We still have some classic philosophical questions about AI. For that, we have to go back to 17th-century England. This is what the philosopher John Locke (1632 - 1704) assumes, that the human mind at birth is a "tabula rasa", a blank slate.
Because the AI program AlphaGo Zero learned the game Go through a lot of practice and not through prior knowledge provided by experts, the program was built from scratch four centuries after Locke.
With his belief that knowledge only comes from experience, Locke marks the beginning of empiricism. In this way, he forms a fault line in Western philosophy.
From Plato onwards, the idea that knowledge is innate was dominant. Descartes and Leibniz, contemporaries of Locke, also fall back on that principle.
Plato considers sensory experience merely as the rediscovery of knowledge that is already unconsciously present.
By presenting man as a "tabula rasa", Locke undermines this tradition of thought. We have no unconscious, innate knowledge at all. Experience is the source of knowledge.
This leads to classic epistemological questions. Some of the current AI applications are based on induction. The issue arose already when we contrasted rationalism with empiricism.
In our thinking we use two approaches to gain knowledge: induction and deduction. Induction is a way of thinking that is opposed to deduction, the method of reasoning that traditional logic uses.
In addition, it is impossible to arrive at a false conclusion from true statements, the premises.
That characterizes deduction and makes logic so attractive for artificial intelligence. A typical example of such an "innate" knowledge model is of course expert systems.
Induction works very differently. You observe swans, perhaps hundreds of them: this swan is white, that one is white, etc. And then you come to the conclusion: all swans are white.
That is, of course, debatable, because the immediate question is: Have you really seen ALL the swans? Of course, that is not the case, so this generalization is not as watertight as a deductively derived conclusion.
Since contemporary AI uses experiences through data, induction is indispensable. If you focus on experiences, this will lead to inductively acquired knowledge.
This means that you never have 100% certainty about the truth of your conclusions, as there is in deductive reasoning.
Here we enter the field of "deep learning", probably the most frequently mentioned concept in the current advance of AI.
Deep learning is a type of machine learning and artificial intelligence (AI) that imitates the way humans gain certain types of knowledge.
Deep learning models can be taught to perform classification tasks and recognize patterns in photos, text, audio, and other various data.
It is also used to automate tasks that would normally need human intelligence, such as describing images or transcribing audio files.
Deep learning is an important element of data science, including statistics and predictive modeling. It is extremely beneficial to data scientists
who are tasked with collecting, analyzing, and interpreting large amounts of data. Deep learning makes this process faster and easier. (www.techtarget.com)
To be continued.....thank you for your attention
Main Sources:
MacMillan The Encyclopedia of Philosophy, 2nd edition
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
[13:19] Max Chatnoir: Thank you, Herman.
[13:19] .: Beertje :. (beertje.beaumont): Thank you Herman
[13:19] herman Bergson: A bit difficult lecture perhaps, today?
[13:20] Max Chatnoir: So AI with deep learning algorithms has both the power and the peril of induction?
[13:20] Nicky (nickyparis.bellecoeur): yes but could be wrong specially if trained in a biased way
[13:20] herman Bergson: Yes, Max...that is why systems sometimes show biases...
[13:20] herman Bergson: For instance because trained on pictures of only men, or white persons
[13:21] Max Chatnoir: So proceed with caution.
[13:21] Nicky (nickyparis.bellecoeur): For example there are documented cases where AI is used to judge criminals, this can end in high convictions to certain etnic groups vs a majority in a country. For example immigrants can be affected.
[13:21] herman Bergson: I heard about a lawsuit where a black person was accussed of a crime, based on facial recognition...
[13:22] herman Bergson: Yes...
[13:22] Lecturehall: 2024-03-26 [22:22] Curious George
[13:22] Nicky (nickyparis.bellecoeur): correct, and that is phrenology which has been proven to be wrong
[13:22] Particle Physicist Bejiita (bejiita.imako): finally! I GOT IN!
[13:22] Nicky (nickyparis.bellecoeur): using facial features to predict personality traits or psychopathology
[13:22] herman Bergson: The man was innocent but the facial recognition AI was inadequate to recognize black people properly
[13:22] .: Beertje :. (beertje.beaumont): hi Bejiita
[13:22] Max Chatnoir: Hi, Bejiita.
