Latest / Rewire To Rise– Beyond The Mindset / AI & The Human Future | Episode 1: AI & Business (Live Panel)
Transcript
- 0:32I'm so excited for this first edition AI and the human future.
- 0:37Rewired to Rise is a home for the overthinkers and the high
- 0:40achievers. I'm joined by three world class
- 0:42experts. And now please welcome our first
- 0:45guest. Hello, Ka.
- 0:49Hello Mirela and thanks for inviting me.
- 0:53So I have a question to start with.
- 0:57You will help us understand the rules and market forces that
- 1:00drive corporate profits, right? How do we build systems that
- 1:04actually demands accountability curve?
- 1:08Thank you. There is quite a story to that,
- 1:10so I've made a presentation. I'll just share my screen.
- 1:16Well, I'm going to let you to your presentation, OK.
- 1:19So see you later on, Cara. Thank you.
- 1:22You're welcome. Yes, that's quite a lot to to
- 1:29this story, but I have these supporting slides.
- 1:33We can also answer questions in the questions and answer section
- 1:37afterwards. After my presentation, we'll
- 1:40hear Mark. After that, we'll hear Torsten
- 1:42and then we still have time for the questions and answers and
- 1:46you will get the presentations in the goodie bags afterwards.
- 1:53AI nowadays is really an infrastructure.
- 1:58I mean, in the early days, Chat GPG was launched in the end of
- 2:032022. It was interesting, it was
- 2:06fascinating. It was also buggy and it was not
- 2:10always that predictable. It still doesn't, not isn't
- 2:15always, but it has by now been built into so many processes, so
- 2:22many companies, so many departments.
- 2:26So it is really by now mission critical for my part as an IT
- 2:32consultant and IT trainer. I meet meet IT in the financial
- 2:36sector, I meet IT in the healthcare, I meet IT in public
- 2:40administration and so many other lines of of business.
- 2:43And I can see so many of these lines of business where the
- 2:47building nowadays depend on this at least as part of the toolbox.
- 2:53It is only part of the toolbox. So it is crucial that we
- 2:59understand what it's all about and it's really the pros and
- 3:03cons and there's a lot to consider.
- 3:05But then again, we have time to dig in.
- 3:10Let's start out from what is the economy behind this because it's
- 3:15quite large number in the left hand side, we can see the annual
- 3:20value potential at least as estimated by McKinsey and these
- 3:25are annual numbers and these are dollars.
- 3:28So it is huge numbers. It is really lifting the economy
- 3:35in part by increasing productivity, in part by doing
- 3:40what was otherwise hard to do, in part by inventing new
- 3:45systems, new ways of of working. And IT has the potential, and we
- 3:53already see early signs of that for boosting actually the global
- 3:57GDP. So we will get more wealthy
- 4:01because the machines can help us do what people would otherwise a
- 4:07decent part do. And then people can do what is
- 4:10even more valuable at this point.
- 4:12It has not created unemployment, but it has created more value.
- 4:18Of course, there are some who will get this, and there are
- 4:22some who will have to rethink their their job positions.
- 4:25But most of all, it is the potential of getting to
- 4:29something even more valuable, creating an even better income
- 4:35for you for for you. And part of it is about
- 4:39productivity, but it is handled quite differently in different
- 4:46areas of the world. These are the three main areas
- 4:51of AI. Of course, the world is bigger
- 4:54than this, but most other countries are really connected
- 4:58to one of these three areas. I will not present it in every
- 5:04word, it's already on the slide, but we can really see how the
- 5:09market driven approach in US is quite different from the state
- 5:14driven in China and then again quite different from the rights
- 5:19driven in the EU. For my part, I live in Denmark,
- 5:24Most of my work is done in Denmark.
- 5:26So I am an EU citizen, which means that the regulations also
- 5:35for me, and I'm actually quite happy that there are these
- 5:39regulations because it is new technology, it is very powerful
- 5:44technology. It can be used, it can also be
- 5:47misused and, and I hope we will manage to put guardrails on at
- 5:54least some of the potential misuse situations.
- 6:00That's much more to the story. And there are links in the booty
- 6:04bag that you will get after this live session telling much more
- 6:09about these regulations. It is also a market with huge
- 6:17investments nowadays. If a company is to make a new
- 6:22frontier model, that could be GBT 5 and the coming GBT 6 from
- 6:28Open AI. It could be the next version of
- 6:31Gemini from Google. It could be the next version of
- 6:35Claude from Anthropic. The cost nowadays is in round
- 6:40numbers, 1 billion dollars. And the the reason why it is so
- 6:46expensive is that they really crawl the science in it.
- 6:50They spin off these data centers, they collect all the
- 6:54data available. Some of it is not publicly
- 6:58available. In the case of Google, they own
- 7:01YouTube. So they can see both what is
- 7:05publicly available, but also what is not publicly available.
- 7:08The same goes for their access to Gmail and Google Maps and
- 7:12whatever other services they have.
- 7:14So it's really also about the size of the data pools.
- 7:18So this spin these servers up and they crunch those data for
- 7:22month, it is often two months, even three months, sometimes
- 7:27even longer that these CPUs or GPUs.
- 7:31It is really run at full capacity in order to understand
- 7:36and structure the data. And from that point you will
- 7:39have a good statistical model and from that point you should
- 7:45fine tune it and really adjust the behavior.
- 7:49We can see that nowadays the GBT 5 model from old May I is no
- 7:54longer new and we have seen both the 5.1, the 5.2, the 5.3 and
- 8:01currently the 5.4. So those are not entirely new
- 8:04models, those are just refinements of the existing GBT
- 8:095 model. Similar patterns can be seen at
- 8:12Google and Anthropic and other players, but those are the big
- 8:15ones. And I don't mention Microsoft
- 8:17because Microsoft do have their own AI department, but a big
- 8:22portion of what they run is really the model from Open AI,
- 8:27the GPT 5 model, and the reason they also made business with
- 8:31Entropic using the plot model, which is the the best one for
- 8:35for coding. So it takes a lot of money, it
- 8:40takes a lot of time, it takes a lot of energy to create an
- 8:44entirely new model. Once the model is up and
- 8:47running, they have improved the the efficiency.
- 8:51It is still a lot of power used, especially if you want to create
- 8:55an image or video because there is a lot of compute to that.
- 8:59But if it's a simple question, it isn't really all that much
- 9:04compute intensive by now because they have improved.
- 9:07A lot. You will often as a company
- 9:16either render platform. That could be the copilot from
- 9:20Microsoft, that could be the Gemini from Google, it could be
- 9:23the Jet CBT from Open AI. It could also be other models
- 9:28from other companies but those are the big ones.
- 9:31But it is also possible to build your own one.
- 9:35You will usually not build it from scratch.
- 9:37You'll really download an open source models and there are
- 9:41plenty of those out there, both from the big players and from
- 9:45the smaller players. You can install it on your own
- 9:48server. You can then do your own fine
- 9:52tuning and you can make sure that it's not publicly available
- 9:57on the on the web. And in the organizations that I
- 10:02serve, I see that many rent standardized platforms.
- 10:06It's easy to get up and running. It's not all that expensive, but
- 10:10with the most sensitive data, they will also build proprietary
- 10:16systems, especially in the healthcare systems or in public
- 10:20administration or in finance. Well, there are data that you
- 10:24don't want to risk putting in a data center of some company and
- 10:29you really don't know exactly how they will use it afterwards.
- 10:35They do have their terms and they do promise to protect it
- 10:38and I hope we can trust those terms.
- 10:44But then again, once it is run by the the Internet services,
- 10:49there are also a lot of touch points on the way to the data
- 10:53center. And in the the case of American
- 10:56systems, there are also American regulations giving the the
- 11:02administration access to that data if they think that they
- 11:06have a need for it. So it is really also about
- 11:09geopolitics. I know that Mark will talk much
- 11:12more about it in his presentation afterwards.
- 11:20I've made a case study just to be a little more specific and
- 11:25maybe get down to a slightly smaller scale because the
- 11:28numbers before for training a new model for the energy
- 11:34consumption and so on worldwide are huge numbers and that
- 11:38they're really hard to understand.
- 11:40And at least from my perspective.
- 11:42So these are our Danish numbers. Denmark is already a very
- 11:47digital country. It is a country with a high
- 11:53level of trust. And I hope that besides trusting
- 11:57other people that we can also trust services.
- 12:01Some are more trustworthy than others, but we will select what
- 12:07serves us the best. I see in the Danish professional
- 12:12market, many choose Microsoft. There is a long history of Mark
- 12:17Microsoft being strong in Denmark.
- 12:19They have bought Danish companies.
- 12:20They have a fairly big office in Denmark, so you can really go to
- 12:23their office and talk to employees.
- 12:26And that is different for for many of the other tech
- 12:29companies. We also have in Denmark high
- 12:35percentage of high speed broadband connections which are
- 12:39relevant for running these services.
