Brain Food for Pioneering Spirits: All of recorded human history up to 2014 versus one cup of coffee’s worth of data today — what does that scale actually mean for opportunity and risk? In Room #881 of the Deep Talk Club, under our overarching theme “Pioneering Spirit with brAIn,” the group digs into comparison-portal data flows, the hierarchy behind who really owns your data, a pregnancy revealed by purchase patterns before the woman herself knew, and why a monopoly isn’t automatically the villain. Listen in on the examples that came up and the perspectives that emerged in dialogue…
Note: This session was recorded live in German as part of the Deep Talk Club, so the video above is German-language audio only. What follows is an English translation of the conversation, lightly edited for readability — the back-and-forth between the moderator and participants is preserved, but the original timestamps and speaker names have been left out.
A wonderful good morning. Today we’re talking about opportunities and risks of big data — what opportunities and risks are becoming apparent in the age of big data? I’m really curious to get into this exchange with you again today, and talk about this topic. We’ve already touched on one thing or another in various rooms — especially this aspect of data protection and so on keeps coming up. And yes, big data is basically about the collection, storage, structuring, and further processing of data. Today we have a whole lot of options to gather information, track things automatically — whether that’s a website we visit, or various videos we might watch on YouTube, and act accordingly. There are algorithms working in the background too, deciding, for example, what more we see, what gets shown to us. And when we go shopping, use cards to pay — so there are a lot, lot of examples where various data gets collected. But the art isn’t just in collecting data, but especially, in inverted commas, in storing and structuring this data. And structuring in particular plays a really essential role here. Because, well, what does it actually mean if I have a name, an email address? It’s primarily about, for example, consumption behavior and similar things that get analyzed here, to make it usable in a certain way. But it’s not just consumption behavior — there’s other information too, all the way to how many properties someone might own, or what their assets are, and so on — assigning certain things to a person, so to speak. Person- and/or company-related. Companies are quite interesting here too, of course. Depending on which field we’re in, you can do a lot, a lot with this data, once it gets structured and used purposefully. And that’s what’s meant by further processing too. And of course we quickly get into an area where, on one hand, the question arises, how transparent do we actually want to be, in inverted commas, what’s this all for? And there’s a high complexity behind this too, again about competitiveness and so on, because ultimately we can’t look at Germany in isolation here either, but need to think and weigh things globally too, regarding developments taking place. And against this backdrop, those are maybe the first impulses for now. I’m really curious what examples you’ll bring in too. So, on one hand, you can pick up what examples actually exist in the context of big data — in what form data gets collected, stored, structured, and further processed. And at the same time, let’s look at the opportunities and risks of the whole thing today too, and get into a discourse, into an exchange, to reflect on the whole thing, to see what insights we can derive for ourselves from this room today. So you’re all very, very warmly invited to join in again today too, get into the exchange, and open up this space for reflection together with me. Just a reminder again, that the room is being recorded for the entire time, and will be published afterward too. So anyone who joins automatically agrees to that. Then we have three room rules to keep in mind — a profile photo where you’re recognizable, first and last name, and at least one sentence in your bio. Anyone who doesn’t want that for whatever reason can of course use the chat too, to bring in thoughts, questions, or similar. Right, with that I’d like to open up the round. Let’s see who’d like to join first. And I already see Marc would like to open the round with me. A wonderful good morning, dear Marc. What are your thoughts and experiences on this topic?
Good morning to the round. Before I share my thoughts with you, I’d like to warmly welcome you all to the workout group with Yasemin, as it felt to me this morning, or brain-shocking. Because with every topic I notice that these topics stir up emotions, and I get to thinking about a topic like this. And then two aspects always come up. One is what you just mentioned, what happens with the data and so on. And the other side is, hey, what could this be good for? So both the negative and the positive. And now I’d like to start with the topic, or the thoughts I had on it — I couldn’t really connect with the word “big data” at first. It’s like “Big Brother.” And to get a better feel for it — maybe you feel the same way — what really astonished me, though it didn’t surprise me, is that if you take the origin of human history up to 2014, so everything ever written and gathered as information, then the amount of big data we generate today, we reach that same amount, from the origin of humanity to 2014, in one cup of coffee’s worth of time. Then you get a sense of what big data actually is, and how much data is in circulation and being collected.
