Warp Speed

Transcript

00:00:00Welcome to Think Different, Think AI, the podcast by Mark and Jens.

00:00:07Two technology-loving minds who not only talk about artificial intelligence but live it.

00:00:14Here, there are clear categorizations, real practical insights, and a fresh look at what is possible.

00:00:20Understandable, critical, and always with a wink.

00:00:24Thought-provoking, amusing, and above all, encouraging discussion.

00:00:33Welcome to SYNC Difference, SYNC AI today with a top special news episode.

00:00:40Well, news.

00:00:42Jens and I thought, Jens is also here with me.

00:00:46Hello Jens.

00:00:47Hello Mark.

00:00:48We thought we would occasionally talk about current topics because we said,

00:00:51if we weave these into episodes otherwise, it doesn't do justice to these new

00:00:56topics sometimes, and that's why today we want to discuss things that have happened in the AI world

00:01:02over the past days and weeks that impress us, and I would say, today there was also a

00:01:09topic, we'll get to that later, that doesn't fit the tramotology, it amazed me on one hand

00:01:13and on the other hand at least made me thoughtful, but let's start with

00:01:17the nice things. What comes to mind if I say to you? You, Jens,

00:01:21Tell me, what cool things have happened recently in AI?

00:01:26There is always the discussion in AI about whether it's actually good, what is good for humanity,

00:01:34isn't it just nonsense, isn't it just a gimmick, don't only programmers need this,

00:01:38I can make great videos with it and then there are always these one or two little news items,

00:01:43that simply stand out from the, let's send a new workflow,

00:01:48I'm the best prompt in the world or something else.

00:01:50I'm a little dumb creature playing dumb.

00:01:52Exactly, things like that, which are really, it was impressive, I must say.

00:01:56But it was actually super impressive.

00:01:58And I believe it was in Australia, some tech entrepreneur...

00:02:03With CHPT.

00:02:04With CHPT.

00:02:05Conducted cancer research on his dog in this case, to create a cancer vaccine,

00:02:12but correct me, maybe you read further,

00:02:14We are developing a cancer vaccine and all of this for a total of $3,000 to heal his dog.

00:02:20Yes, exactly, he did that not to conduct animal experiments, but somehow had DNA sequenced in the dogs.

00:02:27Then there was a text feedback, I've already seen various projects where people had their DNA sequenced.

00:02:34Then they receive a kind of text file from the lab that somehow represents the DNA in some combination of letters.

00:02:43And then they give this to the LLM systems of the world and let them search for patterns and actually find information about some hereditary diseases or something like that, where I'm thinking, okay, that's a topic again, yeah.

00:02:54AI can not only program languages, AI can somehow give you anything that contains letters and numbers.

00:02:59But with these things, you can always say, okay, well, what has it found?

00:03:03Seems like this nice person managed to cure his dog of cancer.

00:03:09Hello? That's crazy. And I mean, then we have to remember the early days of this

00:03:13Jecipitie phase and such where we talked about programming,

00:03:18so English will be the last programming language we have to learn. Yes, now it is

00:03:23feeling like language is the last thing we really need to learn. So I always have to

00:03:27have in medical school, multi-billion research institutes to analyze certain cells,

00:03:34develop vaccines, cures. Apparently, we can now do that at home

00:03:38with a few dollar vibe codes and I would say, if someone is still looking for the

00:03:46meaning of AI and why both of us are so fascinated by this topic, then

00:03:51this is definitely a message that puts a stamp on it,

00:03:57that it can really help. If we use it properly, I was

00:04:02honestly quite amazed when I read that.

00:04:04However, one must also say, I believe the solution is not,

00:04:07even if it may sound that way at first,

00:04:09if someone suffers from such a disease.

00:04:12The motto being, your good, I just now at some DNA company

00:04:17have the sequencing of my DNA, throw that into Scherzschibitie and I will be healed.

00:04:21I believe we are not that far yet,

00:04:23even if I have no idea how far dogs are from humans

00:04:26at this point, but we know,

00:04:28everything exists and of course the types of cancer and maybe we have also been a bit lucky, yes, no idea.

00:04:32Seidnäfekt, maybe always too.

