#42 Cognitive Surrender? When AI Starts Thinking for You with Stephan Ruppert

00:00:00:

00:00:24: And welcome to the Culture Exchange Podcast.

00:00:27: My name is Raoko

00:00:29: Hello everyone, my name is Marie-Christine.

00:00:31: Together we are hosting this podcast for The International Community where we talk about living and working abroad, expat life welcoming special guests who share their expertise.

00:00:44: With our culture exchange podcast We're providing a platform regarding topics of culture identity To raise awareness and give insights on how to navigate the challenges, opportunities of a multicultural

00:00:58: life.

00:00:59: Join us in our journey around.

00:01:24: Today, we're sitting together with Dr.

00:01:26: Stefan Rupert again because We already had the pleasure chatting With you in episode thirty two.

00:01:34: Can I understand culture?

00:01:36: That was The title.

00:01:37: a very successful Episode with many listeners reaching out to us.

00:01:42: So here we are again Stefan and talking about AI And it impacts on human.

00:01:50: Yeah, good morning and thanks for having me again.

00:01:53: It's a big honor to be the second time in your podcast And it is great that you listeners obviously are interested In AI and the impact of AI on culture On thinking about our brains.

00:02:04: That's really great.

00:02:05: Thank You!

00:02:06: Thank YOU for being here today.

00:02:08: For those who have not listened To this specific episode

00:02:12: Our

00:02:13: guest today Dr Stefan Ruppert wears A lot of hats.

00:02:16: He is an intercultural trainer who moves between Germany and China for a living, teaching business communication in Hamburg one week.

00:02:25: And culture studies in Shanghai the next week!

00:02:29: When he's not decoding culture... He is decoding AI through his brains and bot slab, helping trainers and teams actually collaborate with the machines instead of fearing them.

00:02:43: So Stefan welcome back again.

00:02:46: let's get into it!

00:02:47: Yes

00:02:49: Stefan you've been thinking deeply about

00:02:51: AI

00:02:52: and cognition.

00:02:53: You describe

00:02:54: AI as changing cognitive processes slowly and without us even noticing.

00:03:01: Why

00:03:02: does the invisible nature of this shift makes it so much more dangerous than say, the rival of smartphones?

00:03:11: Yeah.

00:03:13: What is your thought on that?

00:03:14: yeah

00:03:15: I would like to start with three short disclaimers before we discuss about The Dangerous Sides Of AI.

00:03:21: So the first disclaimer Is actually i'm a big fan of AI.

00:03:25: It might sound Like i am super skeptical and think We should turn This all off.

00:03:29: That not the case.

00:03:31: I'm training people on AI, i am using AI.

00:03:34: I like AI and think it has a big potential.

00:03:38: The second disclaimer is of course I will polarise in this podcast also my LinkedIn post but its not that human cognition was always fantastic.

00:03:49: when humans thinks it's great the moment we turn on AI everything is bad.

00:03:57: so it isn't black-and-white thinking.

00:03:59: it's more a step-by-step unconscious regression.

00:04:03: And the third thing is I will also quote some studies, um...I think nobody has a final judgment on where this will lead us and what this would do with you in combination.

00:04:14: we see the first studies now in twenty twenty five and twenty twenty six ...and people see that there are dangerous effects which should really worry us.

00:04:24: But there's no final judgment and people use AI in a million different ways.

00:04:30: So it's not that this is the field where we have a final conclusion, um And I think the dangerous part is as you know We still see AI As a productivity tool?

00:04:42: We talk a lot about tools and prompts and stuff.

00:04:45: Um but AI Is actually technology with which changes or which impacts our cognition The way we think, the way we innovate.

00:05:01: We talk to each other in a way that we interact and this happens unconsciously.

00:05:07: so it's not suddenly you realize everything is different but its step by step regression into the middle.

00:05:15: I think these are the biggest risks which we don't realise.

00:05:18: after working for awhile with AI People cannot even tell what is my idea.

00:05:24: or was this the AI ideas?

00:05:25: Yeah, it's true.

00:05:26: Absolutely!

00:05:26: Good point exactly.

00:05:30: So I think if you are aware of whats happening and where what A.I does It not as risky As If You Think This Is My Own Thinking This Is Our Own Reasoning.

00:05:42: But Actually A.i Had An Impact.

00:05:46: Where do you think is the line for, I mean between AI as a tool?

00:05:50: You just mentioned it that actually extends our thinking and working.

00:05:56: And AI

00:05:57: has a substitute for it.

00:05:58: so how do we know when to have crossed that line?

00:06:05: Yeah i think its bit of line between cognitive offloading and cognitive surrender.

00:06:11: So these are two terms which used in certain way I think you also mentioned it with a navigation system, was the calculation we do cognitive offloading.

00:06:21: Yeah?

00:06:21: We say um... It's a burden.

00:06:24: uh..it is big effort to calculate something.

00:06:26: so i have my calculator on my smartphone and just use this.

00:06:30: that's cognitive of loading.

00:06:32: And other thing is really cognitive surrender where they are saying I take everything AI gives me.