Yes, great, thank you, thank you so much. A nice image, a nice comparison, to get a sense of the scale in the first place. Yes, dear Marc, what does that do to you?
Yes, I find really interesting information there too, and I’d like to say something about that too. In the first thought experiment, Marc answered and thought as Marc. In the second thought experiment, Marc thought as an entrepreneur, and I’d like to start with the positive thought too, and a report I saw on TV yesterday moved me on this too — this doesn’t have anything to do with the topic, but we’ve had this topic before in the Deep Talk Club — from 2010 to 2019, negative words, we’ve had this topic before, so negative reports or negative words increased by 300 percent from 2010 to 2019, because that’s how news sells, and because we all react to negative things. And that’s why I decided to start this morning with the positive things, what that does to me. So, big data — you can, or could, and it’s not just “could be good,” but it’s actually been good — in the US, for example, it helped prevent the spread of a flu wave, because the data showed, oh, people are getting sick here, something’s happening, and then you can react and take preventive measures. The second thing I have is that machines can learn, as an entrepreneur, to be more efficient, more productive, or companies get to know their customers better — even though, personally, as Marc, I find it annoying when some platform offers me, person A or B also bought what you bought. That doesn’t offer me any added value as a person, because I don’t care at all who buys what — I’m only interested in the product itself, you get the benefit there, but who else buys it doesn’t trigger anything for me. It annoys me more than anything. And a fourth aspect, for me, would be health. That’s one of the most important aspects for me — that through the data collected, you might be able to find new paths in cancer therapy in the future.
Wonderful, thank you, thank you so much for the examples. And I find it interesting that you say you’re more annoyed when you notice who else buys something, because that of course triggers certain brain mechanisms, in inverted commas. Of course that also has something to do with scarcity, or with strengthening motivation, maybe to buy certain things too. And interestingly, on Amazon for example, they also show, at the bottom, products that are somehow related, or that got ordered in this context too. And I don’t know who among you knows this, but there’s often something there where you look at it and end up ordering it too — or you might even already know beforehand that if I order this now, I’ll also need a second item, and it’s exactly what Amazon actually suggests. And we’ve had this situation often too, that we’re actually glad Amazon suggests it, because then you don’t need to search much further, but can go directly to that product and order it too. Just as an example. So there are a lot of different ways this gets used, and probably very different too, how we personally handle it in our purchasing behavior too. And then there are situations where we might be annoyed. In other situations we might even be grateful for it. So you can see it’s actually quite different, how that plays out contextually afterward too. All the way to, you certainly know this too, on certain platforms, especially with Google or something, you know this — if you’ve looked at certain things, you basically get followed around by ads afterward. All of that is also mechanisms at play, all the way to these connections — and I find that sometimes quite frightening, in inverted commas, it has advantages and disadvantages too. So as a family, if we use one network at home, for example, and the IP address is identical, you can already follow along, to some extent, what the other person might have looked at, or similar, because on the other computer, in another room, or on another floor, or something, you also get shown certain things the other person looked at. And that sometimes gets a bit, you kind of get the feeling it gets spooky. I don’t know how you all experience that. Marc, how do you see this?
I have to say this now just to polarize a bit — then these big platforms have pretty dumb marketing people, because I know what you mean, but you could phrase it differently. If I book a trip and I get information saying, it makes sense to think about insurance for this trip, oh right, that’s true, they’re right about that. Or I buy myself a razor, and then I get the information, cleaning cartridges are available for this device, and they’re often bought together in that context. That makes sense to me, that offers me added value. What I was getting at was, whether person A or B buys some product has zero added value for me. It only has added value for me if I get a selfish benefit. And I don’t get a benefit because another person buys something, but because it’s an additional product that makes sense in connection with a razor, to have cleaning cartridges, so I can keep it clean. And maybe I’ve overlooked it so far, and as I said, what bugs me is just when people advertise. I have nothing against advertising, if it’s clever and offers me added value.