- 12:46But it's still about the data and you can ask questions which
- 12:53have no confidentiality to it. Well, you should be good.
- 12:57But there are data that do contain contain confidential
- 13:02parts to it. And then you should really find
- 13:05your way. So I've made this sketch.
- 13:09You will usually start out from looking into the private company
- 13:17dates, and you might not present those data to those models at
- 13:20all. That might be where you don't
- 13:23use AI at all or where you build your own system, especially if
- 13:27you are a big organization, because of course there is a
- 13:30cost to that. There is a complexity to that.
- 13:33There will be tasks that can be handled by rendered services
- 13:39from Microsoft or Google or May I or even other companies.
- 13:44And in those cases, you can use those global foundation models
- 13:49that are so expensive to train. But you should really be aware
- 13:56of your intellectual property. You should be very aware of how
- 14:03they handle these data also, because if they hallucinate, and
- 14:07these systems will hallucinate to a certain extent, Mark will
- 14:11tell more about it and Torsten will tell more about it later
- 14:14on. Well, you risk that your data is
- 14:18used and presented in a way that is not how you would process it
- 14:24if you were choosing all on your own.
- 14:27So there's a lot to to consider in the in this area.
- 14:33So I suggest that you check your window.
- 14:38I will not read out every word on this slide aloud.
- 14:42You can read it for yourself. These slides will also be sent
- 14:45afterwards. And I also have prepared a
- 14:48checklist that you can use in a normal Word document that is
- 14:52part of the goodie bag that will be sent afterwards because it is
- 14:55really how are they collecting data, who owns that data
- 15:00afterwards and a lot of other questions.
- 15:04For my part nowadays, I mainly do AI.
- 15:08Before it really took speed, I was doing a lot of data
- 15:14analysis, Excel and Power BI and that sort of of, of systems.
- 15:18And, and I could see in in that situation that sometimes you had
- 15:23outdated data or you had data containing errors or you there
- 15:28were missing data. And, and I saw that if it was
- 15:33related to media and marketing, many would use Google services
- 15:37or services from Meta and Facebook, Instagram, that sort
- 15:41of services. And the, the service terms from
- 15:44Google or from Meta in very, very short and slightly
- 15:51characterized is really that they own the data.
- 15:55As soon as you have put some text, it's their property.
- 15:58As soon as you have put a picture, it's their property and
- 16:01they can use it for whatever they want.
- 16:03In reality, they don't start selling your data because they
- 16:07would be so much in trouble and customers would leave them.
- 16:12But there were the Cambridge Analytica scandal with a beta
- 16:16bag in the day. So there is really at risk that
- 16:19the data that you think were yours is now not really yours
- 16:25any longer. But then again, if you pay a
- 16:29good vendor, you're much better off than if you use a free
- 16:33service, because if you don't pay with money, well, you'll
- 16:36probably pay with your data instead.
- 16:41So I suggest that you ask these three questions.
- 16:47It is good that you have data, and it is especially good if you
- 16:51own your own data and with relevant services and with
- 16:56relevant terms. Well, you're good.
- 16:59I see many using Microsoft, but of course, there can be reasons
- 17:03for choosing other providers. For my own part, I mainly use
- 17:07Microsoft nowadays for business purposes.
- 17:09I mainly use Google for private purposes.
- 17:12I also have a Jet GBT from Open AI and I really have currently
- 17:17paid subscriptions to 15 different services.
- 17:20I also have to Adobe and some specialized services for for
- 17:24legal stuff and for search optimization and for a lot of
- 17:27other purposes. And I have logins to even more
- 17:30services. So I really try to test out
- 17:33different services and I can see pros and cons to it to to all of
- 17:38them. And if you get relevant answers
- 17:43to these questions, if you really do your homework, well, I
- 17:47guess you are, you're well off. So the key takeaways from my
- 17:55part of the of this live session is that you should see this as
- 18:01an infrastructure. It's not only a technical
- 18:04services, it is definitely a technical service, but it has
- 18:09connections also to people, to business processes, to how
- 18:14companies work together, to how they interact with other
- 18:17countries, especially the big ones.
- 18:20And there are these geopolitical realities.
- 18:23Well, you could better make sure that you're managing in that.
- 18:27I can see that the US is still the biggest pay play on this
- 18:31market. China is catching up very fast.
- 18:35For my own part, I'm not all that tempted by the Chinese
- 18:39services, but I can be fascinated.
- 18:41And you should really make sure that the trust that is hopefully
- 18:46there from the start can really also be preserved while using
- 18:50these services. And it doesn't come easy.
- 18:52There is homework to to do so when my presentation is over in
- 19:03a minute and when it's time to give the, the, the talk back to
- 19:07Mirella and afterwards to, to Mark.
- 19:10And I will at least advise you to think of these questions.
- 19:17And when you receive the goody bag afterwards, you will have
- 19:22access to the vendor scorecard. You will have access to a
- 19:26curated source list. You will have the hallucination
- 19:30checklist that is made by Mark. I've also made one for
- 19:34filmmaking that is mainly by demand before this event.
- 19:39It's not my, my main topic. I will just very briefly show
- 19:43the two first documents, but but only briefly because you will
- 19:47will receive them afterwards. The vendor systems scorecard is
- 19:56here. You really have some questions.
- 19:59You rate them from 1:00 to 5:00, you count the total score, and
- 20:04based on that total score, well, you have a better fundament for
- 20:08making your decision. Regarding the other document, I
- 20:13have collected links to the EU Artificial Intelligence Act,
- 20:20also to the Standard AI Index report.
- 20:24It's quite extensive to other good sources for additional
- 20:30knowledge and I think that was my 20 minutes so and we will
- 20:40catch up later with the questions and answers.
- 20:45So well, back to you, Marella. Thank you.
- 20:48Thank you so very much. Well, now let's go to the 2nd
- 20:54guest of today's event, which is Mark.
- 20:58He will show us how to use AI to live better without losing our
- 21:03own human judgement. So let's welcome to Mark.
- 21:08Hello, Mark. Welcome.
- 21:10Hey, great to be here. Yeah, so lovely to have you here
- 21:13as well. Mark.
- 21:15In the world of fast AI, how can leaders use these tools as a
- 21:21thinking partner to stay ahead without losing their own
- 21:25critical judgement? Thank you.
- 21:29Yeah, that's a great question and I'll elaborate in my
- 21:32presentation in a second. I'd just like to start off by
- 21:35saying that that exact tension between speed and judgement is
- 21:41exactly what I think we need to talk about more honesty.
- 21:45The leaders I worry about are not the ones who are too slow to
- 21:48adapt AI, they're the ones who start to confuse a IS output
- 21:52with their own thinking. My starting point for this is
- 21:57that AI is generally a useful as a thinking partner, but only if
- 22:03you stay the author of your own thoughts.
- 22:06You can use it to generate options, to stress test an
- 22:09argument, to challenge your own assumptions, and that's
- 22:13powerful. But the moment you let it
- 22:15generate the conclusion and you just sign off on it, you've
- 22:19outsourced the most important part of your job.
- 22:23I think the the irony here is that in the world where everyone
- 22:27has access to the same AI tools, human judgement doesn't become
- 22:31less valuable. It becomes more valuable because
- 22:34the competitive advantage is no longer who has the fastest tool,
- 22:38it's who has the clearest thinking about what to do with
- 22:41it. Yeah.
- 22:42True. So my short answer to your
- 22:44question is treat AI like a very well read colleague who has
- 22:49never had consequences for being wrong.
- 22:52Listen to it, challenge it, but never let it get the last word
- 22:57in the conversation. See if the technology will allow
- 23:03me to share the presentation. Let's find out.
- 23:08I think it's this one, isn't it? It is a.
- 23:13Helping hand. Fine, let's just get back to the
- 23:17start. OK.
- 23:18I'll let you what to your presentation, see you later.
- 23:25Right, so good afternoon everyone.
- 23:30I'd like to start with a small observation claim.
- 23:33Just an observation. Every time a new AI tool
- 23:37launches, the conversation focuses on what it can do, how
- 23:40fast it writes, how much it knows, how efficiently it
- 23:44performs. But I think the more interesting
- 23:46question is a different one. What does AI do to the way we
- 23:53think? That's what I want to explore
- 23:55with you today. Not what AI can do for leaders,
- 23:58but what it might be doing to leadership itself.
- 24:12Let me be direct about what I think the real risk of AI
- 24:15actually is. Now that AI will take over
- 24:19leadership. I don't think that's the the
- 24:21danger, at least not in the near term.
- 24:25The risk is something more subtle and far more human.
- 24:30The risk is that leaders voluntarily surrender too much
- 24:33of their judgement because AI makes it so easy to When
- 24:39something is faster and more convenient and sounds
- 24:43authoritative, we tend to trust it.