00:06:38: I don't question it alot but even in these cognitive off-loadings don't train the muscle of critical thinking anymore in a way we did it before.

00:06:52: And that's, I think, the difference to a calculator.

00:06:55: because yes... Of course!

00:06:56: We lost the ability to calculate in our head.

00:07:00: young people cannot do this any more.

00:07:02: i still learned this in school but younger people can not do it.

00:07:05: But now?

00:07:07: ...we offload the burden of critical-thinking.

00:07:11: We develop from thinkers to selectors.

00:07:16: So we offload this burden of thinking, of researching or questioning things.

00:07:25: Of building our own opinion and instead We do something which is sometimes called reactive writing?

00:07:32: Just select the AI gives us different outputs And then put pieces together so more manage output.

00:07:44: The research shows that if we don't train this muscle of critical thinking, What do I think?

00:07:53: or

00:08:09: what we should actually ask ourselves is, Do i agree with AI and what it's proposed?

00:08:15: So this is what you're proposing here.

00:08:17: And all of us are going forward to think about if they agree on the proposal.

00:08:24: Studies show that there's a study from Google DeepMind which shows for example the autocomplete function which is in my eyes a very mild way of using AI.

00:08:36: It's not that we choose things where we hundred percent agree, it's enough that we don't disagree.

00:08:45: so we chose the next word.

00:08:47: if we say yeah... That's more or less okay!

00:08:52: And with this step by step We have regression towards opinion and our own opinion.

00:08:59: Be honest.

00:08:59: Is there a skill you have already felt is slipping

00:09:06: somehow?

00:09:07: Yeah, I mean...I also feel that this critical thinking skills- I've got PhD in Chemistry so my whole university career was about training Critical Thinking.

00:09:16: It wasn't about Training Chemistry it was about Critical Thinking.

00:09:20: and of course i realize sometimes I'm losing the skill And I consciously choose to do something without AI.

00:09:30: Um, because I feel i'm getting worse at that also creativity.

00:09:36: Yeah this starting with a blank page and having these creative struggle to say what are the first words?

00:09:44: Are right on my blank page.

00:09:46: That is disappearing Because it's so convenient To just Say AI can Just give me A First Proposal And Then I Work With That.

00:09:55: But every time you do this, you narrow your thinking space to the proposals of AI.

00:10:03: And I realize...I mean it's like here.

00:10:07: i'm living in the fourth floor and know that I should take stairs ...to train my muscles but we have an elevator!

00:10:15: Everytime when I stand on the elevators, they're like oh no I shouldn't be doing these, I need to train my muscle.

00:10:20: Then under time pressure say don't do them.

00:10:26: It takes

00:10:26: a lot of discipline,

00:10:27: right?

00:10:29: Yeah it does.

00:10:29: And I can imagine people working in companies where you have a lot pressure and your boss who says why did they take two days to come up with this presentation – your colleague could do that within half an hour!

00:10:41: Then there is also a lot more pressure and incentive for me saying hey I just need to do things much quicker….

00:10:48: Why should i think

00:10:49: so?!

00:10:50: Where's the incentive for ME?

00:10:53: to think by myself and take more time.

00:10:56: But that's in schools, higher education as

00:10:59: well.

00:11:00: so we all have

00:11:00: these challenges

00:11:03: right now.

00:11:03: Yeah there was a new study from China just came out few days ago which clearly showed if they do their homework with AI.

00:11:10: I mean i saw it in action.

00:11:12: That is the dystopia of learning.

00:11:15: They do home work faster but don't remember what did.

00:11:21: Yeah, so in the exam they couldn't recall anything.

00:11:25: And I think that's also a big risk for companies if people just manage output and don't remember what.

00:11:35: such a good point that just copy paste basically.

00:11:37: Yeah, so it's actually really like humans are losing skills because of YouTube AI and I'm thinking more about now our international audience.

00:11:45: you know people lived internationally often learn to navigate actually uncertainty and complexity.

00:11:53: these skills actually have become very important for these internationals or globally mobile people and digital nomads.

00:12:01: So is Will these skills become even more important in an AI-driven world now?

00:12:07: What's your thought on that.

00:12:10: Yeah, I think so!

00:12:11: After a while people will recognize and they're just starting to recognize... ...that we are homogenizing our thinking.. ..that there is regression into the middle.

00:12:22: all of texts sound the same There isn't clear opinion.

00:12:26: And with this i hope at least That people will go back and say, hey we need this critical thinkers.

00:12:32: We meet these difficult colleague now.

00:12:36: I think you also have these people where you know everybody in the room agrees.

00:12:40: And then there's is one difficult person who always says no i see it completely different than any comes up with his weird ideas or G but he's other People we need right?

00:12:49: Now I Have a good friend.

00:12:50: He's exactly like that and uh...that why I Like to talk to him because I know he would challenge me.

00:12:56: he Will challenge my assumptions he will disagree And it's really talking to him is difficult.

00:13:02: That's what we need, We NEED DIFFICULT PEOPLE!

