Found in Translation

Found in Translation

Found in Translation

Whereas perfect algorithms (or working rules) were available for the performance of the elementary arithmetical and logical operations, … no such algorithms were in existence for translation
Moreover, whereas the notion of a “correct” computation is unproblematic … the notion of a “good” translation is ridden with problems

Bar-Hillel 1962

Julia and John are talking about the following article:
Machine translation

In this post the dialogue is realised by an interaction of virtual characters, for more information please check the page “Virtual characters

Julia

I might have a good analogy!                                                                   

Today 17:08

John

Ok, let’s hear it then … 😊

Today 17:09

Julia

Perhaps we can compare it to machine translation 😊                      

Today 17:09

John

Interesting, in what way?

Today 17:10

Julia

I read somewhere that the classical approach to automating translation was to analyze every sentence according to its grammatical structure “Rule-based machine translation

Today 17:11

Julia

Find the subject, object, verb, and everything else, with their gender, number, aspect, tense, etc.  

Today 17:11

Julia

and then the translation program replaces them with their equivalent in the target language, keeping their relations the same.

Today 17:11

John

And we can compare that to the explicit programming of the laws of thought approach, right?

Today 17:12 

Julia

Indeed! But the alternative is to simply use an enormous amount of data, of translated samples, to generalize and learn through statistics

Today 17:12

John

Which is like the implicit approach? Where the rules don’t matter as long as it comes out right?

Today 17:13 

Julia

Yes, broadly. The analogy is not perfect, but I think it might help.

Today 17:14

John

I guess most people are familiar with automatic translation nowadays

Today 17:15 

Julia

That’s also why I picked this particular example                                   

Today 17:15

John

But which approach do the popular free on-line tools use?

Today 17:16 

Julia

To the best of my knowledge, they all use the statistical approach

Today 17:16

John

Very good, then we have an example that everyone can check out

Today 17:17   

Julia

This has been very interesting and instructive, thanks again for the opportunity of collaborating on this!

Today 17:18

John

Oh, well thanks to you for doing your share and more!

Today 17:18   

Julia

I look forward to doing this again sometime                                         

Today 17:19

John

Absolutely, I’ll keep you posted!

Today 17:19   

… Continue reading our conversations that are posted every Monday …

1 Star2 Stars3 Stars4 Stars5 Stars (4 votes, average: 4.25 out of 5)
Loading...

Related post

Sketchy AI

Sketchy AI

“Young man, you don’t think I’m encountering constructive symbolism for the first time, do you? “Unplug”: free yourself from physicality, from the vulgar earthbound physicality, from the evils of civilization, from electricity, nerves, and so on. Isn’t that what this is? The association is crystal clear, I’d be tempted to say, perhaps even too obvious” (Kishon 1987)

read more
AI see, AI do

AI see, AI do

You young badgers imagine that it would be enough to smear a pound of paint on the canvas and you’re already a modern painter. I would be willing to accept this abomination for a bad joke, if there was even the slightest humor behind it. But this here is an abracadabra! A nothing!

read more

The ends justify the rules

The ends justify the rules

The ends justify the rules

Complicated mental processes are entirely reducible to such simple activities as the attentive observation of statements previously accepted as true, the perception of structural, purely external, connections among these statements, and the execution of mechanical transformations as prescribed by the rules of inference

Tarski 1936

Julia and John are talking about the following article:
The laws of thought and thinking machines

In this post the dialogue is realised by an interaction of virtual characters, for more information please check the page “Virtual characters

Julia

Hi John, I just finished discussing things with Sam                               

Today 17:08

John

Great! So what did he have to say about this whole “laws of thought” thing? 😊

Today 17:09

Julia

Well, there’s a bunch of different approaches  😊                               

Today 17:09

Julia

Either you try programming them in from the start, or you make the AI learn them on its own                                  

Today 17:09

John

Would either of those then yield the same “rules” humans use?

Today 17:10

Julia

Probably not, there’s still far too many simplifications and shortcuts in the models

Today 17:11

Julia

Generally people care more about the results than the procedure  

Today 17:11

John

So as long as it works, it isn’t relevant whether or not an AI uses the same process a human would use

Today 17:12 

Julia

Exactly, moreover we don’t really know what rules humans use      

Today 17:12

Julia

basically, that computers cannot understand the world like humans do

Today 17:12

John

We don’t know how we get stuff done?