- 24:46That's not a weakness, it's just human nature.
- 24:50But in leadership, judgement is not a luxury, it's actually the
- 24:55job. So the question I keep coming
- 24:59back to is how do we use AI to sharpen our judgement rather
- 25:04than slowly replace it? Let's start with the common
- 25:15belief, something you hear almost every day, that AI saves
- 25:20time. Well, here's what I actually see
- 25:24AI. It does free up time, but that
- 25:27time almost immediately fills up with new tasks, more emails,
- 25:32more output, more versions. We don't save time, we produce
- 25:37more. Secondly, faster is not better.
- 25:43A faster answer is not the same as a better decision.
- 25:48Speed is not quality. We can now generate 10 mediocre
- 25:53options in the time it used to take to think through one good
- 25:58one. And thirdly, and this is the one
- 26:02I find most important, AI removes friction, but some
- 26:07friction is healthy. Friction is often where
- 26:10reflection actually happens. The struggle to articulate a
- 26:16problem clearly, the discomfort of shedding with uncertainty,
- 26:21the work of pushing back on your own first instinct.
- 26:25That's not inefficiency, that's thinking.
- 26:36This is, if I had to put it into one sentence, what I actually
- 26:40believe. AI can make leaders faster but
- 26:45also worse if they lose the friction in their thinking.
- 26:50I want to be clear, this is not an anti AI argument.
- 26:55AI is generally powerful. It belongs in our work.
- 27:01But we also need to be honest about the real risks that the
- 27:06very ease of AI creates temptation to skip the hard
- 27:10parts of thinking. And the hard parts are often
- 27:14where good leadership actually lives.
- 27:23So let me offer you a framework that I think is practically
- 27:29useful. I call it the three levels of AI
- 27:34use. The first level is AI as an
- 27:39assistant. This is sort of the safe zone.
- 27:43This is where you use it for drafting, summarising, structure
- 27:48during first versions, research leads.
- 27:53You can use AI freely in this zone.
- 27:55This is where it generally saves time without compromising your
- 27:59judgement. The second level is AI as a
- 28:05thinking partner. This is sort of more active
- 28:09territory. You use AI to challenge your own
- 28:13reasoning, to find blind spots, generate counter arguments,
- 28:18simulate objections. This is powerful, but it
- 28:22requires you to stay in the driver's seat.
- 28:27And finally, the third level, AI as a replacement for judgement.
- 28:31This is the danger zone. This is where you have to be
- 28:34aware when AI is allowed to conclude alone on strategy, on
- 28:39people, on risk, without genuine human review, when you hand over
- 28:45accountability without realizing it.
- 28:50So use AI to generate options, not conclusions.
- 28:55That's the principle. Options are EI territory.
- 29:02This is what it's good at. This is where it lives.
- 29:05Conclusions is your territory. And here I'd like to introduce a
- 29:17term called cognitive outsourcing.
- 29:23We talk a lot about outsourcing tax to AI, but I actually think
- 29:28that something much more subtle is happening.
- 29:31We're outsourcing articulation, the work of finding the right
- 29:36words for complex thought. We're outsourcing evaluation,
- 29:42the effort of actually judging whether something is good.
- 29:47And sometimes we outsource the resistance itself.
- 29:52And then it's important to remember that it's the
- 29:56resistance that so often creates insights.
- 30:01Think about the last time you struggled to write something
- 30:05difficult, a hard message, a strategic decision you weren't
- 30:09sure about the the very act of wrestling with it.
- 30:14That that wasn't a waste of time.
- 30:17That was where you're thinking actually happened.
- 30:22When we hand over that discomfort to AI, we lose
- 30:25something real. Let me make this a bit more
- 30:35present or concrete with the demonstration of sorts about AI
- 30:49hallucinations when the model sounds right but isn't really.
- 30:56Core talked a bit about this. A generative AI model produces
- 31:03texts that is fluent. It sounds very authoritative and
- 31:08it sounds very plausible, but in many cases you'll find that it's
- 31:12actually wrong because it doesn't distinguish between the
- 31:17data it's been trained on and the data it generates.
- 31:22It has no concept of truth, only of pattern completion.
- 31:29So it'll come up with fake titation, citations from
- 31:34scientific papers, wrong statistics, false facts, and
- 31:40it'll it'll be confident in what it says it it doesn't
- 31:45distinguish. And I'd like to give you some
- 31:56tools to deal with that. So I've set up a way that you
- 32:00can prompt your way to a better accuracy in your answers.
- 32:11One example of a prompt could be what are the key statistics on
- 32:16AI adaptation in European business?
- 32:19It's a very open prompt. It doesn't specify, and it
- 32:23doesn't make any demands on the system to actually produce the
- 32:28documentation. A much stronger point prompt
- 32:31would be list only statistics that you are certain exists with
- 32:37source and year. If you're unsure of the figure,
- 32:41say so explicitly, and then I put up 44 points.
- 32:48You can. You can actually use for this.
- 32:51Ask for some sources explicitly either in the prompt, or you
- 32:54could do it afterwards. Cite the source for each claim,
- 32:59and then afterwards you have to verify that the sources are
- 33:03correct. You can invite uncertainty.
- 33:09If you don't know, say so put this in your prompt.
- 33:16You sort of have to allow the model to be uncertain and to say
- 33:21if it is uncertain, request a step by step reasoning chain of
- 33:30thought. Prompting reduces the confident
- 33:32leaps to wrong conclusions. Again, it allows for
- 33:36uncertainty. And the last point is to ask it
- 33:41to argue against itself, so it actually debates the arguments
- 33:48that it uses. You will get this practical
- 33:53technique in the grid back. And not least in the form of
- 33:59this, I've tried to put up a checklist.
- 34:05How, when can you trust the output that you get?
- 34:09So first of all, does it cite the sources?
- 34:13Well, if it does, verify at least two independently.
- 34:17Preferably verify all of them. If it doesn't ask, just ask to
- 34:22do it. If specific numbers or dates are
- 34:28mentioned in the output, then you need to be aware, because
- 34:33this is one point where hallucinations often occur.
- 34:39Does it sound unusually certain in the answer?
- 34:43Is the tone very confident and that's not always a sign of
- 34:49accuracy. Use this as an indicator of
- 34:52where it could go wrong. Does it reference name, people,
- 34:57organisations, another point of danger?
- 35:01Check each one. Is the answer in your area of
- 35:06expertise? If yes, does the answer actually
- 35:12match what you know? Do you recognize that this is
- 35:15actually true? And if not, you will be able to
- 35:21tell if there are errors. But if this isn't your your area
- 35:25of expertise, you have to be very carefully and you have to
- 35:27verify. And the last point I invite you
- 35:33to check up on is, would a wrong answer here carry real
- 35:36consequences? And and here I return to the
- 35:53point that Core also made that AI is not towards it's actually
- 36:00infrastructure. Who owns the data?
- 36:11Sort of the same points that core make.
- 36:13You need to ask yourself as leaders, who owns the data, What
- 36:19data do you give access to and who owns it afterwards?
- 36:24What lock in are you creating? Are you building on a foundation
- 36:27that you control or are you renting infrastructure so that
- 36:31you are no longer in control? And finally, can you explain the
- 36:36decisions Responsible leadership requires that you can stand,
- 36:43that you can actually take responsibility for the decisions
- 36:48that you've made. But I don't want to leave you
- 37:05with a picture of AI as pure risk.
- 37:07That would be wrong, and it would miss the real point.
- 37:11And here is the paradox that I find generally interesting.
- 37:15The easier it becomes to generate texts, analysis and
- 37:18proposals, the more valuable the ability to judge what is
- 37:21actually relevant, true, wise, and responsible.
- 37:26When everyone can produce output instantly, the differentiator is
- 37:30not production capacity, it's judgement is the willingness to
- 37:34say clearly and accountably that something is not good enough.
- 37:41That is a deeply human competence, and it becomes more
- 37:44valuable, not less, in a world flooded with AI generated
- 37:48content. So as we're going to watch the
- 37:55end of my segment, I'd like to leave you with three concrete
- 38:00rules things you can actually use.
- 38:05Rule 1. Use AI to expand thinking, not
- 38:09shortcut. It treated as an amplifier, not
- 38:13a bypass around the hard work of reflection.
- 38:17Never let AI conclude alone on strategy.
- 38:20People have risks, the final call is yours.
- 38:24That's not just good practice, it's accountability.
- 38:29And finally, rule 3. Don't just measure efficiency,
- 38:34measure quality, dependency and accountability.
- 38:38Ask who owns the data? What lock in am I creating?
- 38:42Can I explain this decision? The three questions from before?
- 38:54So I'd like to leave you with a question, the one that think
- 39:01that I think actually matters. The crucial question is not
- 39:06whether AI can think for us, it's whether we stop thinking
- 39:10for ourselves because it's become too easy not to.