00:13:05: But I mean you can even

00:13:06: tell AI to challenge

00:13:08: you right?

00:13:10: You could have that as well if you want too.

00:13:12: They're not always like pleasant to you or

00:13:16: trying... This makes a little bit better.

00:13:19: but i think It's a risk that people think.

00:13:23: with the good prompt, I can completely fix.

00:13:25: The problem because AI is programmed in a way and it's trained In a way.

00:13:30: it wants to please you.

00:13:31: And even if you tell it Please be critical challenge my assumptions?

00:13:35: It will do it in A Way Who in the end please?

00:13:38: You So With a Good Prompt and was a Good Setup and Was a set of Agents with Different Opinions.

00:13:45: i Think You Can You Can Improve it But you cannot really fix the fundamental problem.

00:13:52: It reminds me a little bit in the beginning when people realized that AI is hallucinating, I mean it's wrong word but its making up facts.

00:14:03: People said just put into your prompt the disclaimer please don't hallucinate.

00:14:10: Yeah,

00:14:10: that doesn't work because I don't know it's hallucinating.

00:14:14: It's not that you're doing its doing it consciously.

00:14:16: so of course with a good prompt and was the good setup in a set off agents an everything do can improve this a lot but he doesn't fix the fundamental problem that AI always has his regression to the middle.

00:14:35: So i have question more like

00:14:38: on the

00:14:38: AI systems.

00:14:39: So is it true that they are more trained predominantly on Western data, correct?

00:14:45: That's correct like worldwide.

00:14:46: so just now

00:14:48: while traveling

00:14:49: you know people all around the world communicate differently.

00:14:54: yeah

00:14:57: those communication norms and ways of thinking the western way.

00:15:03: how about when actually travel to different countries?

00:15:06: recently, I've just been to Japan.

00:15:09: There is a very different thinking in terms of culture as well

00:15:13: or how

00:15:13: would you reply to an email?

00:15:15: Or

00:15:17: question and not straightforward like maybe Western countries.

00:15:21: so what's your take on this?

00:15:23: when AI was used global base What are the strengths and weaknesses for that?

00:15:31: Yeah, I mean first of all yes.

00:15:32: It's absolutely true.

00:15:33: most of the models we use here are trained on Western data or American data.

00:15:38: some people call it weird data western educated industrialized rich democratic a term coined by Joseph Henrych.

00:15:47: So this is the dataset.

00:15:48: and there are studies which also show that The thinking off AI?

00:15:54: We know it doesn't think but let's humanize.

00:15:58: The thinking of AI is very Western.

00:16:00: It's very individualistic, it's very direct and I talked to a lot people who said the advice that they get from AI for me was always too direct.

00:16:09: so i would never give advice in this way.

00:16:13: So yes...it has these bias.

00:16:16: And the big question is Would models trained on different data sets have a different culture?

00:16:27: And in a lot of regions, people are trying to set up AI systems like in Latin America and India.

00:16:34: We have different system in China.

00:16:37: I didn't find so many studies about this but some studies come to the conclusion first off all still these Chinese models are more Western than we think which seems an indication that also they're trained predominantly on western data sets.

00:16:55: I mean, some people even said maybe deep seek is just the kind of loan off chat to BT and it was extracted from chat to be tea.

00:17:04: And that goes into their direction.

00:17:06: there are other studies which show even between the Chinese models as a big difference in.

00:17:10: some of these models seem too think more Chinese or more collectivistic or more indirect.

00:17:18: uh... The studies clearly show there's a regression towards Western thinking.

00:17:25: There is one study I remember where they gave Indian participants Chatcha

00:17:30: B.T.,

00:17:32: or similar language model, and after awhile they lost their original language in the original thinking.

00:17:40: And it was not just a question of style Their writing styles changed but also their logic, they're thinking style changed more and more to a typical American thinking.

00:17:53: A colleague posted on LinkedIn she's active in the American-German interaction and says when German asked Chachibiti to translate an email into English for an American business partner it doesn't just translate.

00:18:10: Yeah, interesting.

00:18:12: It comes

00:18:13: a little bit more polite if it becomes... A little bit less indirect.

00:18:21: And there are always these things like how you're doing and which Germans would not write?

00:18:27: Germans will just say hey Tom that's the problem with this project because they realize sometimes The way of persuading people, the style of the mail changed and in the end it sounded a bit inappropriate for German.

00:18:45: And this is again case you don't realize.

00:18:48: immediately read email.

00:18:51: that's fair translation what I wanted to say.

00:18:53: but if you read twice no thats not exactly how i want.

00:18:59: I was angry when i was writing this email and suddenly it's sugar coated.

00:19:06: What do you want to write?

00:19:11: This has a big potential, of course there are situations where would say my style is not appropriate for the American Chinese business partners but they're other cases that will keep me original voice.

00:19:29: Yeah, because it's important that on the other side people also recognize my emotion.

00:19:34: My frustration and anger whatever.

00:19:37: And if I use an AI tool to smoothen all this... ...I think in the end communication is not really human

00:19:45: anymore Exactly!

