“AI Is Going to Eat Your Lunch”: Lily Hu on Philosophy Without AI

00:00:00: Hi, my name is Tobias Haberkorn and you're listening to Airlift from Berlin Review.

00:00:06: A quick note before we start.

00:00:07: Airlift as a podcast was launched at the beginning of two thousand twenty six made in produced by our team at Berlin review where be published three print editions year on top of regular online content.

00:00:18: if enjoy listening to us or coming into alive events in Berlin will have special offer for At blnreview.de-airlift you'll get a free trial, including the latest edition of our beautifully designed Berlin Review Reader.

00:00:32: Try this offer at BLNReview.De-airLift.

00:00:37: In September OpenAI announced that its AI had cracked the Navier Stokes Problem One of the notoriously tough Millennium Price problems in mathematics.

00:00:46: And almost right away there was controversy over whether the AI really solved it or if unpublished work by two mathematicians had been leaked to open AI, which the company denies.

00:00:59: The mathematicians who checked the one hundred and sixty-six page proof say that so far this proof hasn't taught them much.

00:01:05: A group of fields medalists even put out a statement That AI companies are being short sided By racing for solutions To problems instead Of seeing the process of solving the problem as the whole point.

00:01:17: This goes beyond mathematics, of course.

00:01:20: AI has started to shape other academic fields including the humanities and even philosophy feel premised on people thinking for themselves.

00:01:28: Today's guest is Lily Hu a professor of Philosophy at Yale University who was trained in both philosophy and applied mathematics.

00:01:36: this summer lily wrote an essay for Berlin review called undoing philosophy about why she refuses to use AI In her work or in the classroom.

00:01:46: We talk about what happens to philosophy in other disciplines when AI intervenes and the pressure that Lily feels from her peers who have embraced the technology.

00:01:56: You also discuss plant intelligence, then why a cold email from a stranger has become a precious form of connection?

00:02:04: Here's my conversation with Lily-who please enjoy!

00:02:14: One of the things you explained is that for now are refusing to use AI in your own philosophy.

00:02:22: So, you're writing philosophy and doing philosophical research but also teaching philosophy And I imagine whether or not students should be allowed to or shouldn't use AI on writing essays is coming up between you and your students.

00:02:43: So I'm interested, how does this conversation actually go?

00:02:47: Well in my classroom the converse... it's a one-directional conversation!

00:02:53: It's a directive on my part which i think ...which i'm sure is not pedagogically optimal but i think we are at ..I mean..my own views have obviously evolved.

00:03:05: everything everyone has to say that says they've evolved since November of And where I am at now is probably the most extreme that i've been actually on this journey.

00:03:16: Okay, and That has led me to the conclusion that I think that I'm going-I'm interested in a return To A more conservative pedagogy which Is to say that... Yeah.. ...I take it as one of my professional responsibilities In One Of My yeah you know my use of expert judgment & discretion to try to Teach students that then get to inculcate in them.

00:03:43: Yeah kind of set up habits of minds and skills And practices but they might not themselves right already appreciate, so Right now my approach is you know do not use AI You don't meaningful eye contact with each student

00:03:57: okay?

00:03:58: Look

00:04:01: there's many different tactics Speak to them person-to-person human and trying To be a little bit shaming trying to be a Little bit.

00:04:14: you know I was like you one day, You Know Like there's all these different techniques.

00:04:19: You know pointing to the studies whatever i do it All

00:04:23: what are they responding?

00:04:25: Honestly I am very lucky.

00:04:27: I have students that Are Interested in my courses And I Have you know like My syllabus It has like A four paragraph screed, gentle screED.

00:04:39: I'll call it a gentle screet about why you know AI use even dialoguing with AI is at part of your ideation process as bad and I forbid it.

00:04:50: And so the students who end up joining my class they say their bottom They're like absolutely Right now?

00:04:58: I think i'm at an eighty percent rate.

00:05:00: Eighty five percent are doing my classes without any use of AI.

00:05:04: Yeah are they're struggling.

00:05:09: Do you

00:05:09: think there's an age component here?

00:05:11: because I think all the people who were until very, very few years ago using AI for writing tasks wasn't an option?

00:05:18: So basically everyone is like leading the AI discourse now and criticism on so forth in even the AI doomsday scenarios.

00:05:29: All of them, I think are too old for having been educated in a world where the default could be that any sort of writing is done by AI.

00:05:37: Including me and including you!

00:05:40: So what i find interesting when I speak with younger colleagues, for example... ...is they're pushing back against AI much more maybe than the older generation does.

00:05:52: Is that something you observe?

00:05:53: Yeah

00:05:53: absolutely

00:05:54: Okay..I'm

00:05:55: not probably one of the youngest on my faculty and we have a range of views about AI.

00:06:01: We just had a meeting about pedagogy in the classroom, you're hard-pressed to find somebody who's going to be more against it than I am but i'm certainly one side of that many older faculty like your seventy eight?