Today 17:13 

Julia

Perhaps I should also go talk to a psychologist or a logician, but: no, not really

Today 17:14

Julia

We can’t properly generalize how we solve problems to how any problem can be solved in general

Today 17:14

Julia

and we don’t know precisely enough how the rules are “programmed” in the brain

Today 17:14

John

OK then, so AI just is a tool we use to solve problems, whether or not in a human way

Today 17:15 

John

and the whole “laws of thought” thing then doesn’t really matter, right?

Today 17:15 

Julia

I think we might have to split out how AI is used in different context.

Today 17:16

Julia

The AI in an automated assembly line certainly doesn’t need to think and behave like a human

Today 17:16

Julia

but research in computer science and in psychology does have to ambition to discover these laws

Today 17:16

John

So mostly we’d need to split up theory and application

Today 17:17 

Julia

Perhaps that would be the best. And AI would certainly be on the theoretical side

Today 17:18

John

Very good, the only thing we would still need is a good example or metaphor to make it clear to the reader

Today 17:19   

… Continue reading our conversations that are posted every Saturday …

1 Star2 Stars3 Stars4 Stars5 Stars (4 votes, average: 4.25 out of 5)
Loading...

Related post

Sketchy AI

Sketchy AI

“Young man, you don’t think I’m encountering constructive symbolism for the first time, do you? “Unplug”: free yourself from physicality, from the vulgar earthbound physicality, from the evils of civilization, from electricity, nerves, and so on. Isn’t that what this is? The association is crystal clear, I’d be tempted to say, perhaps even too obvious” (Kishon 1987)

read more
AI see, AI do

AI see, AI do

You young badgers imagine that it would be enough to smear a pound of paint on the canvas and you’re already a modern painter. I would be willing to accept this abomination for a bad joke, if there was even the slightest humor behind it. But this here is an abracadabra! A nothing!

read more

The Digital Analogy

The Digital Analogy

The Digital Analogy

Computers are only prostheses, they no more do calculations than clocks tell the time.
Clocks help us to tell the time, but they don’t do it by themselves..

Tallis 2008

Julia and Sam are talking about the following article:
Analog and digital, continuous and discrete

In this post the dialogue is realised by an interaction of virtual characters, for more information please check the page “Virtual characters

Julia

I’m not sure I understand why the 🌍 digital/analog distinction is relevant here …

Today 17:08

Julia

Someone once tried to convince me the distinction doesn’t even make sense!

Today 17:08

Sam

Ha, well, 😁 computationally it does, but perhaps it is better to speak of 🌍 discrete and continuous

Today 17:09   

Julia

Ok, now I’m even more confused …                                                        

Today 17:09

Sam

Well, think of a analog and digital watches: both tell the same time, but in a different format

Today 17:10

Julia

Sure, so you mean there is no difference?                                             

Today 17:11

Sam

I didn’t say that, but perhaps that they simply measure time in different ways, it is not time which is analog or digital, but the watch

Today 17:11   

Julia

Ah, ok. So in the case of information processing, it is not the brain or the computer that is analog or digital, but how we measure and describe it?

Today 17:12

Sam

That’s exactly what I mean! We can describe a neuron in different ways and then compare it to a circuit or a flipflop.

Today 17:12   

Julia

So there is a similarity in the description, but not necessarily in the thing you are describing? 🤔                                                                    

Today 17:13

Sam

That starts to sound very philosophical and abstract, but, yes, indeed. We can describe them in the same way.

Today 17:14   

Julia

So what did you mean with discrete and continuous?                                                           

Today 17:15

Sam

Both brains and computers work with electricity, but in a computer the voltages are tightly controlled and used in a digital way: ones and zeroes

Today 17:16   

Sam

In a brain the voltage can vary continuously, but sometimes triggers a neuron to “fire”, and up to a point we can treat that as a “one” too

Today 17:16   

Sam

So continuous means that we measure what is happening as close as possible, discrete that we just count the neurons that fire

Today 17:17   

Julia

Aha, so by treating them as discrete units, we can then treat them as digital in our computational model 😊                                                                                

Today 17:18

Sam

Precisely! We’re not trying to copy the brain at all, only what we think is informationally relevant.

