Tuesday, April 9, 2024

1123: Hume on Induction...

 Where Locke inspired the makers of AlfaGo Zero with his opposition to innate ideas, David Hume (1711 - 1767) provided the material to criticize artificial intelligence a century after Locke. 

For Hume too, sensory experience is the most important source of knowledge about the world. But he goes one step further when he focuses on the question of what cause-effect relationships are based on.

     

We call these relationships causal relationships: A causes B. Hume distinguishes knowledge of the world, matters of fact, from what he calls relations of ideas. 

    

Relations of ideas are created entirely through thought. With a statement like "A triangle has 4 corners" you immediately know that the statement is wrong, 

   

but with a matters-of-fact statement like "The sun will not rise tomorrow" you basically do not know whether the statement is incorrect. You won't know until the next day. 

   

But of course it is a somewhat strange statement. There is a lack of reasons for making this statement and, according to Hume, these reasons can be traced back to causality. 

   

In fact, we are seeing two separate events one right after the other. You bump one billiard ball against another and that ball starts moving. Although one is the cause and the other is the effect, the relationship between them remains a mystery. 

   

We simply count on the way the world has behaved in the past to remain the same in the future. Hume calls this assumption the uniformity of nature, but such an assumption, based on induction, 

  

cannot be deduced from experience, because, for example, a statement such as "All ravens are black" already presupposes the uniformity of nature. 

   

In other words, it is impossible to provide causal relations and inductive conclusions with an indestructible foundation. An important part of AI is induction-based, referred to as deep-learning applications. 

  

These types of applications use so-called neural networks to perform their tasks. I won't bore you with all kinds of technical information about how neural networks work, but we have already noted that this type of applications can make mistakes and be biased. 

    

The AI uses big data, comes to conclusion A and from then on assumes that A fits in the uniformity of nature. The question now is: Is A a lawlike event or a coincidental consequence that emerges from the many data? 

   

In the first case it is possible to make predictions based on the past, in the other case great restraint is required. For AI, making this distinction is a very difficult problem.


Thank you for your attention again...


Main Sources:

MacMillan The Encyclopedia of Philosophy, 2nd edition

Routledge Encyclopedia of Philosophy, 1995
 http://plato.stanford.edu/contents.htm
Guido van der Knaap: Van Arititles to Algoritme (2023(


TABLE OF CONTENT -----------------------------------------------------------------  


  1 - 100 Philosophers                                              9 May 2009  Start of

  2 - 25+ Women Philosophers                              10 May 2009  this blog

  3 - 25 Adventures in Thinking                               10 May 2009

  4 - Modern Theories of Ethics                              29 Oct  2009

  5 - The Ideal State                                               24 Febr 2010   /   234

  6 - The Mystery of the Brain                                  3 Sept 2010   /   266

  7 - The Utopia of the Free Market                       16 Febr 2012    /   383

  8. - The Aftermath of Neo-liberalism                      5 Sept 2012   /   413

  9. - The Art Not to Be an Egoist                             6 Nov  2012   /   426                        

10  - Non-Western Philosophy                               29 May 2013    /   477

11  -  Why Science is Right                                      2 Sept 2014   /   534      

12  - A Philosopher looks at Atheism                        1 Jan  2015   /   557

13  - EVIL, a philosophical investigation                 17 Apr  2015   /   580                

14  - Existentialism and Free Will                             2 Sept 2015   /   586         

15 - Spinoza                                                             2 Sept 2016   /   615

16 - The Meaning of Life                                        13 Febr 2017   /   637

17 - In Search of  my Self                                        6 Sept 2017   /   670

18 - The 20th Century Revisited                              3 Apr  2018    /   706

19 - The Pessimist                                                  11 Jan 2020    /   819

20 - The Optimist                                                     9 Febr 2020   /   824

21 - Awakening from a Neoliberal Dream                8 Oct  2020   /   872

22 - A World Full of Patterns                                    1 Apr 2021    /   912

23 - The Concept of Freedom                                  8 Jan 2022    /   965

24 - Materialism                                                      7 Sept 2022   /  1011

25 - Historical Materialism                                       5 Oct 2023    /  1088

26 - The Bonobo and the Atheist                             9 Jan 2024    /  1102

27 - Artificial Intelligence                                          9 Feb 2024    /  1108 


The Discussion


.

