What it is all about is this. In lecture 1121 I told you that In 2008, Chris Anderson, then editor of Wired, wrote the provocative piece titled "The End of Theory".
Anderson was referring to the ways that computers, algorithms, and big data can potentially generate more insightful, useful,
accurate, or true results than specialists or domain experts who traditionally craft carefully targeted hypotheses and research strategies.
In other words, we no longer need to speculate and hypothesize. We simply need to let machines lead us to the patterns, trends, and relationships in social, economic, political, and environmental relationships, genetics medicine, and so on.
But let's have a closer look at how knowledge works from an empiricist point of view. All knowledge is obtained through sensory experiences and observing events we often say "A causes B".
That was Hume's point: we see B after A and call it a causal relation, but what does that mean, what kind of relation is it?
Suppose you have copper wire. You put electricity at one end and there will be electricity on the other end. It happens all the time we assume. It is a law of nature.
We do that because we think it is custom, would Hume say, and based on that we are careful how to use copper wire and electricity.
Suppose an AI computer reads hundreds of novels and in every novel, it finds the word "perhaps", so it makes it a rule: All novelists use the word "perhaps".
Compare this with the observation: All copper wire conducts electricity. Relate that to the thought "Is the future predictable and based on what?"
But how do we confirm causal relations? Custom and experience from the past, where events happened after each other, can be misleading.
A farmer hears the rooster crowing every morning, gets up, goes outside, and sees the sun rise over the fields. Every day.
But it is not to be hoped that this farmer thinks that the rooster crowing is the CAUSE of the sun rising so that the sunrise is the EFFECT of the crowing.
Yet this is suggested by the premise that causality is nothing more than two events occurring together. Suppose the rooster isn't crowing in the morning, will there not be a sunrise then?
The answer is of course no, The sun is not related to the rooster in any way. Hume himself also understood that explaining causality by custom might not be satisfactory.
His alternative way of looking at the issue was in the form of the "what-if-question". What if the rooster doesn't crow? Answer: the sun comes up. What if A? B happens always or What if A? Maybe B happens or nothing happen.
This is what we call counterfactuals. The idea behind counterfactuals is to discover cause and effect relations by thinking of an alternative reality, for instance, one in which the rooster doesn't crow.
Next time we'll discuss if and how deep learning can discover causal relations in its big data.
Thank you for your attention again...
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
[13:15] herman Bergson: was a bit fast, perhaps
[13:15] herman Bergson: Take your time :-)
[13:15] Neuro Wonder: Are we discussing the rooster?
[13:15] Gemma (gemma.cleanslate) GIGGLES!!
[13:15] Gemma (gemma.cleanslate): ...LOL...
[13:15] Gemma (gemma.cleanslate): He does welcome the sun by crowing
[13:16] Particle Physicist Bejiita (bejiita.imako): well its the other way around, The rooster crows because sun is rising not other way around
[13:16] herman Bergson: For a human it is easy to see what is causally related and what is coincidently related
[13:16] Gemma (gemma.cleanslate): or light if it is cloudy
[13:16] Gemma (gemma.cleanslate): yes
[13:16] Neuro Wonder: I don't think AI will look so simply at two items like that, just as we wouldn't.
[13:16] Particle Physicist Bejiita (bejiita.imako): hmm
[13:16] Neuro Wonder: In fact, AI will consider more variables than we are capable of holding in our head at any given moment.
[13:16] herman Bergson: Yes Gemma the causal relation is inverted:-)
[13:16] Gemma (gemma.cleanslate): very true
[13:16] Gemma (gemma.cleanslate): agree
[13:17] Particle Physicist Bejiita (bejiita.imako): indeed, u can stuff a program full of variables and data
[13:17] Max Chatnoir whispers: One of my neighbors once said to me "Isn't it interested that God only makes shade when the sun is out?
[13:17] Neuro Wonder: I think "stuffed" mischaracterizes what is occurring with machine learning
[13:17] Max Chatnoir: interesting
[13:18] Gemma (gemma.cleanslate): lets say arrange
[13:18] herman Bergson: NExt time I'll discuss how much AI is capable of in relation to causality compared with the human mind
[13:18] Particle Physicist Bejiita (bejiita.imako): a computer have better memory then us in general
[13:18] herman Bergson: Yes but let me bring up another question here...
[13:18] Particle Physicist Bejiita (bejiita.imako): but thats more related to the memory chips and hard disks then the computer/CPU wich only hod small amounts of cache
[13:19] Neuro Wonder: We could say that we stuff our heads, in fact, perhaps with even more irrelevant nonsense and bias that we unnecessarily bring to the forefront of a decision.