00:19:46: And a human gets replaceable basically as well going forward for future jobs or anything like that depending which areas.

00:19:54: But at the same time, it might be also a benefit

00:19:57: I think.

00:19:58: So yeah...I'm

00:20:00: not quite sure so but stay a little bit because i think that's fascinating like this convergence or diversity trap as well Because you've also said that the impact on organizations and society is what concerns you most right?

00:20:16: Because companies have invested enormously in cognitive diversity, let's say different life experiences.

00:20:23: Different cultural backgrounds, different ways of thinking.

00:20:28: and you argue AI actually erases all that.

00:20:33: can you unpack a little bit?

00:20:34: why do think I might be like this silent equalizer?

00:20:40: Yeah so thats again my polarization.

00:20:43: i mean of course AI is not erasing.

00:20:47: There's a tendency and these studies show this very clearly that I mean we all learned.

00:20:54: in order to be innovative, good at problem solving.

00:20:58: Creative it makes sense to have a diverse group because you need diversity.

00:21:04: so come up with ideas also solve complex problems.

00:21:08: what the AI does is on an individual level really interesting Because that's the innovation paradox.

00:21:16: on an individual level, your ideas get better.

00:21:19: Yes!

00:21:20: So if you brainstorm alone at your desk and do it with AI... You will have more or less better idea.

00:21:27: But in a group-level suddenly realize That most people come up with same ideas which are in the same frame And then you lose diversity.

00:21:42: Yeah, so you could summarize it if your really interested in a diverse workforce.

00:21:47: Don't invest in diversity and inclusion programs but first turn off chatcha bt because now You have this fantastic diverse work force with people from different areas of the world With different life experience And you give them chat to BT can they all come up was very similar ideas?

00:22:05: In this experiment.

00:22:07: that was a creative experiment where people were asked to invent a new toy.

00:22:12: A lot of people even gave their new toy the same name, which also built a breeze castle.

00:22:22: and here again The interesting thing is everybody thinks I was super innovative!

00:22:28: i had this disruptive this great idea build a breeze Castle.

00:22:33: That's a fantastic idea.

00:22:35: And then they sit together And I say, OK.

00:22:38: Ten of us have built a breezecastle.

00:22:40: that's all the same.

00:22:43: and here it is so important to first turn AI off then ask people do their creative thinking without AI?

00:22:52: Then in later stage you can work with AI maybe even mirror this saying these are our ideas or we easily see which ideas really original.

00:23:03: but AIs again is homogenizing it, its streamlining the thinking to be predictable.

00:23:11: and what's missing in this unpredictable.

00:23:14: This crazy stuff!

00:23:16: In this study a lot of people had ideas for toys which didn't work where everybody said that's stupid idea.

00:23:23: That's exactly you need in a brainstorm.

00:23:26: Interesting

00:23:31: Maybe, yeah.

00:23:33: I'm not sure

00:23:33: if i agree.

00:23:35: Great!

00:23:37: That's good.

00:23:39: Thanks to a colleague of mine.

00:23:40: actually he came up with an interesting example recently and it is an example from mathematics.

00:23:47: so please repeat everything.

00:23:50: We always say that AI is straight on every thing humans have already written.

00:23:57: So...that might lead to the fact sometimes become like, as you just mentioned the most conventional thinker in the room.

00:24:05: Just repeating the consensus back to us.

00:24:09: but sometimes it actually does the opposite as well because It came up with a different solution.

00:24:15: and this specific Solution to that problem I mean eighty years.

00:24:20: people solve it In the same way.

00:24:22: And then all of A sudden AI came into The game.

00:24:28: Like, the essence of this story is like AI does not make everything alike or it doesn't think alike.

00:24:35: Of course I mean... It does not produce geniuses on demand for sure but i think its more interesting that also can actually discover something new.

00:24:46: nobody

00:24:47: has thought before because we are sometimes so in our beliefs.

00:24:55: For example, these metametricians believe that the best way was something like arranging dots in a giant square grid.

00:25:04: Like graph paper.

00:25:06: and then they have not thought about another field of expertise for example.

00:25:11: or if we think more interdisciplinary actually get solutions.

00:25:17: so people never thought about it.

00:25:18: all of sudden

00:25:19: AI

00:25:21: linked the dots So that might also happen.

00:25:26: Yeah, absolutely!

00:25:29: That's what I said in the beginning.

00:25:30: of course it is not that AI is erasing our creativity completely and there are these cases... ...I know this study where similar cases were in the end turned out.

00:25:45: AI was just good at finding a solution which nobody knew.

00:25:50: so i don't it really took a pattern from different fields and mathematics which is also kind of creativity based on patterns.

00:26:04: So I would not say that the eye is always wrong, and its always hindering creativity.

00:26:12: but the tendency is we lose this human experience with different experiences polish everything and have this regression to the middle.

00:26:26: And that's also why brands are really worried, because they see after a while we will all sound the same?

00:26:33: Yes

00:26:33: exactly!

00:26:35: We'll lose these originality... ...because we all use the same tools.. ..we think we're super innovative cause' we've got better prompt than

00:26:44: others.