[13:23] .: Beertje :. (beertje.beaumont): Thank you Herman

[13:23] Particle Physicist Bejiita (bejiita.imako):

[13:24] herman Bergson: Main point is that induction isn't waterproof, but we as humans can see the difference between lawlike processes and coincidental events.

[13:25] Gemma (gemma.cleanslate): usually

[13:25] Particle Physicist Bejiita (bejiita.imako): i think I'm with u

[13:25] herman Bergson: This has happened at Amazon...

[13:25] Max Chatnoir: Oh, how?

[13:25] herman Bergson: They wanted to make their application procedure easier and faster with AI

[13:26] herman Bergson: Result....women were more often refused than men....

[13:26] Max Chatnoir: job application?

[13:26] herman Bergson: yes job applications

[13:26] Gemma (gemma.cleanslate): ha

[13:26] herman Bergson: cause...

[13:26] herman Bergson: Amazon had filled the AI with old applications and the managers prefered men in those

[13:27] herman Bergson: Or the story of this black man that was arrested for a crime....in the US

[13:27] Max Chatnoir: So the AI was sorting by previous preference?

[13:27] herman Bergson: He was recognized by camera surveillance....and AI

[13:28] herman Bergson: Yes Max

[13:28] Particle Physicist Bejiita (bejiita.imako): outdated input

[13:28] herman Bergson: But the AI was wrong.....

[13:28] herman Bergson: cause...

[13:28] Gemma (gemma.cleanslate): it is with photos yes

[13:28] Gemma (gemma.cleanslate): often

[13:29] herman Bergson: the AI wqas trained by looking at tons of white male pictures and just much less black people's pictures

[13:29] Particle Physicist Bejiita (bejiita.imako): sounds like it was fed data from like the 20s

[13:29] Gemma (gemma.cleanslate): they say Asian people have the most problems

[13:29] Gemma (gemma.cleanslate): at airports for example

[13:29] herman Bergson: Yes...this  leads to all kinds of ethical questions regarding the use of AI

[13:30] Gemma (gemma.cleanslate): I have a feeling it will be figured out for the most part as we and AI learn

[13:30] herman Bergson: As Hume already warnewd us....induction is not waterproof....and big data won't change that...

[13:30] Gemma (gemma.cleanslate): true

[13:31] Particle Physicist Bejiita (bejiita.imako): crap in = crap out

[13:31] Particle Physicist Bejiita (bejiita.imako): no matter how much data u have

[13:31] herman Bergson: Our brain doesn't function like AI does

[13:31] herman Bergson: We stall are smarter to interpret observations

[13:32] Particle Physicist Bejiita (bejiita.imako): AIs are not conscious, cant understand what it actually sees

[13:32] Particle Physicist Bejiita (bejiita.imako): just a machine

[13:32] herman Bergson: Indeed Bejiita

[13:33] Max Chatnoir: We have to learn which criteria are important and which are not.

[13:33] herman Bergson: Expert systems are deductive.....

[13:34] herman Bergson: The point is....deep learning has no criteria on board....

[13:34] herman Bergson: it distilles the criteria  from the big data it gets

[13:34] Particle Physicist Bejiita (bejiita.imako): it swallows it all

[13:34] Particle Physicist Bejiita (bejiita.imako): and then tries to grasp it somehow

[13:34] herman Bergson: and that can lead to strange results....

[13:35] herman Bergson: They tried to use AI t predict the outcome of a World Soccer tournement

[13:35] herman Bergson: One big disaster..

[13:35] herman Bergson: They had trained the AI with data from former championships...

[13:36] herman Bergson: but the resut was garbage

[13:36] Gemma (gemma.cleanslate) GIGGLES!!

[13:36] Gemma (gemma.cleanslate): ...LOL...