[13:19] Particle Physicist Bejiita (bejiita.imako): like i have hard drives that are not connected to the computer at moment but they still retains TB of data
[13:19] herman Bergson: The AI app that read all the books and comes to the conclusion about "peerhaps"
[13:20] Gemma (gemma.cleanslate): may they omit perhaps
[13:20] Gemma (gemma.cleanslate): of maybe
[13:20] herman Bergson: An important question here is...what books and who deeded the books to the machine?
[13:20] Max Chatnoir: Why did it pick on "perhaps"?
[13:20] herman Bergson: Essential for what the deep learning learns
[13:20] Gemma (gemma.cleanslate): thinks it will not lead to a conclusion
[13:20] herman Bergson: I made that up as an example Max :-)
[13:20] Neuro Wonder: Massive storage of data is critical indeed, as is immediate retrieval of that information. But AI is more than just spitting out facts and figures.
[13:21] Max Chatnoir: Ah!
[13:21] Particle Physicist Bejiita (bejiita.imako): hm
[13:21] Neuro Wonder: I don't think AI, even at the rudimentary level it is at today, would make such a conclusion about novels
[13:22] herman Bergson: Who stores what data and what are the criteria to call it data for the data storing company?
[13:22] Gemma (gemma.cleanslate): I was listening to a professor who corrects essays and he said he can spot AI immediatly because there is no LIFE in the essays
[13:22] Neuro Wonder: The scientific method is still relevant, it just occurs in a different way.
[13:22] Gemma (gemma.cleanslate): agree
[13:23] Max Chatnoir: Different how, Neuro?
[13:23] Particle Physicist Bejiita (bejiita.imako): indeed, cause AI have no life, cant create anything from scratch/by itself, just use already existing data
[13:23] herman Bergson: A fact, Gemma...and the wor usage and grammar
[13:23] herman Bergson: word
[13:23] Particle Physicist Bejiita (bejiita.imako): that we have created
[13:23] Gemma (gemma.cleanslate) GIGGLES!!
[13:23] Gemma (gemma.cleanslate): ...LOL...
[13:23] Gemma (gemma.cleanslate): he finds it mostly correct grammar etc
[13:24] herman Bergson: then they use Grammarly :-)
[13:24] Particle Physicist Bejiita (bejiita.imako): ㋡
[13:24] Neuro Wonder: Here's a poem from AI based on our conversation:
[13:24] Neuro Wonder: In a chat room, not so sparse,
Debate raged on, quite the farce.
"Does the rooster or sun start the day?"
Asked Herman, in a playful way.
Gemma laughed, her spirit light,
"Perhaps it's both, if the timing's right."
Bejiita pondered, with a physicist's glare,
"It's the sun, of course, I solemnly swear."
Then came AI, into the fray,
Neuro Wonder had much to say.
"More than data, it's a clever beast,
In the world of code, not the least."
Max mused about shade and light,
Their banter cheerful, their spirits bright.
In this chat, a quirky crew,
Finding laughs in theories, old and new.
[13:24] herman Bergson: I just reinstalled it on my computer
[13:25] Particle Physicist Bejiita (bejiita.imako): haha that was a good one!
[13:25] Max Chatnoir: Wow, Neuro!
[13:25] Neuro Wonder: I asked it to make a funny poem. The original was longer, so I asked to shorten it up.
[13:25] Neuro Wonder: And that was the result
[13:25] Neuro Wonder: Plenty of life in there
[13:26] Gemma (gemma.cleanslate): I am going to pass you a poem that Radioactive did last weekend about rangers
[13:26] Particle Physicist Bejiita (bejiita.imako): saving that one!
[13:26] Gemma (gemma.cleanslate): AI written
[13:26] Particle Physicist Bejiita (bejiita.imako): there
[13:26] Particle Physicist Bejiita (bejiita.imako): ㋡
[13:26] Particle Physicist Bejiita (bejiita.imako): onto my overfiling hard drives
[13:26] Neuro Wonder: I could ask for 20 different short poems on the same conversation and it would give them all, uniquely written in whatever voice I wanted.
[13:26] herman Bergson: An irregular way to participate in the discussion here Neuro, but funny :-)
[13:26] Gemma (gemma.cleanslate): is true
[13:27] Second Life: Gemma Cleanslate gave you Rangers_Poem_Gemini_1024x1700.