- 39:16Thank you. Wonderful.
- 39:20Thank you so very much, Mark. It was really, really very
- 39:23interesting. Great, I thought.
- 39:27I thought it it was yes. OK, then let's go to the our 3rd
- 39:39guest, which is Torsen. Torsen will help us understand
- 39:45how these changes affect our trust and how we truly feel.
- 39:51So let's welcome Torsen. Here we are.
- 39:55Hello, Hello. Welcome, Torsen.
- 39:58Well, thank you for for me to being here.
- 40:02So very much accepting my invitation.
- 40:04So lovely. Thank you.
- 40:05I really appreciate it very much at. 1st I admire my colleagues
- 40:10English, my Angel ain't so good. Again, I'd be better psychology
- 40:18than English, but bear with me I while I search for words.
- 40:23So hang in there. I try to give my best.
- 40:27I'm sure you will and I'm sure that also people will truly
- 40:30understand and they will be patient enough to hear you out
- 40:34what you have to say. So don't worry about it just.
- 40:37And if they don't and if they, if they don't understand me, I
- 40:40have this all in the actual materials, you know.
- 40:43Yeah, yeah, of course. Exactly.
- 40:44They will have everything if they will sign up, register to
- 40:49this event on the website that I have actually at the beginning,
- 40:53I have shared the QR code. They also have the link in the
- 40:56comments. Guys sign up and or register to
- 41:00our website. So you can actually get get all
- 41:03those tools, ideas, articles, even the first chapter of Mark's
- 41:10book, all this for free. So just sign up because
- 41:14otherwise you will miss everything out.
- 41:16So you will not receive it. You will not have it.
- 41:19Thank you. So I'm going to let you first
- 41:21into your presentation now. Good luck.
- 41:25Yeah, thank you. Thank you.
- 41:27I must see what I can do here. Nope.
- 41:34Nope. Yeah,
- 41:56technician. I'm a I'm a psychologist, not a
- 42:00technician. Yeah.
- 42:05Oh, there it is. There it is.
- 42:11Now, what are you seeing here? Well, AI is not only about
- 42:26speed, you know that. And AI is not only about
- 42:31efficiency at all. And it changes us in ways we
- 42:36can't imagine yet, but it's, but we are working on it.
- 42:48Yeah. It's changing the way that we
- 42:53that we notice things around us and the changes to things that
- 43:00we are reacting to. And and this of course, what we
- 43:07react to was also having trust the things we see and perception
- 43:14and the perception of it. So and then when, when, when
- 43:22those things change around us, it also change the regulation in
- 43:30US so we so we can get out of focus.
- 43:35So for me the relation is not what AI can do.
- 43:43My the issue is for me is what AI does to us as humans.
- 43:50Because the design of the AI thinks every everybody reacts
- 43:56the same. But we we are the AI meets the
- 44:04person, the person, not the human species and the person can
- 44:09be like me have PTSD someone ADHD and something but I come to
- 44:18it later on yes how to go Oh yes, some technical technical.
- 44:44My buttons don't work. I'm sorry, I don't know why it's
- 45:05don't work. Well, have the trust on this
- 45:22system isn't good. No, I use too much time for
- 45:37this, maybe I can do it. So Oh yes, I'll try again.
- 45:47One moment, please. It's everything OK.
- 45:52I've got some technical issues here.
- 45:58This happens. Yes, exactly.
- 46:04OK, let's see. It is this one, is it?
- 46:07Yeah, I think I know, I know. I have it.
- 46:13It's the first time for me, you know.
- 46:14Yes, yes, yes, of course. I forgive you.
- 46:16You're forgiven. I hope you can see this going on
- 46:22here. Well, go on.
- 46:27Mark has been a hope address hard to trust you all see it
- 46:36outside out in the Somi media and every everywhere else.
- 46:41It's yeah, we live in a time of deep face.
- 46:46We say the Turk voices with narrative manipulations.
- 46:51So somebody want to move you on to this side or to the other
- 46:57side to so to think, to have your opinion about something.
- 47:06Well, and let's let's make it harder to know it's real if
- 47:11you're what you are seeing and what you are hearing and what
- 47:13you are meant to, to think. And, and if people don't know
- 47:22what's real, they they don't feel safe what's happening
- 47:28around me, you know, and then this matters.
- 47:32This matters very much because it's not only social the the
- 47:39trust is also in your in your body, it's biology.
- 47:46So it affects the nervous system and this is a variety.
- 47:56It feels unstable. The body don't do less, you
- 48:03know, so, so This is why the trust of of what you're seeing
- 48:12can can go in your body and it's reacting to to and say this is
- 48:18this in truth, I, this is not me.
- 48:21It don't resonate. Yeah, yeah, no, again.
- 48:28Again the same problems. Are you using your mouse right
- 48:47now? Torsen or.
- 48:49Yeah, I don't know what I'm doing here.
- 48:56I have to laugh a little better around it.
- 48:58Yes, of course. Next time is there.
- 49:00Definitely not crying I need. To get this.
- 49:19I don't know. Should should I have a proposal?
- 49:25Should. Since KA has your presentation,
- 49:28should KA share your presentation and you can talk
- 49:32about it? I can say next one please.
- 49:35Yeah, yeah. We'll do it that way.
- 49:38I'll I'll just just a second, then I'll I'll show it.
- 49:58Is it this one? No, this is chosen.
- 50:03I have it in a minute, I'm just opening it now.
- 50:12Well, I must ask AI how to do it.
- 50:18Well, I had a pump or something. Yeah.
- 50:28Right. So I will see in the screen just
- 50:34a second and you just tell me when to to
- 50:51go to to. The yes slide #3 please.
- 50:59Yeah, no, it's on I call. Well, whatever right here in my
- 51:09words, yes, this is the slide #3 the fillet feats.
- 51:21You know, it's shape what we see emotional and I don't know what
- 51:31does it means. You know, amputation shapes were
- 51:37what feels urgent attention hijacking shapes what we capture
- 51:43in mind. This is the first.
- 51:48So AI does not only show us content, it only it also shapes
- 51:54perception. What I'm talking about before
- 51:58the perception. And the perception matters
- 52:03because what we are seeing again and again starts to feel
- 52:07important, normal and true. But is it true?
- 52:13How do we, how can we see if it's true?
- 52:16How we can, how, how can we find, find out if it's true?
- 52:21That what shapes, yeah, what shapes the attention also saves
- 52:28our lives in that in that manner.
- 52:32So the attention, the hijacking, the what are you calling the
- 52:45click baits? You know, so, so, so it talks to
- 52:51your to your inner emotions with the with the click baits.
- 52:56Yes. Next slide please.
- 53:02Now again the nervous system respond.
- 53:05In any case, some have lower morale than others and feel some
- 53:15somebody have a high sensitive personality and it depends on
- 53:20how you work with the AI. Again, it's different from
- 53:24person to person. Nobody is the same.
- 53:28So we have to make systems who know the user in any way.
- 53:36Not, not not the, not the overtake the user, but know the
- 53:39user, our things and and and and and and the master.
- 53:51Yeah, Minster on Danish. Yes, yes, yes, yeah.
- 54:05Well, when I see things on the Internet, my body always, always
- 54:14my nervous system always react to something.
- 54:17And then I look, what is this? Is this right?
- 54:20Is this wrong? What does it meant for me and do
- 54:28I need to think about it is it doesn't matter, but we this is
- 54:40an uncertainty of whether intensity would react to
- 54:49emotionally it it it it goes to an overload and you can't you
- 54:58can't stay in there very long before you get like burnouts
- 55:03depression in any way if you use tools or to name not on the
- 55:16right way. Let's take slide 5.
- 55:29Yeah, what is real, what is real, what is relevant, what is
- 55:38mine and what can I trust? Call us all over it.
- 55:44And and Mark 2 now I'm here saying the same thing.
- 55:53So it has something to do with it.
- 56:05Well, in any case, these are not small questions.
- 56:12That does matter if they appear once at management, if they
- 56:18appear all the time, those questions, it's going to be
- 56:23stressful. It's going to overload your
- 56:25nervous system. You ain't the machine if you are
- 56:29human. I don't know if, if this station
- 56:36development, we, we can somebody talk about the singularity.
- 56:41I think it's more about whether human meets us.
- 56:50It's it's part of it's what it can do as a human, what it can
- 56:54bear. The machine ain't a human.
- 57:00It goes wrong as a, as a train, a trigger train who triggers
- 57:07everything in you and shows your, what you are doing wrong
- 57:11and shows your, your, yeah, it shows your, your, your shadow
- 57:19sites, what you are lying about yourself.
- 57:22No, it's not everybody who who can who can manage this and they
- 57:29can get some AI fatigue out of this, you know, and and this
- 57:39further away leads to to to to to to get to resist the the
- 57:49development to be to be against AI because it has has destroyed
- 57:5611's self picture. Yeah, slide number six, please.