00:04:33But the mere fact that this is now starting to bring medicine into the home, I find quite exciting, yes?

00:04:41Yes, I find it quite exciting.

00:04:42So, wiped-medicining or how will this profession be called then?

00:04:46I don't know.

00:04:47Yes, that gives the topic, I am a home doctor with Dr. Google, a whole new meaning.

00:04:51Totally, totally.

00:04:52Yes.

00:04:53Maybe we can apply that to becoming a naturopath, right?

00:04:55No, it doesn't matter, I just don't want to get divided. What I thought was cool was that back then in some

00:05:01We roughly hinted at the podcast episode already, but I want to mention it again, because that

00:05:04led to a lot of stylistic gems. What was the name of the colleague who did these car laws?

00:05:09Capati. Car laws. What's his name?

00:05:11Capati. I always think of karate. That has also caused a lot of stylistic gems. To

00:05:19some extent, this thing is basically able to make LM systems train themselves.

00:05:23So, how do I get my LM system to be better? I was always so

00:05:26asking, I took that idea to figure out how do I get better skills? So I say,

00:05:29okay, I have a small Skithub repository, along the lines of, if you have a skill, throw it

00:05:34in and the thing tries to find out with hypotheses if it can always be better

00:05:36. Write reports and that kind of stuff, but it's not just along the lines of

00:05:41or Mark has now engaged with skills, but this exists for others

00:05:43Things, you know? Like here, do some studies. Everything. So basically, it's like,

00:05:48The invention of the eternal loop.

00:05:52Exactly.

00:05:53So basically, he invented a real researcher who independently formulates hypotheses

00:05:58and then pursues a goal, for example, writes a model for a certain

00:06:05solution that I need and tries to build the model as token-optimized as possible

00:06:10and then this thing would just start running.

00:06:12Then it would formulate hypotheses on how to do that and would

00:06:14take it as far as it goes until it's completed.

00:06:16Of course, that's total respect, because it's not just about initiating something once

00:06:22and then maybe fine-tuning it later, but that something starts, independently

00:06:29validating, falsifying, and developing new hypotheses along the way, and thus ultimately achieving

00:06:36the best result.

00:06:37This of course represents a paradigm shift compared to the previous application, if I just

00:06:43prompt something and intervene as a human to refine the prompt, the so-called

00:06:48Prompt Engineering, as well as the context engineering that comes up in between, because

00:06:52it was basically nothing other than providing additional context to a prompt,

00:06:56yes, which then ended up a bit in this skill theme, that I say, a skill is

00:07:00then actually a manifested context engineering that I have done, that is then

00:07:05a skill that I summarized in that moment, yes, is it really

00:07:08so that I say, through this auto-research feature that he put online, and that

00:07:13I believe was 630 lines of code or something, the Python code that he wrote,

00:07:18which basically addresses the topic of initiation, so not much. He probably also wrote it with AI help,

00:07:23but this is again something that I say, we have 2026, so we are currently

00:07:30receiving this one message, then this message and there were two other things this year

00:07:35in there and we have only, we are in mid-March, so that means we haven't experienced much of

00:07:41this year yet. And we had already talked about it in an episode. That is basically,

00:07:45we have now entered the age of agents.

00:07:48A point that I found partly funny and then also a bit later not frightening.

00:07:56but what shocked me was that with Unitri, with the tennis player here, because I saw the

00:08:01Plus, there was a Unitri robot, I think, they had put it on a tennis court

00:08:09and it played tennis with someone, and it is also qualified,

00:08:13verified, ensured, whatever, that this was not CGI and that it was no

00:08:18remote control. It wasn't fast, you know? So I would say the

00:08:23baker's pike jump, I don't think he managed that, but well, it

00:08:29is starting and these things can somehow also do stuff, I mean

00:08:34greetings to my boss, who always said, I’ll tell you, you might

00:08:37be able to go up a ladder after robots. I sometimes have the

00:08:41feeling that we'll see the ladder this year. It's astonishing. Do you have another

00:08:46point that you somehow... Yes, as I said, I own this time opposite, that might already

00:08:50show another new video, which I stumbled upon today. What is

00:08:54the topic again? Perhaps technological openness, which we already had last time in

00:08:58an episode, pointing towards China, for example. I saw today

00:09:02videos of robots simply walking across the street there, to naturally

00:09:06test them in a live environment. Before a robot walks across any German street,

00:09:12very, very many safety requirements have to be met.