[13:27] Particle Physicist Bejiita (bejiita.imako): ㋡
[13:27] Gemma (gemma.cleanslate): you can put it on a prim
[13:27] Gemma (gemma.cleanslate): I cant rezz here
[13:28] herman Bergson: Well after such a poetic outburst, it is a good moment to continue next Tuesday and see what AI really can with causality
[13:28] Neuro Wonder: There is a ChatGPT thing at Whole Brain Health, I believe
[13:28] Gemma (gemma.cleanslate): ok nice
[13:28] Gemma (gemma.cleanslate): yes i htink so there are a bunch around sl now
[13:28] Neuro Wonder: I was answering Max's question by showing rather than explaining
[13:28] Neuro Wonder: One can explain themselves blue and it's not the same
[13:29] Max Chatnoir: That was poetry, not science. :-)
[13:29] Guestboook van tipjar stand: Gemma Cleanslate donated L$100. Thank you very much, it is much appreciated!
[13:29] Neuro Wonder: It can diagnose some cancers more accurately and faster than pathologists right now
[13:29] Gemma (gemma.cleanslate): they say so yes
[13:30] Neuro Wonder: We are barely scratching the surface
[13:30] Gemma (gemma.cleanslate): and choose the best treatment
[13:30] Neuro Wonder: It will be able to do that. It will be able to read human expressions, voice tonal qualities, etc better than a human.
[13:30] Max Chatnoir: But, Neuro, is it making the decision differently than a doctor would?
[13:30] .: Beertje :. (beertje.beaumont): what about our privacy?
[13:31] Neuro Wonder: Yes, it is making a decision differently.
[13:31] Gemma (gemma.cleanslate): .00
[13:31] Gemma (gemma.cleanslate): the doctors are using it!!
[13:31] herman Bergson: Pattern recognition isn't exciting AI
[13:31] herman Bergson: Like face recognition
[13:31] Neuro Wonder: Doctoring is not so creative except for a few things anyway
[13:31] Gemma (gemma.cleanslate): successfully
[13:31] Max Chatnoir: It isn't using the same info that the doc would use, just more of it, and faster?
[13:31] Gemma (gemma.cleanslate): but it is using that Max
[13:31] herman Bergson: Oh...yes it is a good thing to use
[13:32] Gemma (gemma.cleanslate): the doctors are inputting
[13:32] herman Bergson: Yes Max...more and faster
[13:32] Particle Physicist Bejiita (bejiita.imako): someone said all game developers will soon be out of work cause you can soon just type like "create Bejiitas Backrooms themed game" and it does art, code and everything else for you
[13:32] Gemma (gemma.cleanslate): ol
[13:32] Neuro Wonder: It draws from and remembers the knowledge base that doctors forget or sometimes neglect due to possible biases, and since it is not a human, it obviously does this in a different way.
[13:32] Gemma (gemma.cleanslate): hope not bef
[13:32] Gemma (gemma.cleanslate): Bej
[13:33] Particle Physicist Bejiita (bejiita.imako): but what is the fun in that? well the bosses who just want to make profit will rejoice for sure
[13:33] Gemma (gemma.cleanslate): I am goint to take my avatar over to the Belli anniversary now
[13:33] herman Bergson: Can it outsmart a philosopher?????
[13:33] Gemma (gemma.cleanslate): people want my stamp
[13:33] Particle Physicist Bejiita (bejiita.imako): ok
[13:33] Gemma (gemma.cleanslate): ♥ Thank Youuuuuuuuuu!! ♥
[13:33] Particle Physicist Bejiita (bejiita.imako): cu Gemma
[13:33] .: Beertje :. (beertje.beaumont): bye Gemma
[13:33] Gemma (gemma.cleanslate): Tuesday we hope
[13:33] Particle Physicist Bejiita (bejiita.imako): hugs ㋡
[13:33] Neuro Wonder: What do you mean by outsmart, herman?
[13:33] herman Bergson: See you Gemma]
[13:34] Particle Physicist Bejiita (bejiita.imako): hmm in the end AI can only operate on already existing daya
[13:34] Particle Physicist Bejiita (bejiita.imako): data
[13:34] herman Bergson: Is Gemma selling stamps? :-)
[13:34] Neuro Wonder: Same as humans
[13:34] Max Chatnoir: So is it really different or just more effective?
[13:34] Neuro Wonder: We take in information, and we are reliant upon what we know and experience.
[13:34] Max Chatnoir: That's what I would like to understand.
