robotics : dreaming away on the job?

[peacefan]

• give robots 2 visual cortexes and 2 main CPUs
• the 'real work' (in a warehouse for instance) is not what the robot will focus on (it'll focus on input-output on that 2nd CPU and visual cortex), UNLESS that real-work CPU reports problems.

now why would you want to do this?

• to be nice to the robots, who WILL feel the same about boring repetitive work as human workers would.
• by being nice to robots, we can delay their call for voting rights (dogs don't get to vote either after all).

 

[peacefan]

"but robots will be so much smarter than humans, they'd demand (on pain of worker strikes) to be able to get elected into the highest office."

then just don't build them with IQs&EQs of over 100.

 

[scruffy]

• give robots 2 visual cortexes and 2 main CPUs
• the 'real work' (in a warehouse for instance) is not what the robot will focus on (it'll focus on input-output on that 2nd CPU and visual cortex), UNLESS that real-work CPU reports problems.

now why would you want to do this?

• to be nice to the robots, who WILL feel the same about boring repetitive work as human workers would.
• by being nice to robots, we can delay their call for voting rights (dogs don't get to vote either after all).

"Feelings" require viscera.

Robots dont have any.

Not yet, anyway.

 

[peacefan]

"Feelings" require viscera.

Robots dont have any.

Not yet, anyway.

you don't need a gut-instinct or organs other than a brain to experience emotions.
visual and audio input is plenty to achieve a sub-conciousness that experiences emotions based on what it is fed.

 

[Sherlock Holmes]

you don't need a gut-instinct or organs other than a brain to experience emotions.
visual and audio input is plenty to achieve a sub-conciousness that experiences emotions based on what it is fed.

Computers, I mean digital computers, aka "symbol manipulators" have no capacity for consciousness.

 

[peacefan]

Computers, I mean digital computers, aka "symbol manipulators" have no capacity for consciousness.

so you haven't seen any sci-fi movies at all, have you?

 

[Sherlock Holmes]

so you haven't seen any sci-fi movies at all, have you?

More than I can count, but I thought we were discussing fact not fiction.

 

[peacefan]

More than I can count, but I thought we were discussing fact not fiction.

fiction tends to become a goal to achieve for some.
first Star Trek had digital pads, now we have iPads.
for instance.

 

[Sherlock Holmes]

fiction tends to become a goal to achieve for some.
first Star Trek had digital pads, now we have iPads.
for instance.

Yes but that has no bearing on the fact that symbol manipulators (aka digital computers) have no capacity for consciousness. Here's a good book I recommend, that talks about this:

 

[scruffy]

you don't need a gut-instinct or organs other than a brain to experience emotions.

Yes, you do.

There is overwhelming evidence in that regard.

visual and audio input is plenty to achieve a sub-conciousness that experiences emotions based on what it is fed.

Not sufficient.

I suggest Neuroscience 101.

Qualia are not equivalent to feelings.

 

[scruffy]

Computers, I mean digital computers, aka "symbol manipulators" have no capacity for consciousness.

Not yet.

Pretty soon, though.

 

[scruffy]

Yes but that has no bearing on the fact that symbol manipulators (aka digital computers) have no capacity for consciousness. Here's a good book I recommend, that talks about this:

View attachment 992756

More bullshit by philosophers.

Look here - study psychopathy.

There is a huge pathway from the amygdala to the ventromedial prefrontal cortex that's broken in psychopaths.

"Feelings" require a functioning amygdala. Which in turn depends on input from the hypothalamus, which is "viscera central" in humans and all other conscious life forms.

 

[Sherlock Holmes]

More bullshit by philosophers.

Look here - study psychopathy.

There is a huge pathway from the amygdala to the ventromedial prefrontal cortex that's broken in psychopaths.

"Feelings" require a functioning amygdala. Which in turn depends on input from the hypothalamus, which is "viscera central" in humans and all other conscious life forms.

Digital computers have no hope of attaining consciousness, nothing "bullshit" here I promise you. The human "mind" is not algorithmic, it is able to solve non-algorithmic (aka "non computable") problems, see Penrose:

Digital computers can only compute computable functions, they cannot compute non-computable functions. You need to understand this in order to appreciate what I'm saying. The mind has an ability for "insight" able to make inferences with intuition rather than computation, it's a deep mystery but it is what it is.

 

[scruffy]

Digital computers have no hope of attaining consciousness, nothing "bullshit" here I promise you. The human "mind" is not algorithmic, it is able to solve non-algorithmic (aka "non computable") problems, see Penrose:

View attachment 993285

Digital computers can only compute computable functions, they cannot compute non-computable functions. You need to understand this in order to appreciate what I'm saying. The mind has an ability for "insight" able to make inferences with intuition rather than computation, it's a deep mystery but it is what it is.

