Cognitive Bias

[dallasjustice]

I think everyone wins when this topic is discussed. We don't need to agree on it. You always win if you can get others to just ask the right questions. The answers really aren't that important, IMO.

The problem with other forums dedicate to the my gear is the best sport, is that there's no tolerance for discussing this topic. I'm not sure what bias that is; cognitive disonance?

 

[Thomas savage]

I think everyone wins when this topic is discussed. We don't need to agree on it. You always win if you can get others to just ask the right questions. The answers really aren't that important, IMO.

The problem with other forums dedicate to the my gear is the best sport, is that there's no tolerance for discussing this topic. I'm not sure what bias that is; cognitive disonance?

I believe it is cognitive dissonance, though that shows through posts all over our forum as well... you can't escape being human.

 

[Jakob1863]

And yet... the world is shaped by non-scientists who are prey to every bias in the chart. Almost everything that is designed is created 'sighted'. Every work of art, musical composition, book, newspaper article, piece of architecture, graphic design, political philosophy, business etc. etc. is created in full view and as a direct consequence of those biases. How come anything works at all? Could it be that "bias" is often another word for experience, or even wisdom?

And in addition, a lot of development/progess is made just by accident. "Bias" is not bad per se, it might be another word for experience as you suggested, but it might be also another word for gut feeling. Gigerenzer wrote a lot about people´s use of heuristics, often without knowing about using it. The players trying to catch the ball were one of his examples, as most (if not all) did not know which method they were using to decide where to run.

 

[Jakob1863]

I kind of agree with the paper - I am reasonably convinced that there are philosophical reasons that prevent a digital computer from becoming conscious. However, I don't quite agree with the author that the information processing analogy can be dismissed so easily.

I think he is missing that an artificial neural network can learn to do exactly as he describes, even though it is constructed using an 'IP'-based system (CPU, memory etc.). Similarly, in an artificial neural image recognition system, for example, memories are 'fuzzy' and distributed throughout the whole network rather than a precise 'recording' that is stored in a specific location - which he cites as an example of how brains and 'IP' systems differ.

A digital computer in combination with software can, indeed, behave in ways very similar to a brain.

Maybe i am mistaken, but i think he argues instead, that, although an artifical neural network can be sort of intelligent, our brain still doesn´t work like a computer....

 

[Brad]

That is a really useful chart. I have not come across many of the biases in green before.
But remember that some of that biases listed can be quite useful approximations, and we can't function (is a reasonable time) without approximating our reality. Just keep in the back of your mind that there may be another explanation.
As was mentioned above, the biases help explain a lot of the endless discussion on audio fora.
Another way to look at it, is if someone is doing a listening 'test' they are really performing a psychoacoustic experiment, and subject to those cognitive biases.
If someone is performing a measurement with equipment, they are doing an engineering experiment.
The best practice for those two cases is vastly different (ie different methodologies).

 

[The Smokester]

Real nice chart, but I'd give a nod to this one: Naïve Realism

View attachment 6570

Allow me to tumble down this chart and discuss the article below.

...

I don't know if this is right or wrong, but is an interesting read...

https://aeon.co/essays/your-brain-does-not-process-information-and-it-is-not-a-computer

This article seems to be restricting itself to the commonly used computer architectures, such as the Von Neuman-like machine used in the CPU of your laptop. There are other architectures which also are "Universal Computing Machines", like the Touring machine, fundamentally different in structure, but still Universal.

In the case of Neural Nets (which are not necessarily Universal), it should be recognized that many implementations are simulations executed on a Von Neumman machine: But neural nets are thought to be phyically instantiated in at least some living things. The behavior of a living brain, such as that of a nematode, has been reproduced in computational neural nets instantiated on Von Neuman computers. I see from Google that there is now a project called OpenWorm to completely simulate a worm at the cellular level. (Note: A neural net-based computer is not necessarily Universal.)

Other Universal Computers can be formed from "cellular automata" which can be multi-dimensional arrays of unit cells which see, and respond to, only their very nearest neighbor...the ones touching them. (Conway's Game of Life is a cellular automata.) The number of possibilities here make Go look like Tic-Tac-Toe. These are difficult to understand, let alone program...They can be programmed by training (machine learning). Cellular automata-like architectures give rise to computation-like theories of life, where DNA is viewed as the program which is inserted into a fluid matrix of molecules comprising the computer.

I know this is all just "blah blah blah". Still, I think the article's thesis that "brains aren't computers" is, at best, premature.

 

[RayDunzl]

I know this is all just "blah blah blah". Still, I think the article's thesis that "brains aren't computers" is, at best, premature.

My thought:

"Brains aren't Computers" because "Computers aren't Brains" (yet).

 

[The Smokester]

My thought:

"Brains aren't Computers" because "Computers aren't Brains" (yet).

Yes. Quite insightful and much more succinct. Thank you for your thought.

 

[Cosmik]

My thought:

"Brains aren't Computers" because "Computers aren't Brains" (yet).

Hmm. Surely this is like saying "The human heart isn't a pump because pumps aren't hearts".

 

[RayDunzl]

Hmm. Surely this is like saying "The human heart isn't a pump because pumps aren't hearts".

Good try, but I'm thinking implantable mechanical heart, and not, at least at the human level, a workable replacement for a brain...

http://www.texasheart.org/Research/Devices/abiocor.cfm

Otherwise, yeah, I guess you're right.

 

[Cosmik]

Good try, but I'm thinking implantable mechanical heart, and not, at least at the human level, a workable replacement for a brain...

http://www.texasheart.org/Research/Devices/abiocor.cfm

Otherwise, yeah, I guess you're right.

I a similar vein*, we can repair brains with chips.

A brain–computer interface (BCI)... is a direct communication pathway between an enhanced or wired brain and an external device. BCIs are often directed at researching, mapping, assisting, augmenting, or repairing human cognitive or sensory-motor functions.

* pun not intended