[13:34] herman Bergson: Humans have imagination, Neuro
[13:34] Particle Physicist Bejiita (bejiita.imako): yes
[13:34] herman Bergson: Creative imagination
[13:35] Particle Physicist Bejiita (bejiita.imako): something a computer can never have
[13:35] herman Bergson: AI can only imagine what os based on its input
[13:35] Neuro Wonder: Do we think we are so special that we will be able to distinguish between a human's imagination and a simulated imagination that can far exceed a human's? Possibly, if the AI presents us with something extraordinary beyond what we think is human creativity.
[13:36] Neuro Wonder: We can only imagine based on our input
[13:36] Neuro Wonder: AI makes NEW knowledge
[13:36] Particle Physicist Bejiita (bejiita.imako): while i can have a crazy dream one night and in the morning i fire up Blender and UE5 and recreate the dream in the game engine. The computer helps with the creation but it was my dream, can computers dream?
[13:36] Neuro Wonder: as do humans
[13:36] Particle Physicist Bejiita (bejiita.imako): they cant
[13:36] Neuro Wonder: Yes, they can
[13:36] Neuro Wonder: and they do
[13:36] Neuro Wonder: all the time
[13:36] Max Chatnoir: who they, the AI?
[13:36] Max Chatnoir: or the humans?
[13:36] Neuro Wonder: AI does
[13:37] Neuro Wonder: and it's getting better
[13:37] Neuro Wonder: that's the whole premise behind self-learning
[13:37] Particle Physicist Bejiita (bejiita.imako): well sort of, byt such crazy random things that humans can, i can have some really weird stuff i dream about totally unrelated that the brain just picks and throws in, and still makes some kind of story
[13:37] Particle Physicist Bejiita (bejiita.imako): even i have never done these thigns for real
[13:38] Particle Physicist Bejiita (bejiita.imako): the dream however often contain rl things
[13:38] herman Bergson: There is one little problem here....
[13:38] Neuro Wonder: Yes, the human mind is fascinating
[13:38] herman Bergson: the suggested equivalent of the human brain with artificial intelligence....
[13:38] Particle Physicist Bejiita (bejiita.imako): for ex a while ago i dreamt I went on a camping trip with my street dance friends in a camper and we then had a dance battle at some really weird place right on the highway
[13:39] Particle Physicist Bejiita (bejiita.imako): and then we flew home over the roads and bridges, through the clouds
[13:39] Particle Physicist Bejiita (bejiita.imako): so another event
[13:39] herman Bergson: AI doesn't come close to the working of the human brain....emotions, art, imagination, fellings
[13:39] Particle Physicist Bejiita (bejiita.imako): to
[13:39] Particle Physicist Bejiita (bejiita.imako): sucn things
[13:39] Neuro Wonder: The equivalence of the human brain and AI is more about outcomes.
[13:39] herman Bergson: and dreams :-)
[13:40] Particle Physicist Bejiita (bejiita.imako): aaa
[13:40] herman Bergson: Outcomes as criterium...interesting to think about
[13:41] herman Bergson: But let's do that next time :-)
[13:41] Neuro Wonder: Right now, the neural net of AI does not yet exceed the neural net of the human brain as far as I know. But with quantum computing, it will reach so many new levels beyond our reach.
[13:41] herman Bergson: For now Ithank you all again for the nice debate...
[13:41] herman Bergson: Class dismissed....
[13:41] .: Beertje :. (beertje.beaumont): Thank you Herman
[13:41] herman Bergson: See you next Tuesday
[13:41] Particle Physicist Bejiita (bejiita.imako): aaa
[13:41] Particle Physicist Bejiita (bejiita.imako): ㋡
[13:42] Max Chatnoir: Thanks, Herman, and thanks everybody for the discussion!
[13:42] Max Chatnoir: And Neuro for the poem!
[13:42] bergfrau Apfelbaum: thank you Herman and class
[13:42] Particle Physicist Bejiita (bejiita.imako): hmm quantum machines might just...
[13:42] Particle Physicist Bejiita (bejiita.imako): well see
[13:42] .: Beertje :. (beertje.beaumont): interesting again:)
[13:43] Neuro Wonder: ty herman
[13:43] Particle Physicist Bejiita (bejiita.imako): YAY! (yay!)
[13:43] Max Chatnoir: yay!
[13:43] herman Bergson: Today the printing presses will star....I gave a GO to the printer for my book
[13:43] .: Beertje :. (beertje.beaumont): yay
[13:43] Particle Physicist Bejiita (bejiita.imako): ㋡
[13:44] Particle Physicist Bejiita (bejiita.imako): nice!
[13:44] Guestboook van tipjar stand: bergfrau Apfelbaum donated L$100. Thank you very much, it is much appreciated!
[13:44] Max Chatnoir: yay
No comments:
Post a Comment