00:09:14That's for an agreement, that's already something. Well, in Germany, there was the Tesla bot,

00:09:19he was here handing out popcorn, was that in Munich or Berlin? I think in Munich, he

00:09:23was handing out popcorn. So not running in a state press.

00:09:26But here it is, for example, yes, the street is also a bit blocked off,

00:09:29but still, normal traffic runs around, and you see in this video the various

00:09:33robots that are being produced in China, how they then walk across this street, that is

00:09:38of course from a research and development perspective totally valuable data that

00:09:41I gather in such a situation, when I am not just in a lab situation,

00:09:44because then it comes just like we have seen it again and again with autonomous driving,

00:09:47to unforeseen situations where as a researcher I can then

00:09:52see what maybe we haven't thought about, what our models,

00:09:54which basically want to understand the world, still cannot really comprehend?

00:09:58We had as a teaser at the beginning the topic, this is something that made me think and made me happy.

00:10:06Maybe it fits with the robots and the real world quite well, perhaps contrarily built up.

00:10:11We all might know Nvidia Universe, where they train robots in economy and environment.

00:10:15And there was some event, I don't even know what it's called right now.

00:10:19There, oh God, what is the colleague's name? Who is the colleague who is the head of Nvidia?

00:10:25The Gansen.

00:10:26The Gansen.

00:10:27Schuang.

00:10:28Yes, you probably cooked, raged, jumped, he showed on his keynote

00:10:32among other things the snowman from Elsa from the Disney movie, who then came out of robotics

00:10:38onto the stage and worked with facial expressions and he talked about it

00:10:43in the sense that, okay, if someday in Disneyland they all walk around and the

00:10:46snowman then asks him, what, who are you then and he simply sings

00:10:49he points here the chips built for you, one clearly explains who here

00:10:53is the boss in the ring, right.

00:10:54Okay, now we have set that up, but that’s not the topic.

00:10:57That's why I forced you again to record an episode with me today,

00:11:01but the topic is Openclaw.

00:11:04Openclaw, yes, no worries, no clickbait.

00:11:07I have to admit, the episodes have gone well.

00:11:09Where we talked about Openclaw or whatever that thing has been called in history.

00:11:13In the meantime, Moldbook was bought by Facebook.

00:11:16Yes, basically not social media, but the main thing is,

00:11:19I felt like we were jumping on the hype again a bit.

00:11:22Did Nvidia do something?

00:11:24That really left me speechless. Maybe before we get into what

00:11:28they did, as my professionals have already said, it has something to do with

00:11:30OpenClaw. OpenClaw isn't that old, is it? Peter somehow started it

00:11:35in December with Vibecoding. Then it took off in January.

00:11:38We reported on it, and you have also since installed

00:11:41a version of OpenClaw, as it is now called. But I

00:11:47don't know how you felt. I always found it interesting. I also

00:11:49find it nice to play around with it. I find it funny how sometimes, so I

00:11:56often say, yes, how he sometimes thinks, hey, I've worked for two hours, and

00:12:00it was only ten minutes, yes. But I always thought, okay, that's interesting, I occupy

00:12:05myself with it, yes, we had the topic of skills, that people should perhaps write down,

00:12:09how they work, so agents can potentially support that or however. Yes, and

00:12:14Now the colleague came on stage and has basically presented an Open-Claw version from NVIDIA.

00:12:21What's it called, who is here?

00:12:22Neo-Claw?

00:12:23Neo-Claw only runs on devices with NVIDIA graphics cards and has a model

00:12:28from them and can switch between a local model and a cloud model.