Yes, I understand what you mean, and at the same time, it does work at that point, otherwise they wouldn’t use it. On one hand, through the name-based attribution, you get this aspect that it seems more credible. Nonetheless, you can use these things without those names actually being true either. So of course you’re always allowed to question something like that too. And at the same time, think about it — you have it on one hand, where you don’t just get shown the name of a buyer from three minutes ago, but at the same time also something like, there’s only so much of this product left, so there’s also an additional scarcity aspect on top — that often has an effect on the purchase too. So you can quite massively — this is basically about influence, or manipulation — quite clearly combine certain knowledge you have with the data being collected. So let’s assume for now this isn’t fake data, combined with the data happening in parallel, and that by synchronizing the data being collected on one hand, and then having it shown to the next person almost simultaneously, you can create leverage effects too, at that point. That’s really exciting. Marc, would you like to add to that?
No, exactly, I can understand that — that advertising works is undeniable. Another point I find really important too, is that we as a society also think about all this data collection, and what happens with it, what opportunities — we started with opportunities — what opportunities and risks that offers for us as a society, and how we, as a society, handle it too. Because the reason I’m saying this is, the data we hand over gets collected. And this data is a danger, in inverted commas, for cyberattacks — we’ve already talked about this topic too. That’ll increase, in the coming years too. And I can only carry out cyberattacks if I have as much data as possible — like birth dates and so on and so forth, because those are personal data.
Yes, and above all that becomes especially interesting for hackers too, because the more data there is, the more valuable it becomes as information, so to speak, for me as a hacker too. And that leads to data theft being carried out accordingly, and sometimes you might not even know what happens with the data. And there’ve been reports in the past about this too, that we can follow — there are various sites, maybe, where you can enter certain things yourself and check whether you’ve been affected by this data theft too, and so on and so forth. And that’s always a challenge too. All the way to us being able to observe that companies with an incredible amount of data see their value skyrocket, without them actually having that many assets. And that’s really, really exciting too. So, following this — in the past, it used to be that you’d have, say, large inventory stock, that the number of employees played a big role, and so on. And now we have the situation where the pure amount, and structured amount, of data you have, suddenly makes up an incredible value too, in terms of the company, so to speak, company valuations. Jürgen, a wonderful good morning. What do you say about this topic?
Good morning, dear Yasemin, good morning, Marc, and everyone in the room. I actually wanted to politely, gently disagree, but you got the curve back at the end, in my understanding — because I’d claim, data are the golden nuggets, equal to the assets of the future. But that would actually be the other point, which maybe fits into the risk part, the opportunities and risks weighing. My initial thought was actually that the term “big data” forced the question upon me again, what’s actually the difference between big data and a database, a plain old relational database, for example, which most of us have been dealing with for decades. Databases have existed, in my view, since the first moment some tin contraption worked, and the question for me was, what is big data, in contrast to a normal database, and I think I’ve come to the conclusion for myself that, once again, the “what for” makes the difference, and I’ll claim now, and that’s why data are, for me, the golden nuggets of the future, that a big data inventory is the necessary prerequisite in the first place to establish the construct of artificial intelligence at all. Because big data is basically the feed for the correspondingly developed algorithms, which then, essentially, to put it briefly, lead, through pattern recognition, to the conclusion that Marc recently bought a razor and will now, at some point, need, say, replacement blades. And this quantity of data — not just this algorithm, this instruction, that’s all it really is, linking A with B, if C equals D, for example — but this quantity of data, I’ll say more about that in a moment, is decisive for how this algorithm learns, and only through that, in my understanding, does it become artificial intelligence through machine learning. And what I also wanted to say regarding data — there are three criteria, if I remember correctly, that have existed for quite a while now, for the quality of a big data inventory. And these are based on the three famous V’s — velocity, variety, and volume. So the quantity, with its diversity, and with the speed of access. And basically, the really essential starting point among these three, in my opinion, is volume — the amount of data available to me — because the quantity, as a statistical population, so to speak, because the size of the set ensures that, in the pattern recognition within the algorithm, an increasingly more plausible result can be generated. It makes a difference whether I ask a question, collecting data from the people in this room, or whether I have access to all the data of the people in the Federal Republic of Germany. I think that’s obvious. And now, to conclude, and that’s why I say again, that those who hold this big data in their hands will be the future rulers and decision-makers on this globe. And I’ll go out on a limb and say, right now, three are on the starting line, for me — that’s Apple, that’s Google, and that’s Amazon. And if you look at our daily use of various applications and apps, and go a bit deeper, you’ll always find one of these three. Thank you all for listening.