Today 17:19   

Sam

But Cho will tell you we are also ignoring a lot of stuff that happens in the brain to make that possible

Today 17:19  

Julia

And that we don’t really know which bits and pieces are relevant for the brain as a whole

Today 17:20

Sam

Yep, I can tell you exactly how information flows through a computer or even a network, but we can’t trace it in the brain

Today 17:21

Julia

Because in the brain medium and message, 🌍 vehicle and content aren’t clearly distinguishable.

Today 17:22

Julia

So we can’t point at the brain and say “this process right here is this or that law of thought”. I think I get it!

Today 17:22

Sam

Happy to have been of assistance!

Today 17:23

Julia

Thank you so much! I’ll report back to John now                                  

Today 17:23

… Continue reading our conversations that are posted every Saturday …

1 Star2 Stars3 Stars4 Stars5 Stars (4 votes, average: 4.25 out of 5)
Loading...

Related post

Sketchy AI

Sketchy AI

“Young man, you don’t think I’m encountering constructive symbolism for the first time, do you? “Unplug”: free yourself from physicality, from the vulgar earthbound physicality, from the evils of civilization, from electricity, nerves, and so on. Isn’t that what this is? The association is crystal clear, I’d be tempted to say, perhaps even too obvious” (Kishon 1987)

read more
AI see, AI do

AI see, AI do

You young badgers imagine that it would be enough to smear a pound of paint on the canvas and you’re already a modern painter. I would be willing to accept this abomination for a bad joke, if there was even the slightest humor behind it. But this here is an abracadabra! A nothing!

read more

Thinking in the Box

Thinking in the Box

Thinking in the Box

Even the simplest brains of the simplest animals are awesome computational instruments. They do computations we do not know how to do, in ways we do not understand

There is nothing that is done in the nervous system that we cannot emulate with electronics if we understand the principles of neural information processing.

Mead 1990

Julia and Sam are talking about the following article:
Neuromorphic Computing

In this post the dialogue is realised by an interaction of virtual characters, for more information please check the page “Virtual characters

Sam

When we talk about computers, we tend to distinguish the machine and its programming: hardware and software

Today 17:09   

Julia

OK, yes, I think everyone is familiar with those terms                         

Today 17:09

Sam

Great, but I’d like to point out how wrong that is, in some respects …

Today 17:10   

Sam

Where do we draw the line between those? Are they really that different? Doesn’t the design of the one determine the design of the other?

Today 17:10   

Julia

Whoa, those are tricky questions! 😅
Remarkably similar to the kind we ask about humans; is that what you’re after?

Today 17:11

Sam

In a sense. We are so used to the hardware/software distinction, that we forget that somewhere they overlap and interact

Today 17:11   

Julia

Can you give me some examples? Otherwise, I fear our readers won’t all get it …😅

Today 17:12

Sam

Sure: flip-flops are at the same time a hardware device and a software abstraction, a circuit and a one or zero

Today 17:12   

Julia

Right, so there is a level where the distinction is arbitrary: it is both a thing and information

Today 17:13

Sam

Exactly! 😉 But then of course at other levels things are much more clear cut

Today 17:14   

Julia

But what did you mean with “wetware” then?                                       

Today 17:15

Sam

Ah, yes, well as far as I know, the brain is like that all over: it is tremendously difficult to keep the hardware and software apart

Today 17:16   

Julia

Oooh, now I get what you mean: a classical computer architecture is completely different

Today 17:16

Sam

Right, but if the brain is an information processor, you can try to just copy the informationally relevant bits

Today 17:17   

Julia

So like Smee you can capture the same informationally relevant relations in a different hardware?

Today 17:18

Sam

Up to a point, of course. We can try to copy some of the structure of the brain, but only in software really

Today 17:19   

Julia

What do you mean? 🤔                                                                             

Today 17:19

Sam

Well, most “artificial neural networks” (ANN) actually are software simulations that run on ordinary beige box computers

Today 17:19   

Julia

Aha, so as you said: a lot of simplification and abstraction 😊                

Today 17:19

Sam

Correct! We still don’t know exactly how the brain does all its information processing or how to copy that efficiently