[13:22] Particle Physicist Bejiita (bejiita.imako): puh
[13:22] Particle Physicist Bejiita (bejiita.imako): hi all
[13:23] Particle Physicist Bejiita (bejiita.imako): viewer issues
[13:23] bergfrau Apfelbaum: hi Bejiita:-)
[13:23] Particle Physicist Bejiita (bejiita.imako): ㋡
[13:23] herman Bergson: In the Netherlands we had a system that selected people on their family name...was it not Dutch...then posibbler fraudster...
[13:24] herman Bergson: Became a tremdous scandal
[13:24] Particle Physicist Bejiita (bejiita.imako): hmm that sort of thing, I've heard similar stories
[13:24] Particle Physicist Bejiita (bejiita.imako): not going well
[13:24] herman Bergson: And caused a lot of misery to people
[13:24] Max Chatnoir: like you have said, GIGO
[13:24] Particle Physicist Bejiita (bejiita.imako): hmm
[13:25] .: Beertje :. (beertje.beaumont): GIGO?
[13:25] herman Bergson: This makes induction an interesting subject of debate
[13:25] herman Bergson: Garbage in Garbage out
[13:25] Max Chatnoir: But induction is inquired for learning new things.
[13:25] herman Bergson: Next lecture we'll look at what David Hume had to say about it -)
[13:26] Nicky (nickyparis.bellecoeur): oh I can't wait for it
[13:26] herman Bergson: Of course a lot of science is based on induction and the scientific method....
[13:27] herman Bergson: That is why final conclusions should be open to falsification
[13:27] Max Chatnoir: black swans
[13:27] herman Bergson: They exist :-)
[13:28] .: Beertje :. (beertje.beaumont): yes, i've seen them
[13:28] herman Bergson: hmm...where did that professor go? who was standing here
[13:28] Max Chatnoir: George?
[13:28] herman Bergson: Didn't even say Goodbye...:-)
[13:28] Max Chatnoir: He disappeared.
[13:29] herman Bergson: Curious was his name
[13:29] Particle Physicist Bejiita (bejiita.imako): hmm
[13:29] .: Beertje :. (beertje.beaumont): maybe crashed
[13:29] herman Bergson: Ahh yes can happen
[13:29] herman Bergson: Anyway...if you need some homework....
[13:30] Particle Physicist Bejiita (bejiita.imako): aaa sl seems a bit temeramental, i first thought viewer was broke, because it ran out of memory as i forgot to cluse Unreal and Visual Studio before so was worried i broke the software
[13:30] herman Bergson: check out terms like induction, deduction and deep learning.
[13:30] Particle Physicist Bejiita (bejiita.imako): was stuck at requesting region capabilities
[13:30] Particle Physicist Bejiita (bejiita.imako): but when i logged on to the burning man region it worked
[13:30] Max Chatnoir: I'm glad you got it.
[13:30] Max Chatnoir: in
[13:30] Particle Physicist Bejiita (bejiita.imako): i refused to give up
[13:30] Nicky (nickyparis.bellecoeur): I need to go to sleep, goodnight everyone
[13:30] bergfrau Apfelbaum: i once heard that once AI takes over, there will be enough jobs for humans. Namely AI monitoring :-)
[13:31] Max Chatnoir: Good night, Nicky.
[13:31] Max Chatnoir: And database checking!
[13:31] herman Bergson: Indeed Bergie.....we have to keep a close eye on these developments
[13:31] Particle Physicist Bejiita (bejiita.imako): i guess
[13:31] Particle Physicist Bejiita (bejiita.imako): hmm something is still off graphics is not loaded completley, well i see u at least
[13:32] Particle Physicist Bejiita (bejiita.imako): but walls and floors are grey
[13:32] herman Bergson: So this deep learning is digging into data bases.....but who fills these data bases with what.....so what is the thing learning
[13:32] Max Chatnoir: O see your avatar, Bejiiita
[13:32] Max Chatnoir: I seee..
[13:32] Particle Physicist Bejiita (bejiita.imako): the AI will soon fill it by itself i guess
[13:32] Max Chatnoir: sorryk, having a little trouble with the keyboard
[13:33] Particle Physicist Bejiita (bejiita.imako): aaa i also see myself and all of u
[13:33] .: Beertje :. (beertje.beaumont): and who's data bases are this?
[13:33] herman Bergson: Indeed Beertje!
[13:33] Particle Physicist Bejiita (bejiita.imako): guess I did like the Unreal C++ i wrote before
[13:33] herman Bergson: Russia, China...the US?
