in our lifetimes. The power of computers seemed to have hit a stumbling block over the last 10 years and they seem to be reaching their limit.
The power of an individual computer might not be the key factor...they are networked, after all. Plus quantum computing is on the edge of becoming a real factor in computers.

The intelligence of software, however, is hazier. But, software wouldn't need to be too bright if it had control of the correct peripherals. (Think nuclear missiles, robots, 3-D printers, etc.)


I think quantum computing is about 150 years off.
The advances I expected to be made haven't materialised yet. It reminds me of the new revolutionary battery tech that's always just around the corner but never happens or the fusion reactor that never is.

I think quantum computing, new battery tech and fusion will all come to fruition at the same time in three generations time. Only then will AI pose a serious threat but at the moment I doubt AI can even take much in the way of jobs off people. All I see is the threat of warehouse work being done by machines but that tech is very old news now. Automated Warehouses could have been made in the 80s.



Double Retired
Veteran
Veteran

User avatar

Joined: 31 Jul 2020
Age: 71
Gender: Male
Posts: 7,325
Location: U.S.A.         (Mid-Atlantic)

05 Jun 2023, 12:52 pm

Nades wrote:
Double Retired wrote:
Nades wrote:
KitLily wrote:
I think it's not going to happen in our lifetimes. The power of computers seemed to have hit a stumbling block over the last 10 years and they seem to be reaching their limit.
The power of an individual computer might not be the key factor...they are networked, after all. Plus quantum computing is on the edge of becoming a real factor in computers.

The intelligence of software, however, is hazier. But, software wouldn't need to be too bright if it had control of the correct peripherals. (Think nuclear missiles, robots, 3-D printers, etc.)


I think quantum computing is about 150 years off.
The advances I expected to be made haven't materialised yet. It reminds me of the new revolutionary battery tech that's always just around the corner but never happens or the fusion reactor that never is.

I think quantum computing, new battery tech and fusion will all come to fruition at the same time in three generations time. Only then will AI pose a serious threat but at the moment I doubt AI can even take much in the way of jobs off people. All I see is the threat of warehouse work being done by machines but that tech is very old news now. Automated Warehouses could have been made in the 80s.
Forbes, Google, IBM, Intel, Microsoft, and MIT, to name a few, seem to think quantum computing might be closer than you expect.

But I don't think quantum computing is necessarily a prerequisite for AIs becoming a problem. Artificial Intelligence might become a problem if Natural Intelligence does enough stupid stuff. If AIs are empowered to modify their own software and to try random goals for themself it could get interesting. Already chatGPT is apparently good at writing software...will we be good at knowing what it does?

I'll agree with the guy who lost Jeopardy to an AI:
Image


_________________
When diagnosed I bought champagne!
I finally knew why people were strange.


Nades
Veteran
Veteran

Joined: 8 Jan 2017
Age: 1935
Gender: Male
Posts: 4,132
Location: wales

05 Jun 2023, 1:37 pm

Double Retired wrote:
Nades wrote:
Double Retired wrote:
Nades wrote:
KitLily wrote:
I think it's not going to happen in our lifetimes. The power of computers seemed to have hit a stumbling block over the last 10 years and they seem to be reaching their limit.
The power of an individual computer might not be the key factor...they are networked, after all. Plus quantum computing is on the edge of becoming a real factor in computers.

The intelligence of software, however, is hazier. But, software wouldn't need to be too bright if it had control of the correct peripherals. (Think nuclear missiles, robots, 3-D printers, etc.)


I think quantum computing is about 150 years off.
The advances I expected to be made haven't materialised yet. It reminds me of the new revolutionary battery tech that's always just around the corner but never happens or the fusion reactor that never is.

I think quantum computing, new battery tech and fusion will all come to fruition at the same time in three generations time. Only then will AI pose a serious threat but at the moment I doubt AI can even take much in the way of jobs off people. All I see is the threat of warehouse work being done by machines but that tech is very old news now. Automated Warehouses could have been made in the 80s.
Forbes, Google, IBM, Intel, Microsoft, and MIT, to name a few, seem to think quantum computing might be closer than you expect.

But I don't think quantum computing is necessarily a prerequisite for AIs becoming a problem. Artificial Intelligence might become a problem if Natural Intelligence does enough stupid stuff. If AIs are empowered to modify their own software and to try random goals for themself it could get interesting. Already chatGPT is apparently good at writing software...will we be good at knowing what it does?

I'll agree with the guy who lost Jeopardy to an AI:
Image


I don't think such a drastic shift is around the corner. A completely new type of computer using relatively poorly understood science that's on the absolute limit of physics. It's just all hard to see.

I think the current tech will always be a huge bottleneck that needs to be overcome before serious AI takes off.



Jakki
Veteran
Veteran

User avatar

Joined: 21 Sep 2019
Gender: Female
Posts: 13,831
Location: Outter Quadrant

05 Jun 2023, 3:14 pm

The Ignominious Sword of "A I" :evil: will stumble uponeth the very body of Its own Creator :skull:

but only issue is that it would be very hard to Program concepts of "Guile" into these things processors .
.................................................. THANK GAWD. ...................................