[13:36] herman Bergson: result

[13:36] Gemma (gemma.cleanslate): different teams

[13:36] Max Chatnoir: There are so many factors involved in winning or losing a game.

[13:36] Gemma (gemma.cleanslate): right

[13:36] herman Bergson: Indeed

[13:36] Particle Physicist Bejiita (bejiita.imako): ah

[13:36] Max Chatnoir: Or an election (gulp)

[13:36] Gemma (gemma.cleanslate): I am exhausted from last night

[13:36] herman Bergson: What game did you play Gemma?

[13:37] Gemma (gemma.cleanslate): the UCONN Women basketball team played at 9:45

[13:37] Gemma (gemma.cleanslate): and won

[13:37] .: Beertje :. (beertje.beaumont): played golg with Trump?

[13:37] Gemma (gemma.cleanslate) GIGGLES!!

[13:37] Gemma (gemma.cleanslate): ...LOL...

[13:37] .: Beertje :. (beertje.beaumont): golf

[13:37] Gemma (gemma.cleanslate): right

[13:37] Gemma (gemma.cleanslate): dont forget to buy his bible beertje

[13:37] Particle Physicist Bejiita (bejiita.imako):

[13:37] herman Bergson: lol

[13:37] .: Beertje :. (beertje.beaumont): lol

[13:37] Gemma (gemma.cleanslate): he is hawing bibles

[13:37] Gemma (gemma.cleanslate): sigh

[13:37] .: Beertje :. (beertje.beaumont): i know:))

[13:37] herman Bergson: What a bizar action that is...

[13:37] Max Chatnoir: !

[13:37] herman Bergson: And check for sticky pages...!

[13:37] Gemma (gemma.cleanslate): and is now saying we must bring back prayer to te country

[13:38] .: Beertje :. (beertje.beaumont): hahha sticky pages?

[13:38] Particle Physicist Bejiita (bejiita.imako): lool

[13:38] Gemma (gemma.cleanslate) GIGGLES!!

[13:38] Gemma (gemma.cleanslate): ...LOL...

[13:38] Particle Physicist Bejiita (bejiita.imako): omg

[13:38] .: Beertje :. (beertje.beaumont): ketchup pages

[13:38] Particle Physicist Bejiita (bejiita.imako): was thinking of something else but

[13:38] herman Bergson: Yes...it has such a gold coating  on the edges of the pages

[13:38] Particle Physicist Bejiita (bejiita.imako): ....

[13:38] Particle Physicist Bejiita (bejiita.imako): hmmmm

[13:39] herman Bergson: April 15 begins the trail

[13:39] Gemma (gemma.cleanslate): if it is not put off again sigh

[13:39] herman Bergson: T is doing the weirdest things to prevent it

[13:39] Gemma (gemma.cleanslate): they are fighting hour by hour

[13:39] Gemma (gemma.cleanslate): yep

[13:40] herman Bergson: Yeah...incredible....

[13:40] Particle Physicist Bejiita (bejiita.imako): sigh!(shudders)

[13:40] herman Bergson: Why does he do that...He is innocent....so what is there to fear?

[13:41] Particle Physicist Bejiita (bejiita.imako): Trump?

[13:41] .: Beertje :. (beertje.beaumont): ask AI why he does that

[13:41] herman Bergson: Such  a trail is a walk in the park for such an honest and innocent man

[13:41] Particle Physicist Bejiita (bejiita.imako): or who

[13:41] herman Bergson whispers: I need no AI for that Beertje :-)

[13:41] .: Beertje :. (beertje.beaumont): lol

[13:42] Particle Physicist Bejiita (bejiita.imako):

[13:42] bergfrau Apfelbaum: lol

[13:42] herman Bergson: .

[13:43] Gemma (gemma.cleanslate): wonder if he is using AI for excuses

[13:43] Particle Physicist Bejiita (bejiita.imako): who knows

[13:43] Gemma (gemma.cleanslate): and holdups on the trial

[13:43] .: Beertje :. (beertje.beaumont): he's too stupid for that Gemma

[13:43] herman Bergson: Ok...at least we have learned that  AI may look smart.....but it is dangerous to rely on data from the past. They are not a guarantee for the future.