[13:33] Particle Physicist Bejiita (bejiita.imako): working on a nice dialogue system
[13:33] Max Chatnoir: Yes it AI can create fake data, then that's a problem.
[13:33] Max Chatnoir: if AI...
[13:33] bergfrau Apfelbaum: the Ai is therefore no different from a coffee machine... both require maintenance
[13:34] herman Bergson: Creating fake data is already commonplace stuff, Max
[13:34] Particle Physicist Bejiita (bejiita.imako): indeed and it then fills the DB with that fake data and then we are screwed
[13:34] herman Bergson: Fake movies, fake voices
[13:34] Particle Physicist Bejiita (bejiita.imako): indeed its still just a machine
[13:34] herman Bergson: Ohh...we just had an issue here in the Netherlands...
[13:34] Particle Physicist Bejiita (bejiita.imako): crap in = crap out
[13:35] herman Bergson: The face of a female member of parliament was inserted in a porn movie
[13:35] Particle Physicist Bejiita (bejiita.imako): ok
[13:35] Particle Physicist Bejiita (bejiita.imako): oow
[13:35] .: Beertje :. (beertje.beaumont): ah yes
[13:35] .: Beertje :. (beertje.beaumont): she was not pleased
[13:35] Particle Physicist Bejiita (bejiita.imako): deepfaking/deepshaming
[13:35] Max Chatnoir: I can imagine not!
[13:35] herman Bergson: And she also was a member of a Christian political party.....so imagine the impact
[13:35] Particle Physicist Bejiita (bejiita.imako): (shuddders)
[13:36] Particle Physicist Bejiita (bejiita.imako): well
[13:36] .: Beertje :. (beertje.beaumont): can happen to anyone
[13:36] herman Bergson: Yes, it can...
[13:36] herman Bergson: Horrible but true
[13:36] Particle Physicist Bejiita (bejiita.imako): I guess hmmm
[13:37] .: Beertje :. (beertje.beaumont): it's hard to find the one who didi it
[13:37] .: Beertje :. (beertje.beaumont): did it
[13:37] Particle Physicist Bejiita (bejiita.imako): can imagine
[13:37] herman Bergson: That is always the problem with computer stuff on the internet.....who did it....
[13:37] Particle Physicist Bejiita (bejiita.imako): probably made by a botnet or something
[13:37] Particle Physicist Bejiita (bejiita.imako): impossible to track
[13:37] herman Bergson: so many ways to hide
[13:38] .: Beertje :. (beertje.beaumont): christian or not, it's horrible for every woman
[13:38] herman Bergson: Sure
[13:38] Particle Physicist Bejiita (bejiita.imako): hmm
[13:38] Particle Physicist Bejiita (bejiita.imako): porn sites are so raw
[13:38] .: Beertje :. (beertje.beaumont): how do you know Bejiita:))
[13:39] Particle Physicist Bejiita (bejiita.imako): lol
[13:39] herman Bergson: lol ...got the same question in mind Beertje :-)))
[13:39] Particle Physicist Bejiita (bejiita.imako): research
[13:39] Particle Physicist Bejiita (bejiita.imako): hehe
[13:39] .: Beertje :. (beertje.beaumont): laughs
[13:39] bergfrau Apfelbaum: lol
[13:39] bergfrau Apfelbaum: * Mister PLUPP *
[13:39] Particle Physicist Bejiita (bejiita.imako): indeed almost 100% of the stuff is a direct turn off for me
[13:40] Particle Physicist Bejiita (bejiita.imako): i mean sheesh!
[13:40] herman Bergson: Well..you go on with your research, Bejiita, and after that we go on with the next lecture on Thursday :-)
[13:40] Particle Physicist Bejiita (bejiita.imako): hehe
[13:40] .: Beertje :. (beertje.beaumont): you looked before you turned it off:))))
[13:40] Particle Physicist Bejiita (bejiita.imako): im gonna research my code a bit more after this i think
[13:40] herman Bergson: So.... thank you all again for ayyending the lecture....
[13:40] herman Bergson: Class dismissed...
[13:41] Max Chatnoir: Thanks, again, Herman
[13:41] Particle Physicist Bejiita (bejiita.imako): ㋡
[13:41] .: Beertje :. (beertje.beaumont): Thank you Herman
[13:41] bergfrau Apfelbaum: thank you, Herman and class!
No comments:
Post a Comment