00:06:17: why do feel they need to count out this technology?

00:06:20: The psychology is rather complex.

00:06:24: I have been surprised by the fact that many of my older colleagues Have been much less resistant.

00:06:30: Okay, i can't completely chalk it up to naivete Mm-hmm?

00:06:34: I think part of it is feeling like you speak

00:06:35: about in a piece About you know The de skilling and the argument That if young people are never Never learn how To solve problems with their own brain And mind only then there will be A moment Of deskilling happening.

00:06:47: that's one of the you Know patriarchal arguments towards the young.

00:06:50: but actually It could also happen too

00:06:52: To

00:06:54: the old hands.

00:06:55: Absolutely, I think and i worry about that... ...I think there is this presumption of asymmetry.

00:07:00: That's really not very justified.

00:07:04: Many reasons why we say it's important you read actually yourself It's really important to write.

00:07:12: You experience a struggle with writing.

00:07:14: You linger in confusion In puzzlement.

00:07:18: thats part of normal learning process.

00:07:20: That's all true, and that continues to be true for even when you are a fully-fledged philosopher.

00:07:29: I don't like calling them skills because it makes some seem as if they can have developed them but then become good.

00:07:36: It is really just the practice of philosophy which is something we're continuously doing And this part where the usage of AI is completely antithetical regardless whether in my view or student You know, a seasoned

00:07:52: philosopher.

00:07:52: Yeah since the beginning of the year it really feels like massive acceleration on what AI is able to do.

00:07:58: we continuously read in the news now you know The Doomsday scenario but also and that's very interesting I think over the past couple months A number of mathematical problems seem have been solved by AI That some of them have been formulated almost one hundred years ago And are now eventually being solved with AI by AI agents.

00:08:21: I haven't heard any philosophical breakthrough, thanks to Claude or Chatchapiti.

00:08:27: so is there none?

00:08:30: And if no what does it say about philosophy and what doesn't say about the AI?.

00:08:34: I think that the fact that There has been no philosophical break through yeah i think It's.

00:08:42: you know That The interesting thing About math Is like it solved it Or didn't solve it and You can sort of just You know, you can even have these.

00:08:52: the proofs themselves are now checked by computers.

00:08:55: So like really where are we in that process?

00:09:00: I think The interesting asymmetry with philosophy here is like think about what it would mean for it to Have a philosophical breakthrough.

00:09:06: Mm-hmm In order for us to know Like if they're you know, the categorical imperative is true.

00:09:11: like Kant was correct Somebody would have to be like okay well let me read the argument Right?

00:09:16: Like, it doesn't have that same...

00:09:18: Yeah.

00:09:19: ...that in math you can just sort of have a computer check-it.

00:09:21: but if the argument is persuasive and convincing and true like only YOU CAN CHECK THAT!

00:09:26: Only YOU can see whether they argument What...?

00:09:29: And then

00:09:29: value than this kind of philosophy or this kind revelation or comprehension of the problem Is IN THE PROCESS OF ACTUALLY GOING THROUGH THE ARGUMENT YOURSELF

00:09:41: both going through the argument yourself and seeing whether it rings true to you, whether you are in fact persuaded by it.

00:09:47: Whether that claims being made touch you... That's not true.

00:09:52: with math or many other sciences I presume?

00:09:59: In the natural sciences like biology, medicine

00:10:06: etc.,

00:10:09: all the problems will be solved.

00:10:11: I mean, that's the positive scenario.

00:10:12: and then we have a sort of like a solved world situation as is part of the AI discourse.

00:10:16: I think the thing that's curious about philosophy is one of especially contemporary philosophy... The point for you as a philosopher to try to occupy some part of logical space nobody has tried to occupy before because otherwise it would just be redundant.

00:10:31: It's all brand differentiation!

00:10:33: For philosophers And thats interesting that kind of shows, we're not really converging actually.

00:10:42: In a progressive way you might see in the natural sciences and there are many smart people who have read each other's arguments at end-of-the day their intuitions clash or they had pre theoretical commitments divergent something.

00:10:59: so it is just completely Again, like it would be antithetical to the enterprise.

00:11:07: To think that shape of things this you know let a thousand flowers bloom in philosophy would converge to one solution by Claude or That One Solution By Claude Would Win Over Everybody when everybody's entire trade has been quibbling with most minor details to occupy your own little corner of the philosophical, like logical space.

00:11:32: So this leads us back actually to some really hard distinctions that have been made in history of humanities and history philosophy and history sciences between what actually distinguishes The humanities on one hand and the science is only other right?

00:11:46: And What I find puzzling is that...I mean if we look at a lot these discourses happened at the beginning of the twentieth century And one of the results, I think was that philosophy was turning ever more decidedly towards language.