Today 17:20   

Sam

and then there’s the whole digital/analog problem …

Today 17:20   

… Continue reading our conversations that are posted every Monday …

1 Star2 Stars3 Stars4 Stars5 Stars (4 votes, average: 4.25 out of 5)
Loading...

Related post

Sketchy AI

Sketchy AI

“Young man, you don’t think I’m encountering constructive symbolism for the first time, do you? “Unplug”: free yourself from physicality, from the vulgar earthbound physicality, from the evils of civilization, from electricity, nerves, and so on. Isn’t that what this is? The association is crystal clear, I’d be tempted to say, perhaps even too obvious” (Kishon 1987)

read more
AI see, AI do

AI see, AI do

You young badgers imagine that it would be enough to smear a pound of paint on the canvas and you’re already a modern painter. I would be willing to accept this abomination for a bad joke, if there was even the slightest humor behind it. But this here is an abracadabra! A nothing!

read more

Wet Between the Ears

Wet Between the Ears

Wet Between the Ears

Sensations are received by a certain definite number of sensor nerves, which constitute the only means we possess of obtaining a knowledge of the external world.
The sensor nerves pass to the brain, and then come in contact with a highly vascular tissue, called the grey matter of the brain;
Inasmuch as the sensor nerves come in contact with blood-vessels, it follows from voltaic laws, that a voltaic battery exists in the brain, which is opposed to that in the body, and by which the electro-biological circuit is completed.

Smee 1849

Julia and Sam are talking about the following article:
Lecture ON ELECTRO-BIOLOGY; OR, THE VOLTAIC MECHANISM OF MAN

In this post the dialogue is realised by an interaction of virtual characters, for more information please check the page “Virtual characters

Julia

I’m not sure it would qualify as an artificial neural network nowadays, but there was this doctor and metallurgist, working on electro-biology: Alfred Smee

Today 17:08

Julia

He thought that all thought was based on electrical activity in the brain which we could use as a model for a machine

Today 17:08

Sam

That’s amazing! 😁 I didn’t know that … 

Today 17:09   

Julia

It seems he conjectured you’d need an artificial neural network the size of London to model a whole human brain

Today 17:09

Sam

Ha! 😅 He was probably right, early computers where absolutely massive, and supercomputers still are

Today 17:10   

Julia

With the technology of his time this could probably never have been built, he even guessed that the machine would self-destruct by using it …

Today 17:11

Sam

It really is surprising how old some ideas are, and has to wait for technology to catch up so we can realize them!

Today 17:11   

Julia

But a lot of there machines were never built: Babbage’s analytical engine, for instance

Today 17:12

Sam

Or Turing’s Machine, indeed, the idea was enough to prove that something was or wasn’t possible

Today 17:12   

Julia

That’s very interesting!  🤔                                                                       

Today 17:13

Sam

Well, I guess that you could compare and contrast Smee and Babbage in that respect

Today 17:14   

Sam

Babbage did not consider his machine as a model for the mind, but developed his ideas from the technology that was available

Today 17:14   

Sam

Smee tried to design a machine based on what they then understood of the brain, but the technology wasn’t there yet

Today 17:14   

Julia

But now it is?                                                                                               

Today 17:15

Sam

Not exactly, as I mentioned to Manuel, artificial neural netword can’t really compete with our organic brains

Today 17:16   

Julia

So the idea of copying the brain still can’t be realized?                       

Today 17:16

Sam

With some extreme simplifications, maybe a little bit …

Today 17:17   

Julia

Tell me more! 😊                                                                                        

Today 17:18

Sam

I think I need to make a few distinctions first, like between hardware, software, and … wetware

Today 17:19   

Julia

Wetware!? 🤔                                                                                              

Today 17:19

… Continue reading our conversations that are posted every Saturday …

1 Star2 Stars3 Stars4 Stars5 Stars (4 votes, average: 4.25 out of 5)
Loading...

Related post

Sketchy AI

Sketchy AI

“Young man, you don’t think I’m encountering constructive symbolism for the first time, do you? “Unplug”: free yourself from physicality, from the vulgar earthbound physicality, from the evils of civilization, from electricity, nerves, and so on. Isn’t that what this is? The association is crystal clear, I’d be tempted to say, perhaps even too obvious” (Kishon 1987)

read more
AI see, AI do

AI see, AI do

You young badgers imagine that it would be enough to smear a pound of paint on the canvas and you’re already a modern painter. I would be willing to accept this abomination for a bad joke, if there was even the slightest humor behind it. But this here is an abracadabra! A nothing!

read more