Yes, wonderful, thank you, thank you so much, and you just mentioned Apple, Google, and Amazon in this context. Regarding this aspect with AI, for example, pattern recognition and so on, I’d like to pick up on the fact that this is basically one field, and we already talked about how it’s important not to look at all these aspects we’re considering here under this umbrella theme Pioneering Spirit with Brain in isolation, but to make these connections here too. AI is basically one aspect we can pick up on. What makes the big data topic so interesting, or what also makes the difference sometimes, is, for example, that we used to have the situation — let’s take an airline as an example — that
you had it tangible too. Say we booked a flight with some airline. Then our data was stored with that airline, because we became a customer of that airline. Today there are a lot of different options — we can still go directly through an airline, but a lot of us will probably also use comparison portals, and then our data isn’t just stored with the airline, but there are companies offering these comparison portals, and the moment I look at various things and so on, I also transmit the corresponding data on this platform, where I’m looking at things. That means the company doesn’t offer any flights itself, just the pure information brought together here, made accessible to me, as interesting for booking a flight, and thereby tracks my behavior on one hand — what I look at, and so on, what’s interesting — and can then use that again in other contexts too, for example to increase their sales elsewhere, or something, all the way to the data that gets stored accordingly here too, which is again a central point — we could call it a central point — that doesn’t actually have the assets itself, at that point. And that creates a massive amplification of this data collection, at that point, and of course there are technical backgrounds too, like API interfaces and so on, that play a role here, that enable certain things this way. And then, of course, you can use the data you’ve collected in some way again for other technologies, or in the context of other technologies too, and make it accessible. And so there are certain players, like Apple, Google, Amazon, who discovered and understood the power of this data relatively early on, and use it quite deliberately in various places, not just to collect this data, but to develop new products or services in a structured way, to conquer new markets again. And that’s why it’s no coincidence, for example, that there’s an Apple Pay or something. And more is supposed to come here too. That means we’re already far away from what Apple originally offered, and it’s suddenly moving into completely different fields, and becoming competitive that way too — Apple thereby represents competition, so to speak, to, say, financial service providers or similar, because if we start thinking in that direction, that you can then, say, get into installment payments and so on and so forth — so on one hand it’s a relief for the customer, to get financing relatively quickly and smoothly through this, and similar, and at the same time other providers increasingly get pushed out of the market, or switched off, you could say, at that point. Would you like to add to that too, Marc, and then Jürgen?
Yes, I wanted to briefly get back to Jürgen, because I think that’ll make it clearer. Databases have existed before — what really makes the difference for me, to give an example, is — before, I had data that let me say, Jürgen, or so-and-so many people in Cologne, have a phone of brand XY. If you apply that to computers, then today, and this probably isn’t that bad for any of us, someone has my IP address, someone has my country code, someone knows what language I use, and someone knows what operating system I have. I can live with that, that doesn’t bother me. What’s the difference for me, the really essential difference to before, is that the data isn’t just about the device anymore, but into the device — namely, exactly these kinds of data are also about, what relationships do I have, what friends do I have, what’s my political orientation, what’s my sexual orientation, and so on. And those are data I might not be able to live with anymore, because they intrude much, much more into my personal sphere.
Yes, thank you very much. Jürgen.
Thank you, Yasemin. I wanted to get back to your remarks — I’m really grateful for this example with the comparison portals. Something that’s really significant, in my view, is that this creates a hierarchization in the power structure, based on the available data. Why? The original data, integer and authentic, comes from the user to the comparison portal. This comparison portal then feeds, let’s say, Lufthansa, or the airline. And through the overarching application, namely the internet browser, this data certainly also makes its way, in some form, either to Apple or to Google, depending on what device we have in our hands. That creates, for me, a hierarchization, completely value-free for now, in that various companies have access to my data. And the big vacuum cleaner, as I like to call it, the big vacuum cleaner sitting on top, let’s maybe call it Amazon, and that’s the step for me between big data and toward the algorithm, namely toward the instruction and logic, or toward the system’s learning capability. That’s able to trace where the data actually comes from. Otherwise it wouldn’t meet the basic requirement of a database for authenticity and integrity — so it needs to know where the data comes from, and it needs to be redundancy-free and traceable. And then it’ll find, this comes from Jürgen Kertt. And Jürgen Kertt provided this data in connection with an airline, by having previously used a comparison portal. And that could, under certain circumstances — and I’m quite sure of this — lead to the big data overlord, Amazon, being inclined to instruct this comparison portal, in a certain context, when Jürgen Kertt is searching for flights again, please don’t suggest Lufthansa first this time, but, shoot me now, because there’s a corresponding, business connection at play, bluntly put. This flow of data through various interfaces isn’t actually anything new, dear Yasemin, you certainly know that too — these are the potential connection points for opportunities and risks. You can say that completely value-free. And in a hierarchized database, in this case measured by size, whoever sets the pace is always the one who has the most data. And really important, dear Marc, regarding your example — the one who, in the transition from rubbing, drying, vacuuming, also developed the corresponding logic, to transfer machine learning, for example, into deep learning. Thank you for listening.