00:12:05: It's use natural language and as you say development argument in process really walking through a thought that distinguishes... That is another way to distinguish humanities from hard sciences that end up with a result in there, it's resolved and you don't have to look at anymore because... You move on the next problem.

00:12:28: So what is puzzling about AI actually so good at its language itself?

00:12:33: I know!

00:12:34: It IS very, very good at learning.

00:12:36: That why many of my reasons for not using an AI in classroom or work isn't based how good it really is.

00:12:42: The reality here is that IT'S VERY GOOD.

00:12:46: It has its tick.

00:12:47: But it's NOT good FOR YOU if you study.

00:12:52: I mean, it's not good for the person.

00:12:54: No that

00:12:54: is what i'm saying!

00:12:56: Not you personally but...

00:12:57: Exactly no It certainly isn't good for a human being who has been using it But its really good at producing strings of sentences That court sort-of again.

00:13:09: This is not always the case.

00:13:10: You have to be very skillful user so sort of latch onto each other To reconstruct an argument.

00:13:18: I think AI Is actually Very Good At These Things.

00:13:22: It's very good at drawing distinctions, probably overly-good at drawing distinction.

00:13:25: Yeah yeah the typical sort of like twist of the sentence.

00:13:28: Absolutely!

00:13:30: You mentioned one example in piece a combination of two obscure theories and then obviously the AI finds a way to binding them together into an argument?

00:13:39: Absolutely

00:13:40: absolutely.

00:13:40: so you know.

00:13:43: maybe what should... make us ask as philosophers is whether that kind of philosophy we were doing a lot.

00:13:52: Whether it was really so worthwhile, you know?

00:13:55: If... if philosophy just sort-of these highly technical jargon games where they have different jargon maps and then connect parts to one another like an argument how illuminating or rewarding.

00:14:14: So that would be.

00:14:15: pardon me for this maybe anachronistic distinction, but I was educated in Europe and European universities And we always talked about you know the difference between analytical philosophy and continental philosophy.

00:14:25: An analytical philosophy is In my understanding much more what you described in the beginning of your piece?

00:14:32: You call it points philosophy.

00:14:34: Do you think this strand of philosophy will Because it is dominant.

00:14:39: I mean even in Europe today, in academic philosophy.

00:14:44: do you think this will come to an end?

00:14:45: maybe and slightly more speculative often disregarded continental authors would comeback?

00:15:01: If anyone's coming back... I do wonder whether, yeah.

00:15:08: I think analytic philosophy cannot be the same because if you are in the business of manufacturing points... Okay and then you conceive your task that way?

00:15:19: This is part reason why a lot of analog philosophers have really embraced AI Because they identify themselves as discovering truths which are just these points.

00:15:29: If that does assimilate analytic philosophy more towards the natural sciences, and it'll eat your lunch.

00:15:42: So I think that is true.

00:15:45: now there's going to be all sorts of institutional responses.

00:15:48: what are we gonna do?

00:15:50: However much analytic philosophers love AI, I think they probably like their own jobs a little bit more than that.

00:15:56: So there will be all sorts of I think analytic philosophy or institutions of analytic philosophy try to overcome that difficult truth and maybe something what the mathematicians are trying do with protecting math.

00:16:13: You know an analytic philosophy identifies with math but wants protect its domain.

00:16:19: Might go the way of those mathematicians who have signed all these declarations and sign all these open letters about how mathematics is really about understanding not just about The results themselves, you know that's sort of how a lot of mathematicians Have approached the AI solutions to the Navier stokes problems.

00:16:38: They've tried To say look we still need humans understanding, mathematical understanding.

00:16:45: It won't do us any good if we have all these proofs that no one can understand and We need to you know ensure the legacy of mathematics goes on in the form Of students in a form of training.

00:16:55: so That's one way then like philosophy I think while acknowledging their job at least At The level of points is.

00:17:04: But You describe almost the opposite Way?

00:17:06: You speak about the peer pressure of using it.

00:17:09: yeah Yeah

00:17:10: seems To be very strong which was a surprise for me when i read your piece.

00:17:15: It is very... I think it's very dichotomous.

00:17:19: The field is very split, the people who are very pro or very sort of enthusiastically pro and compared with other fields in humanity they're much more measured by their responses put it forward a hard line because they're philosophers.

00:17:41: They like to really needle you, well why are so against that?

00:17:45: So I think lot of analytic philosophers and would be scared about the numbers but...I imagine plurality use it pretty extensively And everybody is running into this question how we can square those two things?

00:18:02: How do we square our own personal usage while also trying defend The value of me.

00:18:08: I think that's not a good equilibrium.

00:18:11: So everybody has kind of caught in something like attention or paradox.

00:18:16: They're hoping that someone is going to solve for them.

00:18:19: Okay, so but if we now bracket the question whether or not you use AI for doing philosophy AI also poses a number of new philosophical questions doesn't it?