- 58:06But not everyone experienced the same way.
- 58:12I told her about it before. Some people have to have higher
- 58:16higher sensitivity. Some are new age evidence.
- 58:21I hope I expect spell it only pronounce it well.
- 58:26Some have cognitive vulnerability, some simply
- 58:30regret more like 10 times more than other people's.
- 58:38So the same in AI environment can affect people very
- 58:43differently. One person may experience this
- 58:48youthful, another may experience as an overlord like I talked
- 58:54about before, because the design often assumes that all users
- 59:00affect the same way and it doesn't affect the same way.
- 59:04Yes, the same system does not feel the same nervous system and
- 59:11everybody in every person. Yeah, I think it's enough for
- 59:16the slide. Let's take the slide now.
- 59:22This mark was in two, two, and I think this is a, it's a we're
- 59:31trying to do. Yeah.
- 59:38Well, it can become an awareness assistant, as you can see.
- 59:45You can, well, you can simply ask questions like what tone is
- 59:51this text? What biases in this reasoning,
- 59:57and can you rewrite what I wrote here in a comet tone?
- 1:00:06Somebody very angry sometimes, and they just do this and better
- 1:00:13tone. Yeah.
- 1:00:18The eye becomes less a machine for speed and more a mirror,
- 1:00:24mirror for awareness. Oh, everybody makes it faster,
- 1:00:29faster, more, more, more. The quality, the precision goes
- 1:00:34down the hip. And the this is how this
- 1:00:39matters, too. Yeah.
- 1:00:44So the AI can reduce confusion. It can increase clarity and help
- 1:00:55people slow down and notice use well, AI can support reflection
- 1:01:03instead replacement. So so this is very good.
- 1:01:09It's it's in the I've been trying to put it all together in
- 1:01:14the prompts. You can get in in the extra
- 1:01:18material. Yes, Slide 8, please.
- 1:01:23Well, this is the last slide for me.
- 1:01:30It's very, very easy. It's always, it's, it's about
- 1:01:34Christians, it's about what AI do to us.
- 1:01:41What does it do to our trust? What do does it do to our
- 1:01:45attention? What does it do to regulation?
- 1:01:50What does it do to our sense of reality?
- 1:01:54Those questions are important because we without, that's why
- 1:02:03awareness, our awareness about this must must come before the
- 1:02:09blind use of AI. It's it's it's it's not good to
- 1:02:13use AI blind use because without awareness, power becomes
- 1:02:18influence. It's always become the
- 1:02:20confusion. The confusion becomes stressed.
- 1:02:24It's not good. So for me, the most important
- 1:02:29question is not only capability, it's a human impact.
- 1:02:36Thank you. Fantastic.
- 1:02:44Now let's get together, all of us.
- 1:02:53Okie, Dokie, just one second. Hello again.
- 1:03:03Good job, Thorzen. You have two, I have two screens
- 1:03:07so and you all used to work with one screen.
- 1:03:15It was a great presentation and a good team work.
- 1:03:20Yeah, definitely. So I just want to remind to all
- 1:03:25of you guys who are watching today that make sure to sign up
- 1:03:30on our website of today's events.
- 1:03:33Only those who actually subscribe to that website will
- 1:03:38get the full presentation of all three specialists delivered
- 1:03:41straight to their inbox. Don't miss your chance to keep
- 1:03:45this expert insights forever. Moreover, you will have the
- 1:03:49opportunity to download the Reset protocol as well, a
- 1:03:53specific cognitive framework to regain clarity and keep your
- 1:03:57focus during digital noise and the AI Resource Guide as well a
- 1:04:03toolkit curated by me and the guest specialist.
- 1:04:08So yeah, you can use this QR code shown on the screen or and
- 1:04:15you also have the link, direct link into the comments section.
- 1:04:20So feel free to use it anytime. Yeah, I think now we can go.
- 1:04:30I can hear an echo. Why?
- 1:04:35That's the weird. I think we can go to the Q&A
- 1:04:42section just one second, yeah. I think by default we should
- 1:04:51mute our microphones. If we're not talking, then we
- 1:04:54should avoid that echo. Yeah, I'm not quite sure why is
- 1:04:58that happening, but yeah, I don't know.
- 1:05:03OK, OK, OK. So we are at the Q&A section.
- 1:05:10So let's start with the first one, which I have.
- 1:05:20We received a question from Kim, Kim Meyer.
- 1:05:26I hope I pronounced it correctly.
- 1:05:28I apologize, I don't do it on purpose.
- 1:05:32So since it the message was in Danish, I have translated, I'm
- 1:05:36going to translate it in English.
- 1:05:37Of course, it might sound like this.
- 1:05:41So Kim Meyer says, how do we make AIA real help for ordinary
- 1:05:47people in everyday life, so that that technology does not become
- 1:05:54a new barrier with difficult technical language and
- 1:05:58requirements for technical knowledge, but instead a tool
- 1:06:02that is safe, understandable and human?
- 1:06:06And I think this is for Mark. Yeah, Let me just mute one
- 1:06:12second there. I think it's a very interesting
- 1:06:15question and it's it's sort of two different questions.
- 1:06:19How do we make sure that the technology is accessible and and
- 1:06:24we're able to use it even if we're not that tech savvy.
- 1:06:28And I think the answer to that is that is sort of solving
- 1:06:31itself. The technology is becoming more
- 1:06:34user friendly every day. So, so that will not be that
- 1:06:38much of a problem. The problem is, is actually the
- 1:06:43trust and the literacy as we talked about it turns into sort
- 1:06:49of if we outsource our decision making and so on, it will along
- 1:06:56the way erode our judgement and in a sense our freedom to think.
- 1:07:04And that's a much more, that's a much more difficult thing to
- 1:07:09combat. We've talked a bit about this,
- 1:07:14about how we do it, how we, how we use the, the technology as a
- 1:07:23partner and not as a solution in itself.
- 1:07:26And we, we keep our own ability to think, so to speak.
- 1:07:33So, so how do we make AI safe is really by demanding it, both by
- 1:07:44making an effort of ourselves in order to preserve our way of
- 1:07:48thinking, but also by demanding of the people who make it that
- 1:07:53this is the kind of technology that we want.
- 1:07:57We have to demand it from the people who built it, and we also
- 1:08:00have to demand it from the people who regulate it, who set
- 1:08:03the rules and and and not least, not least from the people who
- 1:08:08choose to deploy the technology. Yeah, definitely.
- 1:08:12Thank you so very much for from the same person again, Kim
- 1:08:19Meyer. He has an A second question he
- 1:08:24asked. And how do we also avoid that
- 1:08:27the use of AI leads to dependence, loss of judgement,
- 1:08:32less control over our own data, and a quiet erosion of some of
- 1:08:38our freedom. I think this is might be for
- 1:08:42Kerr. Yes, it is really a dilemma
- 1:08:51because the case is, I know Kim, he's a, he's a great guy and
- 1:08:58he's very curious about these tools.
- 1:09:00He benefits a lot from these tools.
- 1:09:03And I can see that it is very helpful for him that he can
- 1:09:08speak directly to the tools. He can point the camera in a
- 1:09:12direction and it will understand what it's all about.
- 1:09:15But of course, it takes safe technology and it takes relevant
- 1:09:18subscriptions. And there are the considerations
- 1:09:22about data in this specific situation, as I also have an
- 1:09:28ongoing dialogue with the with the Kim.
- 1:09:31He has chosen in some situations to upgrade his subscriptions
- 1:09:35both to get relevant capacity and to get relevant, relevant
- 1:09:39functionality. And it is well about navigating
- 1:09:45and having the next step. So I can see that we can benefit
- 1:09:52a lot from these. Well, let's start like call them
- 1:09:55being questions and answers machines, but it's really still
- 1:09:59about navigating and it's still about having people around in
- 1:10:04order to help you navigate. I think Kim is on safe ground
- 1:10:09and it's such a pleasure to see his curiosity and his benefit
- 1:10:13from this technology. Thank you.
- 1:10:18So we have some other questions, interesting questions which I
- 1:10:22loved. We have someone called Sif
- 1:10:27Muller Gamelhoi. Again, I hope I pronounce it
- 1:10:31correctly, I apologize. OK, he says.
- 1:10:36An angle that I also think could be interesting to dive into
- 1:10:41within my field. Involving men in the dating
- 1:10:44world is not just the work related aspect, but especially
- 1:10:49the human side of it, where more and more people are experiencing
- 1:10:54having an AI partner and what that entails in terms of
- 1:10:59challenges regarding communication and some other
- 1:11:03things. Well, Mr. Sif, I do think
- 1:11:07Congrats because I, I think this is a really brilliant question.