Yes, wonderful, thank you, thank you so much. Marc, would you like to add to that too?
Yes, dear Jürgen, you’re completely right about that. That’s also the next path, for me too — what’s going to happen next, namely the predictability of data, not just in political elections, but also in purchasing decisions, and so on. And you’ve probably also heard how, in America, a woman was really surprised when she got offered pregnancy products, and then found out she really was pregnant, because her purchasing behavior alone showed that something had changed. The woman herself hadn’t noticed, and suddenly got offered baby items, until her doctor said, congratulations, you’re pregnant.
Yes, absolutely, thank you, thank you so much. And there we’re of course partly back in this aspect we’ve talked about repeatedly in other rooms too, that this doesn’t just create opportunities, but also risks arise, whether that’s in election contexts, or other contexts, or simply used and steered for economic benefit, in inverted commas. That means we can assume that through this big data, and through the use and application of this data in a structured way, and then with this amplification that’s there, all the way to whether it’s used for AI or other things, markets can also get steered, based on, and Jürgen put it so nicely, various relationships — business relationships that exist here, for example. It could be politically driven things too, but especially, let’s assume, for now, business relationships. Let me pick up the comparison portal example again too — I think you can see this really, really nicely there too. Even with comparison portals, not going one step higher yet, but staying with comparison portals — if you know what happens behind the scenes, in inverted commas, you also follow along, or know, for example, that it’s no coincidence which company gets listed first, or shown first, but that companies actually pay money for that, to get listed among the top 3, for example, and depending on which position they get listed at, corresponding amounts flow. That means you can go into a corresponding steering here too. And on the next level up, if you look at browsers and so on and so forth, and pick up these big players here and look at it, then really entire markets can get steered here too. And the question is, does anyone even notice this at all? So the topic of transparency is definitely a topic here too. That means, when certain algorithms are working in the background, when we get influenced without noticing it, that slips into a danger zone. It doesn’t always have to be misused, but at least into an area where it gets difficult, and that’s the big challenge in this context too. On one hand the opportunities that arise from this, which are wonderful, and on the other hand, all the way to abuse of power, at that point. And, well, wherever there are people, both developments can always be observed — maybe not per se always good and per se always bad, in quotation marks, but to be used contextually, accordingly. Would you like to tie in with any approaches there, Marc, Jürgen?
I can confirm all of that, dear Yasemin. It’s interesting — I once tried, with someone who knows more about computer science than I do, booking certain hotels through different browsers. So I tried booking a hotel from Italy, then from the Netherlands, then from the UK, and then with different devices, so with an iPhone, with an Android phone, and the interesting thing is, at the same time, of course at the same time, maybe offset by a minute, you get different results.
Yes, another aspect in this context too, let’s not assume the “bad” for now, even if it wouldn’t get used abusively, and you can open up the next door there — Jürgen basically already said this earlier — a really clear imbalance arises here in what players like Apple, Amazon, and so on are able to do, so to speak, opening up new markets too, thinking far-reachingly, analyzing, and picking that up in a positive sense for themselves, to open up further markets too, while other companies, well, they can, in inverted commas, run themselves ragged, and reach a point, if they only work with their own assets, their own database, and so on, where they simply don’t have this potential at that point, if they don’t think differently, act differently here. But not every company can become a player like Apple, Amazon, or whatever else, because that also has to do with skills, competencies, and so on, and so forth. And the resources I have available too, all the way to the means I can deploy, financial ones too, and marketing too. So various components come into play here that work together accordingly. Jürgen, would you like to say something about that too?