[13:44] Particle Physicist Bejiita (bejiita.imako): true

[13:44] Gemma (gemma.cleanslate): his lawyers sound like they are using his words for everything they say the ideas are do dumb!!

[13:44] herman Bergson: Let's wait and see.....

[13:44] .: Beertje :. (beertje.beaumont): are there people still voting for Trum these days?

[13:45] herman Bergson: Observing it from a distance is fascinating....:-)

[13:45] Gemma (gemma.cleanslate): omg yes

[13:45] Gemma (gemma.cleanslate): OMG!!!

[13:45] herman Bergson: OH YES.....

[13:45] Particle Physicist Bejiita (bejiita.imako): GUUUH!

[13:45] Gemma (gemma.cleanslate): his base of believers are so strong

[13:45] Gemma (gemma.cleanslate): scary

[13:45] Max Chatnoir: And giving him money

[13:45] Gemma (gemma.cleanslate): yes

[13:45] Particle Physicist Bejiita (bejiita.imako): STORM THE CONGRESS!

[13:45] Gemma (gemma.cleanslate): most of which pays the lawyers' fees

[13:45] .: Beertje :. (beertje.beaumont): don't they THINK ??

[13:45] Particle Physicist Bejiita (bejiita.imako): GROAN!

[13:45] herman Bergson: There are even people that still claim that Trump is the acual president of the US

[13:46] Gemma (gemma.cleanslate): they just see things and him differently

[13:46] Gemma (gemma.cleanslate): no way to explain it

[13:46] Particle Physicist Bejiita (bejiita.imako): they are brainwashed

[13:46] Max Chatnoir: Well, if he really is, then he can't run again!

[13:46] Particle Physicist Bejiita (bejiita.imako): and bribed

[13:46] herman Bergson: ANd Biden?...That is an imposter...an actor

[13:46] Gemma (gemma.cleanslate): sigh

[13:46] Gemma (gemma.cleanslate): and toooooooo old

[13:46] Gemma (gemma.cleanslate): 3 y ears older than T

[13:46] Gemma (gemma.cleanslate): sigh

[13:46] herman Bergson: three yoers older that that orange

[13:47] Gemma (gemma.cleanslate): yes

[13:47] Gemma (gemma.cleanslate): crazy

[13:47] Particle Physicist Bejiita (bejiita.imako): at least Bien isnt a blubbbering idiot

[13:47] herman Bergson: ridiculous argument

[13:47] Particle Physicist Bejiita (bejiita.imako): like Trump ois

[13:47] Particle Physicist Bejiita (bejiita.imako): is

[13:47] Gemma (gemma.cleanslate): take off the yellow hair and the orange face and he looks darn old too

[13:47] Gemma (gemma.cleanslate): ridiculous

[13:47] Particle Physicist Bejiita (bejiita.imako): trump looks like a rotten orange as is

[13:47] Particle Physicist Bejiita (bejiita.imako): as rotten as he is

[13:47] herman Bergson: I guess this may be enough for today :-)

[13:48] Particle Physicist Bejiita (bejiita.imako): well time will tell who wins

[13:48] herman Bergson: I have lots of fun with Youtube

[13:48] Gemma (gemma.cleanslate): sigh

[13:48] .: Beertje :. (beertje.beaumont): but if Trump wins, fun is over

[13:48] Gemma (gemma.cleanslate): yes

[13:48] Particle Physicist Bejiita (bejiita.imako): (shudders again)

[13:48] herman Bergson: That's for sure, Beertje

[13:49] Particle Physicist Bejiita (bejiita.imako): indeed!

[13:49] .: Beertje :. (beertje.beaumont): "schuilen kan niet meer, heeft geen enkele zin..."

[13:49] herman Bergson: He is a danger to the whole world

[13:50] herman Bergson: SO, let's be patient...

[13:50] herman Bergson: Thank you all again and keep up the good spirit :-)

[13:51] herman Bergson: Class dismissed....

[13:51] Max Chatnoir: We live in hope.