00:18:31: And one of them then i find really interesting and fascinating as this idea Actually basically, yeah.

00:18:39: The term superintelligence and what it entails And there were a number of security breaches this summer AI agents going rogue Coordinating doing hacks?

00:18:52: What is most fascinating about from my point-of view Is that It took weeks or even months to discover That.

00:19:00: So something's out there in operating.

00:19:03: You know There are now a form of intelligence that is probably affecting the social world, certainly.

00:19:12: The real-world and we don't fully know or understand how it operates And We Don't Even Know Whether It's Really There.

00:19:20: So there this term of the treacherous turn in literature on that where you have to actually think about possibility forms of AI don't even reveal themselves to humans and try to do their own thing without letting us know.

00:19:45: It's almost a theological question, isn't it?

00:19:47: Because you have this other agent and intelligence out there... ...and you don't actually know what that does or where it operates.

00:19:57: What great world in which we live!

00:20:04: I don't know how to think about a lot of this intelligence stuff.

00:20:06: I mean, you'll give an example... ...I'm reading this book now on plant intelligence and basically the author is telling us in it-

00:20:21: Which book?

00:20:22: It's

00:20:23: called The Light Eaters by Zoë Schleinger.

00:20:25: Okay!

00:20:27: Um, its great would recommend

00:20:29: it.

00:20:30: And the point is basically plants.

00:20:34: They're doing so much stuff, they are all around us.

00:20:37: Maybe their conscious maybe their intelligent...maybe there communicating.

00:20:41: how can we help and anthropomorphize these things you know?

00:20:44: And definitely the premise of this book is being trying to ask us reconsider what definitions that we've got.

00:20:51: Definitions of consciousness, definitions of intelligence, definition of communication that were used thinking about exclusively in human context.

00:20:59: think whether they applied a plans.

00:21:01: given all this new amazing plant botany, plant biology, plant signaling science that's been done in the past few decades.

00:21:13: And it's funny reading this book and reacting so differently to A forest could be intelligent, or a forest can be conscious.

00:21:22: Like I don't feel any aversion to that like i don't Feel any need to protect the language of intelligence for consciousness For humans

00:21:30: from the farthest

00:21:31: form the corn

00:21:32: yeah okay

00:21:34: But i do feel That you know and it's rather complex behavior.

00:21:39: there

00:21:39: is also The debate on various forms of intelligent animals, other intelligent life forms.

00:21:46: Which

00:21:47: is it by the way a discourse that has become very strong like I would say ten years ago when we were already in a phase of discussing all these AI problems.

00:21:55: but we discussed them as though they where science fiction and now their becoming apparently real so fast.

00:22:04: But i guess my point there are complex behaviors around us life, like plants and animals.

00:22:12: And in the world of digital things whatever they are.

00:22:18: How should we choose to understand that complexity?

00:22:23: Whether we choose it as forms or not forms of intelligence is unease with rewarding some categories Like, I think that's sort of the interesting thing.

00:22:40: Right?

00:22:41: I was just reading on this book and thinking about...I have no problem calling a beach tree intelligent but i really am so adverse to calling AI agents intelligent or calling them super-intelligent God forbid!

00:22:52: Or calling them conscious even though their behaviors are complex.

00:22:55: Of course they're behaviors are unpredictable And extremely intricate.

00:23:00: It is better at me than doing so many things.

00:23:09: preserve the language of consciousness or intelligence as a sort of honorific?

00:23:13: Yeah, for humans.

00:23:16: Or even orb and plants!

00:23:18: Okay okay not for machines...

00:23:20: You want to preserve it as an honorific I think.

00:23:22: And so i guess in response to your question you know what are we supposed to think about this?

00:23:27: It's like- What do we suppose to think?

00:23:28: that fact that y'know..I've been using this analogy Like there was my- The COVID might have been result of a lab leak.

00:23:35: It escaped, it went rogue.

00:23:38: The novel coronavirus went rogue and escaped some lab.

00:23:43: would we award intelligence to that?

00:23:45: To an extremely fast evolving virus that in fact creates a huge global pandemic And is able to basically evade so many different forms of interventions by humans?

00:23:58: I mean i think We wouldn't be likely to call the virus like the novel coronavirus, for instance, intelligent or conscious.

00:24:05: And that complexity makes us uneasy maybe but I think it's interesting that kinds of complexities displayed by AIs are really making this uneasy and threatened in a way to the complexity of plants.

00:24:25: I guess.

00:24:25: whether or not we allow the term of consciousness, you know to be a apply-to AI.

00:24:31: The question then becomes what does it actually do?

00:24:33: What is this capable often and what are the guardrails?

00:24:36: And You have worked...I mean you also trained in mathematics So i suppose you have like some sort of technical understanding Also of the underlying technology.