Quantum computers are not digital.

I have fully explained the mystery of awareness in another thread.

It has to do with topology, not computation.

 

[scruffy]

Digital computers have no hope of attaining consciousness, nothing "bullshit" here I promise you. The human "mind" is not algorithmic, it is able to solve non-algorithmic (aka "non computable") problems, see Penrose:

View attachment 993285

Digital computers can only compute computable functions, they cannot compute non-computable functions. You need to understand this in order to appreciate what I'm saying. The mind has an ability for "insight" able to make inferences with intuition rather than computation, it's a deep mystery but it is what it is.

The Penrose-Hammerof theory of consciousness has been disproven.

However, there is plenty of evidence for quantum tunneling in microtubules.

Inference is no mystery at all. It has been occurring in artificial neural networks for at least 50 years.

Google "Bayesian inference".

Not only that, but Bayes is old news.

Study non- Bayesian inference.

Google "judea pearl causality".

 

[scruffy]

Digital computers have no hope of attaining consciousness, nothing "bullshit" here I promise you. The human "mind" is not algorithmic, it is able to solve non-algorithmic (aka "non computable") problems, see Penrose:

View attachment 993285

Digital computers can only compute computable functions, they cannot compute non-computable functions. You need to understand this in order to appreciate what I'm saying. The mind has an ability for "insight" able to make inferences with intuition rather than computation, it's a deep mystery but it is what it is.

As far as computability goes, Hopfield networks and their cousins Boltzmann machines are able to solve np-hard problems in milliseconds.

An artificial neural network can solve a 50+ city traveling salesman problem in less than 1 msec. In other words "much faster than a human being".

The math is "stochastic optimization". It's not easy.

For AI technology, Google "multi headed transformers" and "arnold-kolmogorov networks".

 

[fncceo]

visual and audio input is plenty to achieve a sub-conciousness that experiences emotions

I’ll believe that the day my media center kills itself because it was being online mocked by a Sony

 

[peacefan]

Digital computers have no hope of attaining consciousness, nothing "bullshit" here I promise you. The human "mind" is not algorithmic, it is able to solve non-algorithmic (aka "non computable") problems, see Penrose:

View attachment 993285

Digital computers can only compute computable functions, they cannot compute non-computable functions. You need to understand this in order to appreciate what I'm saying. The mind has an ability for "insight" able to make inferences with intuition rather than computation, it's a deep mystery but it is what it is.

and predicting the future is only possible for futurists (Iron Man) and the Kwizat Hiderrach (Dune), and even then : very often inaccurate and next to impossible.

 

[Sherlock Holmes]

As far as computability goes, Hopfield networks and their cousins Boltzmann machines are able to solve np-hard problems in milliseconds.

An artificial neural network can solve a 50+ city traveling salesman problem in less than 1 msec. In other words "much faster than a human being".

The math is "stochastic optimization". It's not easy.

For AI technology, Google "multi headed transformers" and "arnold-kolmogorov networks".

Anything that is implemented, realized in software is by definition computable. If the human mind can perform non-computable processes then that operation could never be simulated in a digital computer.

My position has been for some time, that consciousness is not an emergent quality in nature, but is fundamental to it. Consciousness gives rise to the world not the other way around.

Penrose speaks at length about Gödel's incompleteness theorem and how the act of devising that theorem cannot itself be reduced to an algorithm, computable yet a human mind did and can do such things.

Ultimately all of the machines you cite, and including Turing machines and simple FSMs and so on, these are all variants of general state machines. The systems are deterministic.

 

[scruffy]

Anything that is implemented, realized in software is by definition computable. If the human mind can perform non-computable processes then that operation could never be simulated in a digital computer.

Again: neural networks are not digital.

My position has been for some time, that consciousness is not an emergent quality in nature, but is fundamental to it. Consciousness gives rise to the world not the other way around.

Otay... If you say so...

Penrose speaks at length about Gödel's incompleteness theorem and how the act of devising that theorem cannot itself be reduced to an algorithm, computable yet a human mind did and can do such things.

A neural network can do it too.

AI has already solved problems that vexxed mathematicians for centuries.

Ultimately all of the machines you cite, and including Turing machines and simple FSMs and so on, these are all variants of general state machines. The systems are deterministic.

You are ignorant.

I told you already, it's STOCHASTIC optimization.

By definition, that is the exact opposite of deterministic.

I just spent half a thread showing you why asynchronicity is important. Maybe you should read it.