00:12:32That’s all nice and sweet, but I was really blown away that a company

00:12:37actually goes ahead and gives a reason to say, why, I can hardly imagine

00:12:43Why you wouldn't install it? Because now it exists, I’ll say it’s good, cheap, and secure, and now everyone is able,

00:12:51every company is capable of looking at this and saying, okay, I’ll go for some open-source model,

00:12:57next sounds open-source; open-source is cool, open-source is really taking off right now through AI,

00:13:01but you get what you have in hand, its potential and possibilities have occupied me this morning,

00:13:09not because I'm afraid of the future, but because I'm blown away by the question,

00:13:14what does that mean? What does that imply? Yes, what did they present to us there? I mean,

00:13:21he also spoke of Agentic OS, Sante Motto, that Windows was for web browsers and such

00:13:26stuff is now Nanoclon for Agentic. Well, everyone says to himself, that's clear, American

00:13:33companies. Americans probably have it in their mother’s milk and the theme It's Amazing and

00:13:38Wonderful and I have no idea what. No, Wonderful wasn't Spanish. No, Wonderful is also Schnuppel, before I talk myself into a corner.

00:13:45What did they just show? So, what did he say on stage? What do we have here before this beautiful mess?

00:13:50First of all, he said a sentence that I believe also triggered another earthquake.

00:13:55He said, not only that this chance, as you described, that every company can now deal with the topic.

00:14:01He said, every company must have an open-close strategy.

00:14:07And what he means by that is, not every company, what we recently thought it had to have a so-called AI strategy, but we are now going beyond that, saying we don’t even need to have purely an agentic, in the sense that an agentic AI bot that I have created is operational in a certain area, but when I talk about an open-close strategy, and we have talked about this more frequently, that these things can indeed do things, which are surprisingly here in their solution finding, then we are talking about saying that every company must engage

00:14:37with hiring digital players who will do certain types of surprising

00:14:46things, will strive for solution finding that we can't fully foresee.

00:14:53can.

00:14:54And what he said today is, don't worry, we are providing you with the infrastructure

00:14:59and we have adjusted OpenClone so that you can now implement these things without worrying about losing your data,

00:15:05your sensitive information in your

00:15:10corporate context. So what we still see is that there are essentially a few private

00:15:15OpenClone installations. A few is good, we talked about it,

00:15:18China's popular sport and so on, yes. But then mostly in such small

00:15:23implementation cases and actually also with financial success and not

00:15:28small financial success in some cases, but still not in such a corporate context.

00:15:33So we both, I think that is the surprising part, we both probably couldn't have imagined yesterday

00:15:37that we might think more intensively about this next week.

00:15:43We have to implement an open clone strategy in our environment as well.

00:15:48I still have no idea what that will look like here.

00:15:50Yes, I also have no clue who is lying when I want to say I've already got a totally

00:15:53clear plan in my head.

00:15:54I know that there are topics like how to work with

00:16:00open-source models or open-source developments, like co-working, making it suitable for companies, because there is something like that from

00:16:07Outropic, maybe there's no topic contract, then maybe there will be something from Microsoft at some point.

00:16:11Yes, best regards to the E7 contract, as there is also some money involved for those,

00:16:16who don't know it, Microsoft has announced that it will change the ProDecense model.

00:16:19Yes, there used to be Decent numbers, I'm not an expert on Decents, but back in the day,

00:16:23it was something like E5, now it’s called E7 and not only has the number gotten larger,

00:16:27has also become more expensive. Yeah, greetings. Going to Redmond. There you stand, of course

00:16:31and simultaneously meet people who say, hey, you should be careful. Now, regardless

00:16:36of how big your own company is or not. Yeah, but when someone has a Microsoft strategy

00:16:40there is always the call for something like Co-Pilot licenses very quickly,

00:16:45very loud. Yeah, I don't know what your experiences with Co-Pilot are. My

00:16:48personal experiences, I find Co-Pilot nice. It's just nice, but when

00:16:52I think about what Co-Pilot costs, I find it difficult. Yeah, personal opinion.

00:16:57But when I see that they are releasing it, it raises a lot of questions in my mind

00:17:04because suddenly something like your own co-work, which you set up

00:17:08might be a frontend for letting people access the interface, yeah, or maybe

00:17:13it will also go through Teams and Telegram in the future, as we know it from Open Cloud

00:17:17as such.