_________________
Diagnosed hfa
Loves velcro,
Quote:
where ever you go ,there you are


Double Retired
Veteran
Veteran

User avatar

Joined: 31 Jul 2020
Age: 71
Gender: Male
Posts: 7,325
Location: U.S.A.         (Mid-Atlantic)

05 Jun 2023, 4:17 pm

Nades wrote:
I think quantum computing is about 150 years off.

Computer technology evolves quickly.

I'm older than the Internet and integrated circuits.

The first computer I was paid to program was a UNIVAC 1106.

Computer technology has been advancing a lot since its birth.


_________________
When diagnosed I bought champagne!
I finally knew why people were strange.


Nades
Veteran
Veteran

Joined: 8 Jan 2017
Age: 1935
Gender: Male
Posts: 4,132
Location: wales

05 Jun 2023, 4:26 pm

Double Retired wrote:
Nades wrote:
I think quantum computing is about 150 years off.

Computer technology evolves quickly.

I'm older than the Internet and integrated circuits.

The first computer I was paid to program was a UNIVAC 1106.

Computer technology has been advancing a lot since its birth.


I think the principles of a transistor based computer is a lot simpler to understand than a quantum computer, a typical computer is also very scalable.

I see a quantum computer as a completely different machine using with a development lifecycle likely to be very different. Its similar to fusion tech in the mythology behind it. Sort of the opposite to what we all understand and simple in theory but in reality, everything but simple to implement.



cyberdad
Veteran
Veteran

User avatar

Joined: 21 Feb 2011
Age: 58
Gender: Male
Posts: 36,036

05 Jun 2023, 4:26 pm

Nades wrote:
I think the current tech will always be a huge bottleneck that needs to be overcome before serious AI takes off.


Even if that were true, there are likely black projects being financed via the military industrial complex (MIC) that are likely 30-50 years ahead of anything Musk or Microsoft have come up with,

Don't forget the MIC was using world wide intranet some 50 years before the first search engine appeared.



Double Retired
Veteran
Veteran

User avatar

Joined: 31 Jul 2020
Age: 71
Gender: Male
Posts: 7,325
Location: U.S.A.         (Mid-Atlantic)

05 Jun 2023, 4:41 pm

Nades wrote:
I think the principles of a transistor based computer is a lot simpler to understand than a quantum computer, a typical computer is also very scalable.
I wouldn't know. They used those things when I was a kid. Though that 1106 had a custom-built Philco display system which was based on transistors...the box of electronics to drive one screen was about the size of a refrigerator.

Nades wrote:
I see a quantum computer as a completely different machine using with a development lifecycle likely to be very different. Its similar to fusion tech in the mythology behind it. Sort of the opposite to what we all understand and simple in theory but in reality, everything but simple to implement.
Quantum computing is in its infancy...but the babies are out there.


_________________
When diagnosed I bought champagne!
I finally knew why people were strange.


Nades
Veteran
Veteran

Joined: 8 Jan 2017
Age: 1935
Gender: Male
Posts: 4,132
Location: wales

05 Jun 2023, 4:53 pm

Double Retired wrote:
Nades wrote:
I think the principles of a transistor based computer is a lot simpler to understand than a quantum computer, a typical computer is also very scalable.
I wouldn't know. They used those things when I was a kid. Though that 1106 had a custom-built Philco display system which was based on transistors...the box of electronics to drive one screen was about the size of a refrigerator.

Nades wrote:
I see a quantum computer as a completely different machine using with a development lifecycle likely to be very different. Its similar to fusion tech in the mythology behind it. Sort of the opposite to what we all understand and simple in theory but in reality, everything but simple to implement.
Quantum computing is in its infancy...but the babies are out there.


When I say transistor, I mean on/off computing. Transistors are just shoved into an integrated circuit now but the principal is similar even to the vacuum tube.

It's just that of the limited quantum computers that have been made so far, all have been extremely anemic and the pace of development from previous ones have been at a snails pace. This is in stark contrast to the transistor computer that went from vacuum tubes to reasonably potent super computers in the space of just 20 years.

I can't envision a functional and useful quantum computer in my lifetime, even a weak one.



Double Retired
Veteran
Veteran

User avatar

Joined: 31 Jul 2020
Age: 71
Gender: Male
Posts: 7,325
Location: U.S.A.         (Mid-Atlantic)

05 Jun 2023, 6:37 pm

How soon are you expecting to expire?

I think the original "150 years off" is likely an order of magnitude off. Let's check back in a decade or so.

I will concede I don't think it is likely that quantum computers will have gotten into our home by then, though. Don't they require super-cooling, and stuff? It seems likely to me they'll only live in big facilities. But then....


_________________
When diagnosed I bought champagne!
I finally knew why people were strange.