00:26:45: but of course I mean, they are as smart as you and have the same font.

00:26:50: They come up with very similar stuff... The life lab for this is linked in.

00:26:57: I mean, you see it already because their people are really super lazy and people are annoyed Because they say everything sounds the same.

00:27:06: It's a good point.

00:27:07: actually People need to realize as well.

00:27:09: Some people don't realize because They don't read other posts maybe on LinkedIn.

00:27:13: So that's a great point.

00:27:14: Critical thinking Realization Aswell And then Actually action For every individual.

00:27:22: My metaphor for this is just imagine you're driving on a highway, You use your navigation system and see there's traffic jam.

00:27:33: And then suddenly Google Maps suggests to different route which goes through some... I never

00:27:41: do that again!

00:27:44: But at the moment think you are a genius.

00:27:47: For a moment you think I have the solution.

00:27:51: It saves me half an hour, i'm so fantastic!

00:27:56: You know that your algorithm in Google Maps is same as for all others and then look at people who stay on the highway... That's

00:28:04: great point.

00:28:05: thankyou for sharing.

00:28:06: And then half-an-hour later you stand with other smart people because of course google suggested This route would have made sense for you, but it doesn't make sense if you send hundreds of people in this direction.

00:28:23: Let's

00:28:26: talk about the human connection.

00:28:28: at the age of AI like can authenticity be automated?

00:28:33: For example our listeners we all know.

00:28:36: when you travel or go to a restaurant and increase automation check in might be digital for Hotel of Light, but then people still... or want to feel warmth, attentiveness and personal service right?

00:28:56: So is there actually still someone personally help the customer find a way in

00:29:05: what aspects

00:29:06: of human connection cannot

00:29:07: be

00:29:08: replicated by AI.

00:29:09: And how do you see this whole scenario?

00:29:13: could convenience come at the expense?

00:29:17: AI is fantastic at simulating empathy, of course.

00:29:20: Yeah?

00:29:20: We all know this and that's also why people use AI most.

00:29:24: so in all the statistics how do people use

00:29:28: A.I.?

00:29:28: And what are the use cases?

00:29:30: The top-use case is coaching help with life challenges and stuff like that.

00:29:36: So that's when people go to AI because they get good advice.

00:29:39: Other studies show especially in coaching it makes a difference if you know It's an AI which has no cost and no effort, or if you know there's a human.

00:29:51: And the humans spend some effort... ...and the human is concentrating on me and invests

00:29:55: sometime.".

00:29:56: That makes a difference for us!

00:30:00: The biggest risk I see in that similar to social media.. ..is typically when we interact with other humans it always comes with a cost.

00:30:13: So this is the Social Exchange Theory.

00:30:15: Um, sociologists say whenever we interact with someone it has a benefit and it has cost.

00:30:22: So we invest something because We have to open up the other person.

00:30:27: The other person might judge out or disagree with us Or not understand what they are saying And the others may expect something in exchange.

00:30:38: So if I tell you about my relationship problems and you listen to me, You expect that.

00:30:43: then I listened to your problem.

00:30:46: So there's always a price for human interaction And AI removes this prize.

00:30:53: The AIs are free.

00:30:55: There is no psychological price i have to pay.

00:30:59: and studies show That the same as with social media some years before After interacting with AI People are less satisfied with human interaction because they say it was so easy.

00:31:14: It's helpful, the AI listened to me and gave good advice.

00:31:22: Now I talk my friend and suddenly everything is complicated.

00:31:28: My friend disagrees but does not always prove what i said.

00:31:33: And I see a big risk here because the same happened on social media with this instant reward.

00:31:38: If you post something in Social Media, You get a response instantly and feel great!

00:31:45: This morning i posted something linked-in after five minutes at first like... Fantastic!

00:31:50: Instant

00:31:51: Reward!!!

00:31:52: Compared to that human interaction is a burden.

00:31:58: it needs an effort.

00:32:00: studies show that.

00:32:01: That's the reason why people prefer to post online instead of talking to their friends because they get these instant rewards.

00:32:08: and I think with AI, it could be even worse... ...that people avoid human connections or they judge human connections in a different way?

00:32:19: Because suddenly everything is so effortful.

00:32:22: Maybe

00:32:24: about conflict as well.

00:32:25: you mean right like

00:32:25: for human interaction in general.

00:32:30: yeah If I tell you more about my personal problems, You will somehow judge me.

00:32:34: Yeah?

00:32:34: Even as a coach in certain way humans always judge.

00:32:38: You might criticize me if i talk to My difficult friend About my relationship Problems.

00:32:44: he might say Stefan I guess your completely wrong.

00:32:47: Now Your girlfriend is fantastic and you see it wrong!

00:32:50: You made a mistake.

00:32:51: That's not what I want to hear.

00:32:54: So I would rather turn To the AI.

00:32:56: And The AI Will Always Say yeah poor guy and you have been treated very bad.

00:33:03: And I give your solution, there has been examples for example where people in the end got divorced because the AI confirmed their view.