00:24:49: In your philosophical work, you've looked at how certain characteristics that show up in data are interpreted and then attached to real-life social concepts such as race for example.

00:25:04: Without now going into the details because I don't fully understand what your papers really were about but it's sort of like adjacent The kind of research that should interest AI developers also.

00:25:22: Because the models are constantly inferring from data to the real world.

00:25:28: Do you know how receptive the companies and the developers?

00:25:36: Well, given their interest in hiring philosophers which seems to be that they are often very interested in hiring philosopher's I think They Are probably Very Interested and these sorts of questions.

00:25:46: They want i mean especially Anthropic where?

00:25:49: They're very interested In you know understanding Claude as it were.

00:25:53: yeah You Know what is It like To Be Claude?

00:25:55: How Does Claude Experience the world or its Chats Or Whatever?

00:25:59: um I Think That It is no accident that these companies are finding themselves foray into the world of philosophy both because There are just very interesting philosophical questions like you said they have been raised.

00:26:14: I mean, there were.

00:26:15: i was working at a time which it's so hard to believe Because if feels like eons ago but to talk about algorithms and machine learning.

00:26:22: Yeah Oh my god You're so dated.

00:26:23: yeah Like you know That was like The cutting edge uh, ten years ago.

00:26:29: Right?

00:26:30: And now we've totally moved past that.

00:26:33: but there's many of the themes are just... The same

00:26:37: right.

00:26:37: how do I understand what correlations mean about the world?

00:26:41: What is the gap between correlation which isn't a feature data or a feature statistics and for instance causation Which Is A Feature Of The World How you know how our how slash to what extent can machines Grasp that net that gap.

00:26:57: That's always there between data mirror data and the real world of things and objects And feelings in people, and so on.

00:27:05: I think that as these tools have become So much more sophisticated beyond the grasp of not just me but seemingly even their very own creators We we are beginning to just ask philosophical questions.

00:27:20: You know what is it?

00:27:20: To know something?

00:27:21: yeah this is a philosophical question.

00:27:23: you know What does it do?

00:27:25: Yeah, what is it to understand something?

00:27:27: You know as opposed to merely predicts.

00:27:30: Something yeah What's the gap between you knowledge and understanding?

00:27:34: What's that gap between prediction And explanation?

00:27:39: these are kind of classic questions in philosophy That The more that AI companies are interested In Kind of arguing that or presenting their tools as able To do them complex thing, not just know but understand.

00:27:55: Not just predict but explain the more that they need to venture into this basically these land of philosophy.

00:28:02: Yeah That's fascinating.

00:28:03: But then do you think that philosophers can like play a role in?

00:28:08: In the regulation off these machines?

00:28:10: I Think that The right philosophers me for instance if anyone.

00:28:19: I think it's very hard To think about What kinds of academic expertise would be actually practically useful?

00:28:28: Because many of the big questions here are political questions.

00:28:31: Yeah And philosophers Are very bad at political question because you know philosophers like to abstract.

00:28:39: politics is about concrete

00:28:40: yeah, but

00:28:41: concrete and You know I Think philosophers tend to idealize.

00:28:46: in politics it's really about The real.

00:28:48: so i think that philosophers have a couple of traits that actually make them rather poor, I think practical guides to what we ought do.

00:28:59: And i think thats why a lot of philosophers are in my view not very wise about AI because they tend to abstract the tool rather than thinking as it concretely exists and how people infact are oriented towards it or use it.

00:29:19: You know given that aversion to dealing with the messiness of reality I would say philosophers are probably

00:29:24: not at the forefront Of proposing some solutions, but what about the institutions then?

00:29:29: I mean you alluded to that before When we talked about your students and also about your colleagues.

00:29:35: What our-what is your university for example?

00:29:39: yay.

00:29:39: Are they putting in place any sort best practices for either the teachers or some form of like, institutionalization... Yeah.

00:29:50: ...of the way you deal with that new technology?

00:29:55: Or is everyone confused and...?

00:29:57: Second thing!

00:29:58: The short answer is no there has been no kind of broad guidance except for general guidance which is be aware this tool exists now transforms the classroom, whether you like it or not and transforms your course.

00:30:14: You need to AI proof your assignments for instance.

00:30:17: there's guidance at that level which is be aware that students can use this tool to complete your assignments in about ten seconds.

00:30:25: but as too how actually deal with beyond simply being aware of it?

00:30:30: There isn't a lot of guidance.

00:30:32: then we just had meeting last week where somebody from our teaching learning center came.

00:30:38: They came in and they just started by asking us, like each of us introduce yourself.

00:30:42: And explain how you use AI.

00:30:44: You know it's kind of not really wanting to get a lot people different opinions and voices trying to allow for pluralism as opposed to the best practices guide.

00:30:57: I think that is for many reasons but The technology changing very quickly putting forth a guideline that will be obsolete within three months.