Yes, I’d like to say something about that too, because, forgive me for saying this, but we’ve now spent almost 40 minutes talking more or less about risks and the “bad image” of big data. Of course there’s a positive side too, and the positive side — and I keep saying, or I’ll say it again — big data is, for me, the foundation, really the foundation in the first place, for a corresponding algorithm, and algorithm I’ll now set as a synonym for the type and quality of thinking, or of learning, how an algorithm learns. I have a current example. The level, which depends on the power of the data behind an algorithm, starts with so-called unsupervised — sorry, supervised learning. My example is cancer detection. This supervised learning is based on a basic set of data, where I predefine for this amount of data, if this and this appears, then this and this is given. So basically a screening test — if your PSA value is higher than XY, then there’s suspicion of prostate cancer. That’s this supervised learning. Unsupervised learning — there the pattern isn’t predefined yet. That’s certainly possible with basically any basic set, which doesn’t need to be large. The second one, unsupervised learning, there the machine decides, based on the cleverness of the underlying algorithm, that a pattern not yet recognized by doctors, and now here it comes, based on necessarily much larger amounts of data, can also be a sign of cancer. Something that’s absolutely no longer feasible from the isolated view of doctors, medical teams, clinics. That’s unsupervised learning. And the third point is reinforcement learning. There the algorithm gets a little reward for what it’s achieved and recognized. And this little reward, this positive conditioning, or reinforcement, leads the algorithm to develop toward recognizing this as positive, and gets encouraged to keep searching. And if you look at these three levels, it’s basically nothing other than, on a meta-level, roughly, the way we can learn too. I’m saying this deliberately carefully — there are other forms too. And I’ll come back to my original statement — golden nuggets — the bigger the data set, and whoever has this data set, will, and I want to emphasize this, ultimately be able to actually carry out this reinforcement learning in the first place. And that brings me, dear Yasemin, dear Marc, of course, for me too, in the first moment — you’ll enlighten me or reassure me, Yasemin — that’s human, that I say, oh, that’s a risk. If someone recognizes that I have cancer, and my family doctor isn’t even able to. Then I say, well, that’s in someone else’s hands, but they recognize it. And that’s the advantage for me. And I have to say, I honestly don’t care where this information comes from. As long as I know it. Thank you for listening.
Yes, thank you, thank you so much. I’m totally with you on that. And there we’re at this benefit aspect, ultimately, and so, what does that establish? This information, which I recognize as a benefit for myself as a consumer, whether that’s in a medical context, or in other contexts, leads to me using that for myself too. And I’d just like to open up another perspective, because you said earlier we’d spent about 40 minutes talking about risks, or the flip side. I honestly don’t feel that way, because for me it’s really important, in this context, to first show the connections. Only once I recognize these connections, what’s actually happening here, am I actually able, as an individual, as a private person, to get into a certain kind of self-regulation too — always back at the topic of education — as well as, from the perspective of a self-employed person or entrepreneur, to realize what significance big data actually has, what’s happening there, how these hierarchizations, so to speak, take place, and to realize and consider what that means for me — what opportunities can I derive from this for myself, as a business, if I have certain background information, and use that accordingly for myself and for my business, at that point. And there we’re back at this context with AI too, where a lot of people, especially in German-speaking regions, in these discussions, you notice people say, I’ll just watch the developments first, I’ll lean back for now. I think, especially when we open up topics like this and show these connections, it should become relatively quickly apparent, at the latest once we open up these various perspectives, that, especially as a company, you don’t actually have the chance, or shouldn’t take the time, to say, I’ll lean back here for now and just observe first, because it becomes difficult, if you look at this amplification, what’s happening in the background, to catch up on anything at all. Then I actually get into a dangerous situation too, if I wait too long, because I can no longer catch up on this amplification, at that point. And that also applies, by the way, if we look at the past — you could well ask the question, how far can you even go with players like Amazon, Google, Apple, and so on, which have already come up here as terms — do you even have a chance? And to what extent can you recognize and use certain things positively for yourself here, at that point? Jürgen, you wanted to add to that too.
Yes, I’ll still address your last statement. How can you catch such big players? Absolutely, fundamentally. And I think you know me well enough by now that I’m a pioneer for creating the necessary foundation of understanding and connections, all the way to the precision of terminology, before talking about something.