00:33:14: that's a right decision.

00:33:15: they were convinced by

00:33:19: AI.

00:33:21: A good friend might've said sorry maybe you're a little bit too emotional right now?

00:33:26: Maybe we should see it from different perspective.

00:33:29: Maybe you should have a long-term perspective and then you start thinking about your view.

00:33:38: And I can completely resonate with this, i also discuss really heavy life problems with AI... ...and realize it's less effort than calling friend or talking to friends.

00:33:53: It's the easier way.

00:33:54: And if I don't like the answer from chat to BT, are just close to BT and open Gemini?

00:33:58: Yeah that is true!

00:34:03: If i open up a friend and I do not agree with what this friend says to me... ...I cannot turn him off or say okay forget it we never talked about this.

00:34:15: So you think everything he mentioned has an impact on our behavior as well But what's then the consequence?

00:34:25: So, What should we consciously protect as AI becomes As you just shared like part of everyday

00:34:33: life.

00:34:36: Yeah I think We should protect this plurality This diversity.

00:34:41: i Think this is really super important to keep this diversity Of views and To be aware that AI is pulling us towards a statistical average of human or American thinking.

00:34:57: And the other thing, it's really moral judgment because AI doesn't have a morale now?

00:35:05: It cannot make a moral judgement and you can say I think this right in an ethical way.

00:35:19: And then this ability to challenge the patterns.

00:35:25: Because that's what moves societies forward, there are people who say yes we always did it in this way but I challenged them and have a vision of different world.

00:35:37: See for example things like racism gender equality If you look thirty years back our society had completely different view on that.

00:35:46: There were people who said, yes but I don't agree to this anymore.

00:35:55: Yeah, and human qualities like the human

00:36:18: qualities.

00:36:20: We all appreciate it really Which are even becoming more and more important as you mentioned in the AI driven world?

00:36:28: yeah And also this being difficult creating friction.

00:36:34: We need people who create friction would disagree and we're disruptive and I will not necessarily do this job.

00:36:42: of course if your prompted in a certain way I can also somehow take this role But I think it's not a substitute for these people who just say no.

00:36:52: You all agree on this, i see it completely different and try something completely new And everybody thinks that will fail.

00:37:00: but i tried anyway.

00:37:01: Do you think our society is thinking specifically about educational system actually appreciates the value of those people?

00:37:10: Are we better than AI or

00:37:17: society?

00:37:23: I think it's a great question because i also worked for twenty three years in the big corporation and these people always had a difficult time.

00:37:32: It was this non-conformists that were always the people who got not promoted, everybody said you're so difficult but they are still there!

00:37:44: So some of these people were still there and the company also made sure we keep some of this.

00:37:53: And I don't know if we lose.

00:37:55: in the age when everybody agrees, AI is always confirming your views instead And maybe in the age of AI, people start to value this even more.

00:38:14: So I talked a lot of people recently about these tendencies and about these impacts on cognition and human communication.

00:38:22: At least people listen and they are interested and say that it's really time now for us talk.

00:38:32: what does our company do?

00:38:34: Especially young people who joined the company grew up.

00:38:40: I think they didn't grow up with AI, but they spent their professional life and time at the university.

00:38:48: They don't bring these skills or critical thinking... ...and more people are concerned about them!

00:38:55: But you know what?

00:38:55: I just realized in a recent workshop where teams

00:39:00: had to

00:39:00: present their findings.

00:39:02: There were five out of those teams that really have authentic, I would say original thoughts and you could tell.

00:39:12: And then there was one team as you can immediately tell okay this is AI generated in terms of the structure...in terms of pictures how they presented it?

00:39:23: Actually everyone realized that right.

00:39:28: so that's obvious!

00:39:30: The other teams actually complained about this one team.

00:39:34: So I think

00:39:34: there is like, this

00:39:35: awareness.

00:39:36: it's growing actually and people are a little bit not offended by.

00:39:40: they told me yeah but i mean This obviously AI generated.

00:39:43: so we put all the efforts in.

00:39:47: We did some research And had brainstorming and discussions going on.

00:39:52: Also students know that.

00:39:54: Yeah also student aware.

00:39:56: It's Not That The Young Generation I mean, they're not stupid.

00:40:00: They see the impact and if i do my homework with AI... ...I don't learn anything!

00:40:07: They are well aware of this.

00:40:09: The question is how we deal with it as a society?

00:40:12: Because I think that awareness there also.

00:40:15: young people say like me.. ..I should take stairs but instead elevator.

00:40:21: So How to create space where people say I train my critical thinking.

00:40:27: I trained this muscle instead of saying, hey actually now you can do more homework than before that's fantastic.

00:40:34: so i give you more tasks

00:40:35: a

00:40:36: different task.

00:40:40: So it is something we have to fix on our society or company level.

00:40:47: To say how do we deal with these?

00:40:49: because for the individual its not very attractive.

00:40:54: That doesn't make any sense.

00:40:56: If I train my critical thinking muscle, the others will outperform on a societal level.

00:41:03: and if you talk to students then please let's turn off your computers.