00:31:10: I think some other institutions have taken charge and put forward more restrictive AI policies, so you see Berkeley Law School they are like no-AI.

00:31:24: basically there's this first year course sequence at Chicago but many places it is.

00:31:33: The guidance that you get is if anything like here's how you could incorporate it into teaching rather than Here are ways to reaffirm traditional You know ways of teaching.

00:31:44: It's more like, you Know we need to be open to what this technology can do.

00:31:49: How?

00:31:49: Do you feel how do you AI proof your classes for example?

00:31:52: well apparently it's impossible because a fairly

00:31:55: fairly it's only in person access.

00:31:57: Yeah in-person examination that can exclude the use of Yes,

00:32:02: and it's a huge cost.

00:32:03: So I'll tell you what i'm doing...I'm doing an in-person essay so they're gonna write an essay for two hours.

00:32:10: Blue books, you know?

00:32:11: Compositions

00:32:12: like Seated at the Table

00:32:15: Physical writing implement

00:32:17: Yeah

00:32:19: But I really went back and forth on this because when we all go to our time of school or college university, graduate school whatever.

00:32:32: The amount of time that one spends in the classroom is a fraction Of the time you spend learning like everybody's memories of school or like pulling an all-nighter at the library Or like, you know Frantically writing their fifteen page paper.

00:32:44: The day before it was due.

00:32:45: You know It's these sorts of extended struggling sessions That take place necessarily because they are so extended outside of the classroom Where you might meet.

00:32:54: and

00:32:54: also reading

00:32:55: reading my god I know

00:32:57: getting to like understanding trying to summarize and.

00:33:01: Absolutely,

00:33:03: yes absolutely.

00:33:04: so I you know i was going say look im gonna give up on all of those longer term exercises.

00:33:13: Im just going do the kinds assessments that could be in a classroom but in end thought thats too limiting.

00:33:20: my students are still writing long-term paper.

00:33:23: will they use AI?

00:33:25: We'll see!

00:33:26: Are you confident?

00:33:29: There's a range of tactics that I'm using to try and get them not to use it.

00:33:33: And we'll see, if people... If i get fifty percent participation rate-if i got fifty percent of the people who are not using AI

00:33:42: That is an amazing

00:33:43: score!

00:33:43: Isn't this so sad?

00:33:44: Fifty

00:33:45: percent?!

00:33:46: Yeah.. If can get seventy five per cent ,I will say like thats great success.

00:33:50: At this point where you know We cannot hope to recover The learning experience Of twenty ten what's to do.

00:34:00: But you can think about this in a sort of like defensive way, how do I AI prove my classes?

00:34:06: In order make sure that my students really read the assignment or really do their writing themselves?

00:34:12: but let's maybe try and think about it progressively.

00:34:16: What from your point-of-view as a philosopher teacher should young people learn today To prepare them for the world off tomorrow?

00:34:25: We're at an interesting time now because we keep hearing constantly that we're at the either, you know The precipice of total disaster and extinction of humanity or basically Total abundance

00:34:40: a

00:34:40: world without work Etc.

00:34:42: etc.

00:34:43: And everyone all the big experts seem to agree That they have no idea what their words gonna be like in ten years.

00:34:49: Yeah

00:34:53: I definitely feel that the future is completely open right?

00:34:56: Like there's this totally frightening, you know.

00:35:03: What's gonna happen to us feeling?

00:35:07: I think the students themselves have that feelin'.

00:35:11: You can't go to college now thinking your'e going get a computer science degree and you're gunna be done

00:35:18: safe!

00:35:19: When only five years ago this was exactly what u were supposed to

00:35:21: do!

00:35:22: Exactly!

00:35:22: That is most risk-averse Prudent decision that one could make.

00:35:28: an investment in your future was and that is as it turns out the most easily Automatable job.

00:35:36: And meanwhile, you know being a masseuse like the most protected job because they eyes can't You don't not yet?

00:35:43: They can give you deep tissue massage.

00:35:45: so Given that's the world which we're living in a World where you know there's just its really difficult to understand what it's good at What will actually be able to incorporate what kinds of pipelines it'll actually be able to be used in for work.

00:36:02: I don't really, you know...I wouldn't be able give an argument to my students

00:36:06: like... Let's try make this through not from a practical and utilitarian point-of-view.

00:36:13: What should i study?

00:36:14: To be ready for the world but would maybe want to study and enjoy In order to prepare yourself what's to come, you know.

00:36:28: Doesn't this actually lead us back too?

00:36:31: You should study more humanities or read literature...

00:36:37: You're not gonna get a job anyway!

00:36:39: Read the great books and...?

00:36:42: I think so.

00:36:44: It is a challenge because we've been used to justifying those enterprises reading the great book set in Humanities instrumentally for so long.

00:36:56: That it's a means to and the end of having a brilliant degree?