Yes, that’s right, that’s absolutely necessary, and I’d venture to say, if a larger part of people were aware of what actually happens in the sequence and nature of this big data collection, which a lot of people actually know about, but basically had an understanding of what happens with it, and how our future gets structured through it, then we’d probably have a much broader discussion, similar to the one we’re having here in this room. And, dear Yasemin, regarding your
question — and now I’ll take this consistent step and say, how do you catch these big data monsters? Well, please, let’s first clarify what for, and if I want to catch them, that also means, for me, if I catch someone, I’m limiting them, restricting them. And with that I also have to reckon with negative consequences, under certain circumstances. See my example with this reinforcement learning, which might no longer be possible then. Thank you.
Yes, wonderful, thank you, thank you so much. And I’m curious what Marc has to say about that too.
Yes, well, the first thing I’d counter with is the aspect of market monopolies that develop here. That means, we’ve always established this, there are always these different sides. So on one hand it’s certainly a really wonderful aspect to see what sequences result from this, and that opens up new opportunities for all of us too, for the future, in inverted commas. And on the other hand, if we look at it from a different perspective, like competitiveness and so on, there are certain disadvantages too, that result from this. And that doesn’t mean a restriction, from my point of view, doesn’t mean there are no more positive sequences at all, but maybe reducing the speed a bit again here too, so others get a chance to play along on this big field too, in inverted commas. I don’t know what you two think about that. Marc, Jürgen?
Sorry that I’m jumping in like this — I just wanted to say, it’s totally exciting to experience live what Jürgen said, in terms of perception. We have an incredible number of doom-and-gloom moments, and precisely so that doesn’t happen, I’d just like to remind you, I started the room by saying that people recognized American flu patterns, that it’s good for cancer therapy, that machines learn, and that companies understand customers better as an introduction. But still, the feeling arises, and I feel the same way, that after 30 minutes you feel like you’ve only talked about doom-and-gloom moments. And sorry, Jürgen, for jumping in.
Wonderful, thank you, Marc. Jürgen, what do you say to that?
I like your jumping in, dear Marc. It’s as always, taken straight from life, and I’ll touch my nose and bow my head in apology. Okay, yes, regarding monopoly, dear Yasemin, this is now an example for me, and our discussion increasingly an example, of what I appreciate, what I often wish for — that we move between two poles — and there too I’d like to gently, politely disagree, and say, dear Yasemin, a monopoly doesn’t have to be bad. Monopoly is unfortunately negatively connotated for most people. But a monopoly, and now let me remind you of one of our recent topics, sustainability — a monopoly that, by virtue of its availability and its competence, upholds the three essential components of sustainability, namely ecological, economic, and social — and, again, a topic from our room recently, in the field of ethics and morality — then I can only say, what would the alternative be? This monopoly wouldn’t exist. So their solutions and benefits wouldn’t be available to me either. This shows, for me, and I think for all of us, that this topic, alongside its complexity, always has two sides. Decisive is, and this has come up often too, and you highlighted it as a point, dear Yasemin, ultimately it’s not the data set. Ultimately it’s not the algorithm either. I’m firmly convinced that there will never be a machine, in my remaining time on this planet, that comes anywhere close to human cognitive ability. But I think this is a development we always need to see from two sides, and one that’s still made by humans. Behind the machine stands a human, and nobody has been able to tell me yet — nobody has been able to tell me that an AI will ever run out of control at some point, because there will always be the option to simply switch off the power. Where is the problem? Thank you for listening.
Yes, wonderful, thank you, thank you so much, and that’s actually already been done in the past, after an AI developed its own language — I think that was at Facebook, and at some point they had to pull the plug, because the developers, so to speak, could no longer understand what the AI was exchanging and talking about among themselves. And so we have a good example here too, that this can ultimately be, and will be, the solution too, at that point, and shows that we have the ability to act accordingly here, and aren’t somehow at its mercy in any way. We’ve opened this up in a lot, lot of places, as Jürgen also just said, that it always stands and falls with the human behind it, because ultimately that’s nothing other than a programming, which, while able to keep learning, based on the algorithms, refine it, bring things together, and so on and so forth, still always has something to do with our own programming. And if things happen that aren’t wanted, then we have the option to pull the plug. I find that such a nice image too. And also this monopoly aspect, I think we can maybe pick that up again — we’ll be bringing these topics together and discussing them on another level too. So I find it quite nice too that this monopoly aspect got picked up, or came up, today too. Because these three aspects too — ecological, economic, social — of course raise the question, to what extent is it even in a company’s interest to actually act ecologically, economically, socially? Or does it maybe still need a central authority that steers one thing or another accordingly? No idea. Question mark with a question mark. Those are aspects we can gladly pick up again in the context of the political and societal discussions we’ll open up too, and get into exchange here too, to consider what options exist there. Does it make sense to impose restrictions here, or maybe rather not? Or, what are the possible developments in one direction or another, with the various scenarios we’ll have accordingly too. I’m really excited about these rooms too. And, yes, we’ll continue tomorrow with quantum computing first. So we’re staying in this aspect of high speed, and we’ll stay on this speed level too. So what exactly is that, actually, and what’s happening there — let’s look at that too. I’m looking forward to that, to going into this deep dive together with you tomorrow too, and seeing what points and aspects come together here again. And with that I’d like to move into today’s closing round. Marc, Jürgen, what are your highlights, or thoughts you’d like to underline to close out this shared room today? Marc.