00:41:07: Let now practice critical thinking without AI.

00:41:11: they agree with this.

00:41:13: They know it is useful.

00:41:19: Yeah that what i think The awareness

00:41:32: actually exists.

00:41:33: But can you imagine in a company saying I want your ideas on a new strategy,

00:41:42: please

00:41:43: do it without AI.

00:41:44: It's a

00:41:44: different exercise, if you already say it as is at the beginning.

00:41:48: It's an exercise for in the company or students are wherever

00:41:54: and The first step is always the awareness.

00:41:56: That's why I also like to talk about this because i think its important that people Think about it.

00:42:01: they're aware They realize Im more creative with AI but we lose creativity We loose diversity We loose experience and then think about the next step.

00:42:14: What do we do with that?

00:42:16: And what are the tasks?

00:42:18: where AI is fantastic, I mean for ninety percent of this stuff maybe AIs fantastic but what other task were I say here?

00:42:27: AI really this homogenization it's not what i want.

00:42:31: That means you need a good leader within the cooperation companies who actually encouraging individuals to back from it.

00:42:42: Yeah, and you have to establish a certain culture.

00:42:46: I'm saying this is our company culture.

00:42:50: now it's really about our point of difference.

00:42:52: And here we need this human experience This guy from India single mother with three children.

00:43:01: We need all these perspectives in the product development And we don't want to smoothen everything with AI and then come up the same solution as everybody else.

00:43:13: The word increasingly is shaped by AI, Stefan!

00:43:19: Looking from today going forward like in future what gives you hope?

00:43:24: What's your excitement or change?

00:43:27: First of all I think that AI can do fantastic things help us in drug discovery.

00:43:34: In a lot of technologies, I mean you mentioned mathematics?

00:43:38: There are lots areas where AI can improve our life.

00:43:41: that's the first thing which gives hope and other things is people are realising their negative effects.

00:43:50: it gives me hope they will take action.

00:43:54: You see this on LinkedIn.

00:43:55: everybody is annoyed about what happens.

00:43:59: People also, when I talk to them they say yeah.

00:44:01: Yeah i saw it in my company and some presentations that were all very similar but don't want this anymore.

00:44:08: so there will be a swing back asking exactly the same question you are asked on a society level what can we do to stay human?

00:44:25: discussion here about this politician who held a speech in Auschwitz, very empathic speech and it turned out that was all written by AI.

00:44:35: And people say that's not okay or at least we had the discussion.

00:44:38: is it ok?

00:44:39: Is it not OK?

00:44:40: People realized as well.

00:44:42: so thats bad thing.

00:44:43: they realize its an AI.

00:44:44: you

00:44:45: know Yeah, and if someone holds the speech and says wow this was a very emotional moment for me.

00:44:51: And then you realize that what's all made up by AI there wasn't no emotional moments You question people authenticity.

00:44:59: I think it is happening now.

00:45:01: This exactly discussion we need right know i don't have final answer but also feel If somebody gives such an emotional and sensitive moment It's wrong to do it with AI.

00:45:16: There were other cases where journalists did opinion pieces, and I was like okay if you tell me that is your opinion i want to hear your opinion And maybe the A.I.

00:45:30: Opinion was good but interesting.

00:45:33: But I wanted to know Was this your opinion or an AI?

00:45:37: That makes a difference for me somehow.

00:45:41: At the same time, I think sometimes we might judge somebody who's really a talented speaker for example

00:45:49: or is quite

00:45:51: gifted in terms of finding the right words whatever it is.

00:45:54: So and immediately assume so that must be AI That not human rights.

00:45:59: If you're thinking about these

00:46:02: AI detectors For

00:46:03: example i mean We have them at uni as well for sure.

00:46:08: My point is sometimes I'm not quite sure if people who are actually gifted in terms of being able to have give like a wonderful speech, but everyone assumes that this isn't human.

00:46:20: It's just.

00:46:21: it can be true.

00:46:23: you punish the ones or maybe quite excellent.

00:46:28: for me This happens with images.

00:46:30: whenever i see an image was something which Yeah,

00:46:34: that

00:46:34: is all this movie where the weather in Brazil.

00:46:37: The lady was pushed from the cliff and they forget to fix it.

00:46:42: I thought there's

00:46:44: exactly yeah

00:46:46: so sure.

00:46:47: And then i did some research and i realized wow This really happened here.

00:46:50: It was a real accident and someone died In this accident.

00:46:53: but i couldn't believe this because i see So many ai created videos.

00:46:58: I don't trust anything anymore.

00:47:01: Are

00:47:02: we using our trust?

00:47:04: that might be another thing

00:47:07: to discuss?

00:47:09: Second question, they're actually doubting it right there and questioning is this a real image the speech or AI?

00:47:18: people are looking twice now.

00:47:19: you look at the corners edges of the AI generated.

00:47:22: so yeah what has actually developed the last couple of months and years.

00:47:29: I

00:47:29: think that's also a risk, with every speech was ever linked in post instead of listening to what people say?