00:37:01: Yeah,

00:37:01: critical thinking you know something like vaguely like critical thinking skills or being able to understand the news with a critic.

00:37:08: yeah this sort of vague gestures at being able navigate the world more productively because your person-of-letters And its really not clear whether that is true.

00:37:23: what was the intrinsic value of the humanities?

00:37:29: or philosophy, reading?

00:37:32: What was that again?

00:37:33: We've kind of forgotten where it is.

00:37:36: When I talk to my students and try now... It's an opportunity for academic institutions schools I suppose it's true that we have all the time in the world is saved now.

00:37:53: Like, The Best Case scenario materializes and you never have to write... All of the former English majors had to write ad copy.

00:38:00: they'll never have a single word of add-copy ever again You know?

00:38:04: It's all to save us times so we can do other stuff Which has to be

00:38:08: what?!

00:38:09: So i think We are really pushed back To ask fundamental questions about What gives our life meaning what we choose to do with the time that we have, What kind of a person you want to be?

00:38:21: You know these fundamental questions That in The past We would say oh your no.

00:38:29: What A luxury To Be able to just Ruminate on Questions like what Do I do With the Time i Have!

00:38:35: You don't have any time you have to go to work yeah.

00:38:38: so i think it is an opportunity not that These institutions are taking that Opportunity to see schooling as Not Just Slotted into the capitalist sort of progression, right?

00:38:52: But think about it actually as a place for the development Of The Human To A Certain End.

00:38:59: What About A More Muscular Or Even Militant Stance Toward It?

00:39:04: I mean in the piece funny segment of your essay you say that When You Claim That You Don't Use AI As A Philosopher You Have Basically Two Attacks.

00:39:13: One Is That Other People Will Say That You're presumptuous.

00:39:16: You think you're a manual con?

00:39:18: Like, you can do this without AI.

00:39:20: And the other one is that your being a Luddite.

00:39:22: So I think it goes back to the nineteenth century and textile workers who destroyed machines we were arriving at making them obsolete.

00:39:33: so what's the appetite for?

00:39:35: Luddism?

00:39:36: really forms of sabotage or refusal of the AI infrastructure also that's being built.

00:39:50: Is there any appetite among your students, our colleagues for that?

00:39:54: Oh my colleagues I doubt it.

00:39:56: but i think one funny consequence bots were the worst thing that could happen.

00:40:06: You're like, oh it's filled with bots!

00:40:09: The companies wanted to get rid of the bots because they are making their platforms unusable.

00:40:12: and now the company is themselves trying to create bots as fake accounts?

00:40:18: And I was talking about some my students in there... social media isn't fun anymore you know, like it's just not very usable.

00:40:26: You know even things like dating apps.

00:40:29: these are not really useable.

00:40:30: at some point when people using AI agents to respond your pick up line on hinge It is so depressing.

00:40:39: So I think that there may be a minor leadism in the form of Not being online in the same way as five years ago.

00:40:50: The more muscular forms of refusal are probably to be found in sort-of, the US especially in anti data center movement.

00:40:59: It's such a remarkable groundswell across the ideological spectrum opposition to data centers and it comes with little nimbyism.

00:41:13: but I think that we can understand as people just don't want this life, like in their local lives.

00:41:22: And what can we do to resist that?

00:41:24: Well, we can resist the physical infrastructure.

00:41:28: What's the role of a place like Yale?

00:41:30: or what's the rule over place like an academic institution?

00:41:32: and I think they're probably

00:41:33: clear.

00:41:34: yeah it's less clear with them.

00:41:36: Yeah so speaking about your colleagues also about you students um The thing with AI is also that its one of best kept secrets that people are using it.

00:41:46: I mean, you write about who's going to be transparent about their use of AI?

00:41:51: And you come out and say, I categorically refuse to use it and everyone should not use it.

00:41:57: so that is a very strong stance.

00:41:59: how did people react?

00:42:00: what was the feedback from your colleagues

00:42:04: and students?

00:42:05: Yeah...I have used it in this sense.

00:42:10: I don't use it as a part of my work or are part of regimented uses.

00:42:13: It's not apart of my philosophical work, or part of teaching and generate notes but definitely try to do tasks with that like am i using this tool wrong?

00:42:24: I think what is so interesting about AI?

00:42:26: the totally divergent reactions people have too.

00:42:29: It's amazing.

00:42:30: I don't do any of these other things that used to, and the people who just find it really difficult.

00:42:34: they found very difficult use And i have found a kind of difficulty.

00:42:37: us if tried to be like visit every single one Of these websites tell me Like Who is on editorial board for instance?

00:42:43: That has been Very hard To Do.

00:42:46: So i Have Used In Obviously.

00:42:50: i have to Arm up against my students.

00:42:54: Yeah, so I read a lot of AI text because i get A lot of emails and Because you know my students are AI adult But...

00:43:05: You don't have to handle a lot Of excel sheets in your professional

00:43:09: life?