Dear Jürgen, you gave me quite a chunk to chew on today. A monopoly doesn’t have to be bad — I need to digest that first, and think about it. Thank you for that. I’d like to close out the room with one more positive and current real-world example — big data can also be used in shipping and logistics, because it was never possible before to plan truck routes in a way that factors in weather, traffic jams, and fuel, gas prices. All of that also gives businesses new options for planning, saving money, and being more efficient and productive. Thank you, and a wonderful day to all of you.
Great, thank you, thank you so much. Jürgen.
Yes, thank you very much, dear Yasemin, for the moderation, excellent as always, and dear Marc, for your contributions, excellent as always, and it’s a topic that generates so many connection points — I think, or I’ll take this away again, that we’re increasingly facing the challenge of designing an overarching meta-level ourselves, a neural network, for all the topics we address. And I have to keep saying, chapeau, dear Yasemin. That’s actually the foundation, just like big data is for artificial intelligence — the foundation for us to fall constructively and value-creatively from one topic into the next. That’s really, really refreshing. A nice day to all of you in the room too, and hopefully see you tomorrow. Ciao, take care.
Yes, thank you, thank you so much, and I can only agree. It becomes clear that, in order to reduce the complexity, so to speak, on one hand, it’s important to go deeper into individual fields, and here, in this exchange, look at what information, what knowledge we can bring together, what experiences we might have personally had, what we’ve observed, then expertise flowing in from various fields, to then bring it back together, and not look at the chosen topic focuses, so to speak, or the technologies, in isolation from each other, but pick them up again in their interplay, so as not to leave the complexity out, but bring it back in, and then ultimately discuss it on various levels, to reflect on our own thoughts too, at that point, to sort them out. And there we’re partly back at big data too. Basically that’s nothing else, because big data doesn’t just work with data collection, in inverted commas, but a really essential part is also structuring this data, on one hand, and then doing something with it again, meaning applying it again subsequently. And basically I find it nice to draw these parallels here too. And by the way, even when we talk about artificial intelligence or something, it’s no coincidence that the whole thing is called artificial intelligence too. That means parallels actually do get drawn here, to thinking, to acting, in relation to the human brain, and certain things get mirrored too. And so, whether we open this up regarding thinking processes, regarding learning processes, and so on, and these keywords came up here too, you see, in inverted commas, that this concerns our meta-level on one hand, and at the same time reflections, in inverted commas, take place, which get drawn on here, in inverted commas, to make one thing or another possible in a technological context too. And that’s really mega fascinating, when we show these interfaces here too, show these developments, and then also recognize that the human is always the basis, at that point, both in how it later gets applied, what gets made of it, as well as in the creation of all these technologies, where processes of the brain also get used, get mirrored, at that point, to develop technologies too. And that makes it so mega exciting, and also the discussions on an ethical level, on a societal level, and on a political level, looking at what that actually means, and how we handle it, what opportunities and risks we see, so that we don’t block out either side, and at the same time make use of these opportunities for brain-juggling, as I like to call it, to generate new insights for ourselves personally too. And with that I’m really looking forward to tomorrow — quantum computing, I mentioned it, is the keyword for tomorrow. We want to go deeper into that, and so you’re all very warmly invited to join again tomorrow at 8. Thank you, thank you so much for participating, and for the stimulating discussions we had together today. Until then, ciao!