00:47:36: I try to find out is it integrated then M-dash init or if there are certain structure in it which is also not good?

00:47:45: i don't know how much EU regulation helps because beginning of August we will have this regulation.

00:47:53: At least for images, there needs to be a mark.

00:47:56: And now they also published the official symbol of this and text which has not been changed by human.

00:48:07: There must have been a mark on it... I don't know if that really helps?

00:48:13: It is more transparent maybe as well!

00:48:15: It helps people quickly understand what's actually shown.

00:48:22: But it also shows me, It makes a difference for people.

00:48:26: They are not just say I don't care if the text is good or bad but that's what it means to say.

00:48:33: Is this really authentic?

00:48:35: Or is AI generated and you can present me AI-generated texts but i want to

00:48:42: know they don't accept them as easily anymore.

00:48:45: So its okay with me when reading a piece on Spiegel online.

00:48:52: But if someone says this is an opinion piece and then I realized that was not your opinion, it's just the statistical average of all opinions in the world.

00:49:03: That ticks me off somehow.

00:49:06: Stefan what are you working on at the moment?

00:49:09: What's in the pipeline for you?

00:49:17: Yeah, for me I really mean...I'm not an IT person.

00:49:20: So i look at this whole topic of AI and the develops from a outside perspective which is important.

00:49:27: we cannot leave it to the IT experts.

00:49:30: We need people in different areas Interculturalists, sociologists People form all different areas To look at that And currently trying Get into contact or get in to discussions with companies about that because I think it's a time now too.

00:49:49: Go step back and say, hey we all work with AI know for two years We introduce them our company is we are worked with co-pilot of chat GBT And maybe now It's the time for the leadership team To look at these other aspects and to discuss.

00:50:05: do you see this already?

00:50:06: How relevant is it for us?

00:50:08: what did we do?

00:50:09: was it?

00:50:10: So that's what I'm preparing now, really doing workshops with companies to discuss about.

00:50:16: And i see a certain openness and the question is when will companies be ready?

00:50:22: To pay for workshop?

00:50:25: invest some time do this.

00:50:28: but people are responding too.

00:50:30: they say yeah we already seen it.

00:50:33: you don't have to convince us all these studies.

00:50:36: We saw this already in our last innovation workshop and we see these all ready, you know how strategic proposals an everything.

00:50:45: And if anyone wants to reach out what's the best way?

00:50:48: contact here?

00:50:49: LinkedIn.

00:50:50: I'm on Linkedin and i just posted a series of nine posts about exactly what were talking about.

00:50:57: so people want read it and see fancy AI generated infographics.

00:51:07: Yeah, yeah.

00:51:07: Of course I mean that's funny!

00:51:09: I talk about this and when i post on LinkedIn... ...I use AI generated infographics which were done with notebook lm.

00:51:17: So.. ..i also had to laugh when i started with the series because i realized how bizarre it is to say ''i use AI'' , ''i talk about the dangers''.

00:51:28: And still i used it to create the infographics.

00:51:33: But i think that mirrors probably how everyone is navigating these

00:51:37: years

00:51:38: or times right now.

00:51:39: Right?

00:51:40: Yeah, and at least I would of course tell notebook LM this is my opinion about the topic.

00:51:46: This Is What i Want to Say And Then I Would Check Is My Opinion Really Reflected in The Infographic Or Did It Smoothen it?

00:51:54: For Example for the Diversity Topic yeah I Was Very Very Sharp on This Diversity Topics To say We Lose Diversity if you are interested in diversity, which is of course a polarized statement.

00:52:09: It's a hook and every time I wanted to write the LinkedIn post with it or blog posts A.I took this out little bit And i said no that was my statement.

00:52:24: So always had to bring again.

00:52:26: So LinkedIn, I think it's a best way.

00:52:28: Also have my website.

00:52:29: you can put in the show

00:52:31: notes.

00:52:31: Thank You Yes In The Show Notes i Can also Send You If You Want to Send Some Of The Studies If The Listeners Really want To Invest The Time And The Struggle And The Critical Thinking To Read The Studies Your Own.

00:52:48: Of Course You Can Do This.

00:52:49: It'S A Fascinating Topic.

00:52:50: For Sure If you could give our community, or each individual of our community I have to say now one

00:53:00: rule.

00:53:01: One

00:53:01: line to put on their wall in the journals about using AI without losing themselves into process what would it be?

00:53:14: I will give them a quote and i'll read because It's a quote that just borrowed from someone else.

00:53:22: So Dave Boyce, and I don't know if he invented it or just re-quoted.

00:53:26: He said automate the predictable so you can humanize the exceptional.

00:53:35: It resonated with me.

00:53:36: Automate the predictable to humanise the exceptional?

00:53:41: If we use AI in that way... We could really use a potential.

00:53:48: keep our humanness, our human experience and our plurality.

00:53:52: Our ability to challenge patterns instead of just working with these patterns forever – our moral judgment or ethical compass for everything!

00:54:04: For me this is a guideline which makes a lot sense.

00:54:14: Thank you so

00:54:16: much

00:54:23: for sharing.