- 1:11:11Even though we might not seem so, I do believe it is.
- 1:11:16And I will answer to this if you don't mind, guys, because it's
- 1:11:20kind of challenged me in such a way, not only because I'm a
- 1:11:24cognitive coach, but also because I'm single for a long,
- 1:11:27long, long time. So, yeah, life experience, you
- 1:11:30know, so see if I think it, this particular question connects
- 1:11:37sickly to something really deeply human, which is the
- 1:11:42illusion of connection. We are, since we are human, all
- 1:11:47of us, right? We need connection.
- 1:11:50And from a cognitive perspective, our brain brains
- 1:11:54are wired to productive friction.
- 1:11:57For example, in a relationship, we have to navigate moods,
- 1:12:01disagreements, and growth. And that friction is actually
- 1:12:06what builds our emotional resilience and identity as well.
- 1:12:11So an AI partner is programmed to be perfectly agreeable,
- 1:12:15right? That means that it gives you the
- 1:12:19right affirmation all the time, constantly.
- 1:12:23It's like looking into a mirror. And that only shows you what you
- 1:12:28want to see. And from a cognitive
- 1:12:30perspective, we stop practicing the muscle of empathy because
- 1:12:35the AI never challenged us, right?
- 1:12:37Makes sense. So in the dating world, if men
- 1:12:41or anyone else rely on AI for intimacy, they risk to losing
- 1:12:48the ability to handle the beautiful, messy
- 1:12:51unpredictability of a real human being.
- 1:12:55So since we rewire ourselves to prefer perfection of a presence,
- 1:13:01and that is a very lonely place to be.
- 1:13:05This is my personal answer to him.
- 1:13:10But thank you for your question. I think it was really, really
- 1:13:13cool. So the same person, Sif also
- 1:13:16have a second question. Let me see here.
- 1:13:23He he says humans have always tried to optimize work flows
- 1:13:29from the wheel to the calculator to the washing machine,
- 1:13:33etcetera. But the difference is perhaps
- 1:13:36that there isn't as much concern regarding safety when it comes
- 1:13:41to the wheel as there the wheel as there is with AI.
- 1:13:46Because where does where does the data go and what is it used
- 1:13:51for? Very few people are afraid of
- 1:13:54their data from the vacuum cleaner being misused.
- 1:13:59Well, I think this is for Curve. Yes, and you're right, I mean
- 1:14:08some data are are not problematic at all.
- 1:14:11The, the example with the vacuum cleaner is actually quite good.
- 1:14:15And, and it seems that products that were not previously digital
- 1:14:20starts to being digital and they collect data and who knows where
- 1:14:27they are sent. I can see for, for my part, I,
- 1:14:31I've sometimes served the, the military.
- 1:14:34I shouldn't go into details because part of that is
- 1:14:36confidential, but I know that there is a very big interest in
- 1:14:42what cars are actually entering those areas because there are
- 1:14:47cameras everywhere and, and, and data is sent to service servers
- 1:14:51and, and, and where are those servers located.
- 1:14:54Currently there is a big interest in in in Chinese
- 1:14:58electric cars and it seems that they communicate with servers in
- 1:15:02China. And as it is so much connected
- 1:15:06to the state, well, it is unpredictable.
- 1:15:09So it, it, it really is a trade off.
- 1:15:13I think part of it points back to to my point about choosing
- 1:15:17the, the proper system, because some systems are trustworthy.
- 1:15:22Some systems can be put with a guardrails.
- 1:15:27You can even create your own system if you have the, the, the
- 1:15:31money for it and, and the resources for it.
- 1:15:33That's not always an option, but for for some organizations it
- 1:15:36will be. And, and there are really a lot
- 1:15:39of dilemmas here because the convenience is convenience is so
- 1:15:43tempting, but there are two sides of of that coin.
- 1:15:50Thank you. So the next song will be from
- 1:15:54Jan Mikhail Khanna. He says the first question that
- 1:16:00he asked is about market structure.
- 1:16:05Will AI centralized power on a few platforms or create new
- 1:16:09decentralized markets? And the second one is which
- 1:16:14industry will experience the greatest structural
- 1:16:18redistribution of value? I think this is for you.
- 1:16:22Yep, definitely. So again, he said he asked will
- 1:16:27will AI centralized power on a few platforms or create new
- 1:16:31decentralized markets. And the second one is which
- 1:16:35industries will experience the greatest industrial
- 1:16:38distribution, this redistribution of value?
- 1:16:43Well, let's let's start out from the platforms.
- 1:16:46Currently more and more points in the direction that the big
- 1:16:51players in this market are really the big tech players that
- 1:16:54we already know. When we saw Open AI, that was a
- 1:16:59new player in the market, but it was initially supported by Elon
- 1:17:05Musk. It was afterwards supported by
- 1:17:07Microsoft and and afterwards as Microsoft and Google and mates
- 1:17:13got up to speed. Well, those are are basically
- 1:17:17the the usual suspects. We do see new players in in this
- 1:17:22area and I hope that we will see a place for new players.
- 1:17:29When we saw the, the new Chinese strong model, it, it had a price
- 1:17:39tag of 1/10 of what would usually be the price of creating
- 1:17:43a new model because you can really learn from existing
- 1:17:46models. So I think there is still a
- 1:17:49possibility for, for an open market.
- 1:17:51I can also see that the, the French guy Yan de Kun, who for
- 1:17:58many years was the the, the leader of the, the, the major AI
- 1:18:03my activities has now created his own company located in
- 1:18:09France. So we might see new European
- 1:18:12players. So those are about the
- 1:18:16companies. Many other users suspects, some
- 1:18:20are new players and I hope that's still the room for new
- 1:18:23players. What was the second part of the
- 1:18:27question once again? Which industries will experience
- 1:18:31the greatest structural redistribution of value?
- 1:18:35Yeah, It seems that it starts out from, let's call it office
- 1:18:40work. Really the more digital it
- 1:18:43already is, the more it is possible to add this on top.
- 1:18:47I can see that it has changed marketing a lot.
- 1:18:50It has changed legal services a lot.
- 1:18:53It has changed some customer service segments, but little by
- 1:18:59little, as we also see human humanoid robots powered by AI
- 1:19:06and as we see AI entering other kinds of systems.
- 1:19:09I mean, even in, in, in the military industry.
- 1:19:13Well, it's hard to predict what will be next, but it seems that
- 1:19:18what can be done by a computer can really be done by by AI.
- 1:19:22What can be done mainly by a human being will mainly maybe be
- 1:19:27done even better by a human being with with the help of
- 1:19:32relevant tools. And those tools need not only be
- 1:19:35about AI, but but of course it can be part of the package.
- 1:19:39Thank you, Nicole. So the next one, it's more
- 1:19:44related to leadership. I can see what skills do leaders
- 1:19:48need to develop to not only implement AI, but manage its
- 1:19:53consequences. And the second one, always ready
- 1:19:59to leadership, how do you balance innovation and
- 1:20:02accountability? I think this is for Mark.
- 1:20:08Yeah, I'll try try and answer that.
- 1:20:10So again, what skills or what skills do leaders need to
- 1:20:14develop, to develop, to not only implement AI but manage its
- 1:20:18consequences? This is the first question.
- 1:20:23I'd say one of the most important things is sort of
- 1:20:26humility, knowing what you don't know and being honest about it.
- 1:20:34I think most AI failure occurs in implementation because
- 1:20:39systems are implemented that is not fully understood.
- 1:20:43So you have to, you have to be able to say, I don't know, we
- 1:20:46have to find out. A second skill you need is is
- 1:20:51what you might call consequence thinking.
- 1:20:56The ability not to just ask just does this work, but what does it
- 1:21:01actually do to the people it affects employees, most
- 1:21:06certainly, but also customers, citizens and and communities.
- 1:21:10As such, the third ability and I'd like to I'd like to use a
- 1:21:18third ability here is because it's important in my opinion,
- 1:21:23made it the most important is the ability just to stay in the
- 1:21:26room to actually be present with the with employees and also when
- 1:21:36you make the decisions. So true.
- 1:21:40So I I agree. The second one was how do you
- 1:21:43balance innovation and accountability?
- 1:21:49Yeah, sometimes things that think that that too much
- 1:21:54importance is put into innovation in that balance
- 1:21:59between the two, because I actually think that something
- 1:22:01like accountability can be a competitive advantage as well.
- 1:22:06So they it doesn't need to be opposites.
- 1:22:11I think you need to value both sides OK.
- 1:22:16The same person Jan asked. Also he's best of highlights the
- 1:22:21fact that it's more for a debate.
- 1:22:23He asked. Companies can develop strategic
- 1:22:26dependencies on few AI vendors. I think this is for.
- 1:22:35It is surely a a theme. I mean before this all took took
- 1:22:40took speed, Microsoft was already a strong player and now
- 1:22:45Microsoft is grabbing even more systems and even more customers
- 1:22:49and even more dependency. The same goes for for Google.