00:43:09: Exactly!

00:43:09: So if did I would have no problem using an AI for Excel just To be clear.

00:43:16: Um..I've gotten Emails that were very about the piece being like.

00:43:22: thank-you For saying how I felt but I haven't.

00:43:24: um,

00:43:26: that I haven't been able to articulate and That has been you know very very heartening to hear.

00:43:32: But i think what's interesting is how many people like?

00:43:35: I Know People who've read the piece.

00:43:37: Who have not said anything to me And I wonder What that Is about.

00:43:40: I mean You know I I remember my.

00:43:44: I'm Not on Facebook I'm not on any social media, but my

00:43:48: Facebook was ten years ago.

00:43:49: Okay?

00:43:49: Well okay But this is why all the philosophers are indeed on it.

00:43:53: So so i've Not on these.

00:43:55: on These um apps with my apps platforms and My friends were setting me screenshots of like people in The profession sharing my article

00:44:04: On facebook

00:44:05: on faith.

00:44:06: ok And you know It's Like.

00:44:10: You Know Is so happy to get All This Traction.

00:44:11: But the Comments Were like Horrifying.

00:44:15: Seems like she's taking a really hard line stance doesn't seem that informed.

00:44:18: You know, it is exactly the kind of thing I said was just like well you know She so presumptuous.

00:44:23: Yeah hubris or like she such a moralistic figure Mm-hmm It's impossible to say and think because people feel shame they feel uncomfortable.

00:44:33: yeah They don't like having to confirm.

00:44:36: And they also you know and there's just open as best kept secret Like you said?

00:44:41: That when someone speaking to her directly you have to find other reasons to undercut them.

00:44:46: Oh, it's because they're moralistic and they are not... It

00:44:50: is funny when I was preparing the episode thinking about historical analogies into a moment that we were now in transformative.

00:44:58: We don't know how its gonna end up.

00:45:02: And one of the analogies i thought of Was the so-called discovery of the unconscious the beginning of the twentieth century, or around the turn-of-the-century.

00:45:11: The second to last turn of the century with Sigmund Freud and that happened also at the same time as theory of relativity was formulated roughly at the sametime And a lot of philosophical literature in the beginning speaks about how anthropologically like humans were sort uprooted and unsettled by the discoveries when suddenly, oh actually it's not my free will.

00:45:40: There is something called the unconscious that somehow like you know directing me in these unknowable ways.

00:45:49: I think by now we've gotten used to that and there's a language around it, which has evolved over hundreds of years.

00:45:56: But maybe the thing with AI is so unsettling but makes people insecure or even more defensive because there are no social rules nor conventions about this thing.

00:46:08: This was at the very beginning for us in our new social world also.

00:46:17: I get a lot of AI emails from random people, but also from students.

00:46:21: And I respond very matter-of-factly which is like...I typically do not respond to AI e-mails But i want make eye contact with you try and instill a norm because think your right we're in a normal or there are different norms for different spaces There's so many overlapping norms.

00:46:41: Is it okay for student reach out to me using an AI email?

00:46:47: Asking me to let them into my class like no, I would say.

00:46:52: No that but Like There's a way.

00:46:56: you know You know they're sending these emails on in mass To so many teachers But there was also away.

00:47:00: which like email is the communication?

00:47:02: We are you and Me Are communicating right now yeah And So that's contested, like.

00:47:08: it's not clear where that particular message falls under some bureaucratic arrangement.

00:47:12: Where you're trying to get into my class or under a direct person-to-person communication.

00:47:16: and I have definitely found myself trying to reinstate certain types of norms of communication because of AI.

00:47:24: in previous areas were I would eye was lot more lack.

00:47:29: so i got all these emails being talk, and it's like a random person from Ireland.

00:47:41: And I'm like what?

00:47:42: I don't know you!

00:47:43: Now i respond to those emails.

00:47:45: one of the things that AI has pushed me do is be like this is the person who was reaching out to me trying to make human-to-human contact.

00:47:52: That Is A Precious Thing.

00:47:53: now that's precious today in twenty six whereas five years ago there was an incumbrance It Was An intrusion on my inbox.

00:48:03: The way that AI has transformed social space, at least for me personally... ...has made me kind of re-root myself in some of those very basic and fundamental forms of human connection.

00:48:16: That I totally took for granted like a cold email.

00:48:21: Maybe it's gonna be the nail on the coffin of mass e-mailing.

00:48:24: Oh my god!

00:48:25: Here here is hoping.

00:48:37: You can read Lilly's essay Undoing Philosophy and everything else we publish at blnreview.de.

00:48:45: As I mentioned, go to blnrevue.de-airlift for a free first month including the latest edition of our beautiful print magazine.

00:48:53: This episode was edited by Caitlin Roberts And recorded at Chapters Bookshop in Berlin.