- 1:22:54So I mean during my presentation I presented as an alternative
- 1:22:59that it is possible to actually have your own model, download it
- 1:23:04and and there are strong models for free.
- 1:23:07Install it on a relevant server. Make sure that that you train it
- 1:23:10for the purpose of the the organization.
- 1:23:13And in some situations that will actually be the need, especially
- 1:23:17when there is a big level of confidentiality.
- 1:23:24But well, sometimes you get the convenience, sometimes you get
- 1:23:29the functionality and the price tag will affect so, so it's
- 1:23:33really well, one factor affects the the other.
- 1:23:39Thank you, the same person has the last question, the 4th one.
- 1:23:46AI is not only changing jobs but also how humans understand their
- 1:23:51role and value in work. And I suppose this is for
- 1:23:55Torsen. Yeah, can I hear it once more?
- 1:24:03Yeah, sure. So he asked.
- 1:24:05AI is not only changing jobs, but also how humans understands
- 1:24:12understand their role and value in work.
- 1:24:17Yeah, well, I don't hear the question.
- 1:24:19It's more a statement, but it's but I'm but I'm.
- 1:24:26Might be that he's or she confused or something like that
- 1:24:30and might need some, you know, advices or tips or whatever.
- 1:24:36Yeah, but what he's saying, I give him, right?
- 1:24:40Yeah, that's right. What he's saying this, it
- 1:24:44changes, changes our way to work in our free time.
- 1:24:49We, we, we can't sit there and then look in the air.
- 1:24:53So we take another job and another thing and this and
- 1:24:59that's, that's why it changed the work.
- 1:25:01It's changed how we are working with it every day.
- 1:25:06So we get we have to have a lot of free time on the payroll.
- 1:25:15Thank you. We have a question for you,
- 1:25:22Torsten from Nikolai. There is no way that AI is going
- 1:25:31to change our cognitive ways. Is there a path that makes
- 1:25:36sense? And is it good or bad from
- 1:25:40whatever standpoint you wish? Well.
- 1:25:52Yeah, this is. Yeah, I can't, Can't.
- 1:26:00Could I read it again? Yeah, please.
- 1:26:04Yeah, sure, So sure, he says. There is no way that AI is is
- 1:26:13going to change our cognitive ways.
- 1:26:16Is there a path that makes sense and is it good or bad from
- 1:26:22whatever standpoint you wish? Yeah, from my my standpoint is
- 1:26:28how how I work with it. It's a, it's a, it's a, it's a,
- 1:26:32it's a good, good one because it can help us in, in this way, we
- 1:26:38can someone who, like me, have a difference to, to, to make
- 1:26:46myself clear. So and that way my, my, my for
- 1:26:52what I mean will become translated.
- 1:26:57So I give sense to to the people around.
- 1:27:01And so it's also also with the with the cognitive abilities,
- 1:27:06but but we need to have some some framework in it so that
- 1:27:11don't take the human part of it, but teach a human how to use the
- 1:27:16human parts if it gives sense. Yeah, it makes sense for sure.
- 1:27:23Hopefully. I hope, I hope it's enough for
- 1:27:26this, Nikolai. I will find out.
- 1:27:33Yeah, they can still. Anyway the guys, you can still
- 1:27:36add some comments in in the comments section no problem if
- 1:27:41it's not. And, and, and, and addition to
- 1:27:43this, I, I, I will, I would invite people to dive into it,
- 1:27:51what it does to, to us as, as human.
- 1:27:56Everybody may maybe write something about it, but but it's
- 1:27:59very important we we get that knowledge.
- 1:28:02Thank you so very much, Dawson. So we have an anxious keynote
- 1:28:07studio here asking a question. Thank you.
- 1:28:11By the way, he says what happens when AI creates massive
- 1:28:17unemployment? What will people do and how will
- 1:28:21they earn money? Well, quite interesting, simple,
- 1:28:24but very too strict to the point.
- 1:28:26I think it's very important, guys who wants to.
- 1:28:29I have an answer for for that. Maybe I'm not the only one, but
- 1:28:33can I go first? Yeah, yeah.
- 1:28:37Look, at least on the short horizon, AI hasn't yet created a
- 1:28:43massive unemployment. It has improved efficiency.
- 1:28:48It has mixed up some some things, but it has actually,
- 1:28:55well also created new jobs and at least on the short horizon,
- 1:29:00it seems that for instance, programmers can now program more
- 1:29:05efficiently and there is such a big need for programming.
- 1:29:09So the volume is just bigger, the price point is lower.
- 1:29:12We see the same thing for legal services.
- 1:29:15Legal services have such a high price point that many legal
- 1:29:20services are not taking into consideration in the 1st place.
- 1:29:24And now with the lower price point, you will actually get
- 1:29:27more. So it seems that we will
- 1:29:30basically get more. And it seems that there is such
- 1:29:34a big need for actual human hands.
- 1:29:37I see it in the healthcare sector, I see it in schools, I
- 1:29:42see it in many places where people actually meet.
- 1:29:46So I I think that it will create new jobs that will restructure
- 1:29:53existing jobs, but it doesn't seem to create unemployment on
- 1:29:59the long horizons. It is.
- 1:30:01It's of course hard to predict. Yes, Mark.
- 1:30:07Yeah, I agree with the core. The science right now is isn't a
- 1:30:11mass unemployment, but we also need to sort of find the balance
- 1:30:17of the employers need to find the balance in how we do things.
- 1:30:20I hope most employers will actually choose the path of
- 1:30:23restructuring jobs as as core is pointing out that we are looking
- 1:30:28at what, what jobs are human good, humans good and what are
- 1:30:32the machines good at, so to speak.
- 1:30:34And, and, and so maybe it won't create mass employment, but it
- 1:30:38could also be an opportunity for us to maybe work less and that
- 1:30:43would mean that more people could actually get, get employed
- 1:30:46along the way. I also think we need to look at
- 1:30:52solutions like universal basic income and that sort of thing in
- 1:30:56connection to this. But it's it's, it's a complex
- 1:31:00issue and all things have to be brought into the discussion.
- 1:31:05Yeah, I agree totally. But there there is so much to
- 1:31:09say here like definitely. Also I agree.
- 1:31:13I have another one from the same person Keynote Studio.
- 1:31:20So he or she says AI makes us think less.
- 1:31:25Over time the cognitive brain weakens and the limbing system
- 1:31:33takes over, increasing anxiety. That is why I teach the limbing
- 1:31:38system. I think Chosen might be for you.
- 1:31:41Yeah, the limbic system. Yeah, of course the machines
- 1:31:49takes over the thinking of people.
- 1:31:52They, they, every time they have something in mind.
- 1:31:56Oh, I must ask my AI what I should do, what I should do.
- 1:32:00This. I want to think.
- 1:32:03And what we need to slow down with it is go, go, go, go.
- 1:32:08Much, much, much too fast. So we out search our thinking
- 1:32:12process to the machine, but we need to again have some
- 1:32:17framework so the machine teachers to think, not to give
- 1:32:22the thinking process to the machine, to the AI.
- 1:32:27That's what I must say to this. Thank you.
- 1:32:30Thank you so much. We have another one for postural
- 1:32:34feet. Don't you think we will be too
- 1:32:38controlled in the future by using AI too much?
- 1:32:44Yeah, Mark. I I completely see what the what
- 1:32:51the person asking the question means.
- 1:32:54I think it's very much a matter of what kind of direction we
- 1:32:58choose. I think this has to do with the
- 1:33:03our choice of how we want to use AI.
- 1:33:05What what do we, what we wanted to do?
- 1:33:11And we have to find the balance and as Torsten says, find a
- 1:33:15framework for using AI. So will we be too controlled?
- 1:33:21We could be if we let it happen. And I think the keyword here is
- 1:33:27LED it. We have to choose not to.
- 1:33:32So we have from the same posture on feet, another question how we
- 1:33:36can how, how we can use artificial AI in a positive way.
- 1:33:45Yeah, yeah, more. Yeah, I think we've, we've
- 1:33:49talked about it all three of us in our presentations to, to use
- 1:33:53it to challenge us in, in our existing thinking, use it to to
- 1:33:59create ideas, not to make decisions, but use the
- 1:34:03technology to make us better basically.
- 1:34:08Yeah. Well, OK, we have another one
- 1:34:15from from Keynote Studio again. So he or she says 99% in Denmark
- 1:34:24are not really using AI. They only see it as a simple
- 1:34:29tool. We are for far beyond behind,
- 1:34:33especially compared to country like China.
- 1:34:37This is very interesting. Thank you for your question.
- 1:34:39Please guys, the floor is yours. Yeah, Yeah, Mark.
- 1:34:47I guess I could start out. I think 99% is is a very high
- 1:34:53figure. Yeah, I think so too.
- 1:34:55Yeah. When I'm out talking to people,
- 1:34:57I, I, I, I meet quite rare that I meet people who doesn't have
- 1:35:02some kind of interaction. Me too, same.
- 1:35:05Here with these questions, but, but it could be that, that on a,
- 1:35:10on a general level, we are, we're not as engaged as a
- 1:35:16country like China. But you also have to remember
- 1:35:18there, there are different rules when we're talking China and,
- 1:35:22and we have a much more voluntary based system in, in a
- 1:35:28country like Denmark. So, but, but, but when that is
- 1:35:32said, I think you also have to say that that a country like
- 1:35:35Denmark has a much has a really good starting point because
- 1:35:42we're, we're used to using new learning, new system, using new
- 1:35:46systems, integrating them in existing structures.
- 1:35:50And I think that is what's going to happen here as well.
- 1:35:56Someone else want to add something else?
- 1:35:58Yeah. I often have a dialogue of, of,
- 1:36:04of this kind when, when I do courses and I, we often talk
- 1:36:10about whether we live in the right part of, of, of the world
- 1:36:16or the wrong part of the world. And, and it really depends
- 1:36:19because it, it is true that China is using technology in
- 1:36:24much different ways than the EU. From my point of view as a Dane,
- 1:36:29I think we do it in a better way.
- 1:36:31I think we do it in a, in a more respectful way to, to humans.
- 1:36:36I, I can see that we can miss out on efficiency and, and, and
- 1:36:41the pros and cons to everything. But I really, I, I like the, the
- 1:36:45European approach. I think we, we also focus on the
- 1:36:52potential problems and, and how we will still protect human
- 1:36:58rights and, and, and, and, and, and personal information and,
- 1:37:03and that is under pressure in, in many parts of the world.
- 1:37:06I mean even in, in, in Denmark. But I hope we can we can keep on
- 1:37:11a good track in that aspect. Torsten, would you like to add
- 1:37:16something as well or? Yeah, I can say that most of the
- 1:37:22users are the people in Denmark I have spoken with.
- 1:37:27They see AI as cheat, as something to make fake movies
- 1:37:37about a mouse dazing a cat with a baseball bat or anything else
- 1:37:46I don't know. Influences make a lot of this,
- 1:37:54Put it on the net, make overload of it.
- 1:37:57And we get in our and our faces and we have this what is this
- 1:38:04what I'm looking at? And that's you.
- 1:38:09You can see in that that far most of the people don't know
- 1:38:14what they are holding in their hands, this power, real power.
- 1:38:21And maybe, maybe it's, it's a little bit of scary the, the
- 1:38:28people don't know what they, what they have in the, in the
- 1:38:31hands. And this is not to talk down to
- 1:38:34people. It's to talk up to them.
- 1:38:36It's to, to show them how, how, how, how, how they use, how they
- 1:38:40can use this kind of of tool to every good benefit for the
- 1:38:45humankind. Not, not only you think about
- 1:38:48the money, it's the money, always the money.
- 1:38:52Who, who, who makes things worse in in in some kind of this.
- 1:38:59And we need to get the focus on the human behind the screen.
- 1:39:05We use it so we can benefit from the AI and maybe earn some money
- 1:39:15off it. Yeah.
- 1:39:17Thank you so very much to us. Yeah, yeah, yeah, Mark.
- 1:39:21Yeah, just to say that maybe it's not such a bad thing that
- 1:39:24we're using AI in a different way compared to China, for
- 1:39:28example. Maybe we need the more
- 1:39:31afterthought like Tosten says, we need the human IT to be in
- 1:39:36the focus of our thinking and and then learn how to use the
- 1:39:40technology from that standpoint. I think the risk of just blindly
- 1:39:46implementing and forcing people to use AI would be, is also what
- 1:39:51we've been talking about this, this confidence of outsourcing,
- 1:39:54this losing of, of mental abilities that we that we need
- 1:39:59as humans. So maybe it's not a bad thing
- 1:40:02that things are a bit critical about this.
- 1:40:04And like I said, they're probably faster than most people
- 1:40:07at getting the system once we figure out how to use it.
- 1:40:13The same person added as well as again Keynote Studio.
- 1:40:19So they say 99% use AI casually or not at all.
- 1:40:26It is only the 1% where businesses use it probably to
- 1:40:31create structural efficiency and real results.
- 1:40:36What do you think guys? Yeah.
- 1:40:41I think part of it has to do with the structure of of Danish
- 1:40:46companies, because I can see that big companies in most cases
- 1:40:52work a lot on this. But Denmark does not have all
- 1:40:56that many big companies, at least not on on an international
- 1:40:59scale. So we're so, so it has to do a
- 1:41:03lot with with really how we structure our, our professional
- 1:41:07market for the good and for the well, whatever it is.
- 1:41:12I mean that pros and cons. Yeah, someone else wants to add
- 1:41:17something else. OK, from the same.
- 1:41:20Let's take the last one and then we have to close here again.
- 1:41:25Keynote Studio Dynamo has around 300,000 businesses and nearly 6
- 1:41:30million people. Most people are not businesses,
- 1:41:34business owners, which means over 90% are not using AI
- 1:41:39properly. That is the real gap.
- 1:41:42What do you think guys? Yeah, Mark.
- 1:41:47I think the question here is also what is proper use?
- 1:41:51How do you use AI properly? Because in I see the point of
- 1:41:56the person asking the question because there's definitely not
- 1:42:00enough structure. That's at least my experience in
- 1:42:03most business of how you actually use the technology.
- 1:42:07In a lot of businesses you just get a direct order.
- 1:42:11You have to use AI and you don't.
- 1:42:13You don't actually get any instructions on how to use it.
- 1:42:17So, so I agree with the fact that we, we don't necessarily
- 1:42:21use it properly, but that is also part, like I said, part of
- 1:42:26getting to know the technology, getting to know the proper use
- 1:42:30of it and how do we get the best results.
- 1:42:38Again, OK, let's help as much as we can.
- 1:42:46Keynote Studio again says you don't have to be a big company
- 1:42:50to use AI. Yes, Mark.
- 1:42:55I agree. I agree, I think there are
- 1:42:57immense benefits for smaller companies to to use.
- 1:43:02AII use AI all the time in my business and it helps.
- 1:43:06Yeah, we do. In a lot of ways it helps.
- 1:43:08Yeah, exactly. But again, like the the the
- 1:43:13previous question, we do need to find some kind of structure and
- 1:43:18and that is something I'm working on as well to find the
- 1:43:20the most optimal way of using the systems.
- 1:43:23But great, great advantages to small companies, I think.
- 1:43:29Torsten. Yeah, I'm a one man business and
- 1:43:32I use it and and yeah, that's what I what I want to say, what
- 1:43:42I want to say. One man business is 1000
- 1:43:47businesses in any way. We all the we all use AI in
- 1:43:53different ways. 6 million in Denmark, Six million ways to use
- 1:43:58AI. Yes.
- 1:44:02Yeah. Well, we kind of exceed actually
- 1:44:05quite a lot the term timeline of life event today.
- 1:44:11But but before we close this session, I would like to kind of
- 1:44:15give to people as a reminder, Please remember that you have to
- 1:44:20calibrate guys, strategic integration, as we have talked
- 1:44:24throughout all this time and cognitive resilience as well.
- 1:44:28Because if you want to have a job in the next few five years,
- 1:44:32literally there are there are the this AI episode will need,
- 1:44:38you will need to attend to. It will help you a lot when you
- 1:44:42shift your thinking about AI, you can actually make 10 times
- 1:44:47the impact you normally you you would and therefore you will
- 1:44:52starts to treat it as a like a digital companion because you
- 1:44:56have more agency that you actually realize.
- 1:45:00So that's what actually I wanted to add.
- 1:45:03And if they're you, one of you, my guest, want to add something
- 1:45:07else before we close this live session, they're welcome to do
- 1:45:11so, no? I think we covered it quite well
- 1:45:19and I mean the dialogue can quite likely continue both on
- 1:45:23the YouTube channel and the social media channels.
- 1:45:27Exactly, remember guys that you can reach out, you have the link
- 1:45:33in the description on YouTube, on Facebook and on LinkedIn.
- 1:45:38You can definitely find every detail or informations in in
- 1:45:43those platforms on those platforms.
- 1:45:45So reach out to us. No, no problem.
- 1:45:48And yeah, we will reply to you in no time.
- 1:45:53Thank you for attending today. Thank you so very much to all my
- 1:45:56guests and to everyone as well. Thank you and see you on the
- 1:45:59next episode. AI and education.
- 1:46:02Please remember to sign out the register so you can actually
- 1:46:06receive all the updated dates and days and informations
- 1:46:12regarding the next episode #2 AI and Education.
- 1:46:16Thank you.