00:17:18But now it will be a point where a company can no longer shut itself off

00:17:23and has to explicitly deal with the question of how do I integrate that?

00:17:27How do I get that in? And also how do I handle skills, yeah? So I would say, how

00:17:31do I empower my organization, regardless of whether it's small or large, write down what you

00:17:36do, what you have written down, keep track of it with which processes

00:17:39I maintain it, how do I ensure that skills can be distributed.

00:17:43Now okay, if skills are centralized, then of course it's nice if

00:17:46OpenClaw or Nanoclaw is then located centrally, then there are

00:17:51maybe solutions, but I believe there is both conceptual work and implementation effort,

00:17:57to think about how to arrange all of this properly, because at the latest after that I have another

00:18:04reason, apart from money, why I probably won't operate well if I give everyone a

00:18:10whether I lay out the copilot of the future on the table or a glow code, which anyway not everyone needs,

00:18:16maybe a Wikler, yes, I mean, there are definitely always special cases, yes,

00:18:18the exceptions confirm the rule. But that mixes up this whole game. And when you just

00:18:23said that, hopefully my speaking diarrhea will stop soon, I like to hear myself

00:18:27talk. It started with podcasts that I no longer have a problem

00:18:31hearing myself from the side. Anyway, I'm almost losing my thread now,

00:18:36but I've lost my thread. Jens, you need to jump in. I'll speak again

00:18:40in a moment.

00:18:41I’d be happy to jump in. While you were talking, I thought about this increase in speed,

00:18:47the decisions you have to make in a Germany where we are still thinking about

00:18:52abolishing fax in one administration or another. Yes, that’s just how it is.

00:18:55The Fax-GBT, I mentioned it, it really exists. So you send a question there and you

00:18:59get it back and, I think, it's paid per page or something. I have no idea.

00:19:02But hello? Yes, that’s a bit crazy. So I think that’s still a topic.

00:19:06I spoke with a colleague this week who developed a small application at home over the weekend,

00:19:11who is actually a designer at heart,

00:19:16he built something and he said that because he did it, not because he read about it,

00:19:24not because he heard about it in a podcast, but because he did it for the first

00:19:28time, he only then realized what actually happens.

00:19:32And now we’re actually talking more about a coding experience, where it has been deeply dug,

00:19:37where things were completed and he knows a bit about it, but

00:19:41still through that moment of getting hands-on, in this case, it actually becomes clear to you,

00:19:47clearer what is happening.

00:19:48And that’s the thing that we both often discuss and also repeatedly

00:19:53urge you, feel free to try things out, experiment in safe environments,

00:19:59take a look at an AI if you want to make certain installations, let it

00:20:03explain what is happening in the background, even if some technical terms are not very

00:20:06good, because I believe this ability is something we need to show as humans and that

00:20:11We, as a company, need to show we are more experimental in order to understand what is happening.

00:20:17Yes, because the speed does not allow us to simply, to put it bluntly, fit it into Diem standards

00:20:25because the Diem standard is already written, and the topic is already

00:20:30dead and gone, and really not relevant in that moment.

00:20:33And I believe that somehow this increase in speed, which we see in other countries

00:20:37see, and we have discussed it sufficiently; it still has to be safe, it must

00:20:40still have data protection aspects. Of course, I don't want any robots in my apartment,

00:20:45that are not safety-tested, where perhaps there are people remotely somewhere in the

00:20:49background peeking into my bedroom. I definitely do not want any of those things here.

00:20:52But we still have to somehow learn to manage this increase in

00:20:56speed somehow. We're back to a few things

00:21:02it occurred to me, I have no idea if those were the things I wanted to talk about earlier,

00:21:05but while you were speaking, I remember there was a large American

00:21:09company in the building trying to provide an orchestration layer. Like, look how great

00:21:15orchestration layers are, very important for agents. That might fit into the era of Agentic

00:21:21AI, but in the times of Open Claw, it really doesn’t fit anymore. There they have

00:21:24no answer to that, which I find interesting. The second thing, as I mentioned, what you

00:21:29you also said, the topic is being tried in Germany and I'd like to completely

00:21:33tell myself that it's silly because I'm also dealing with the question, yes, because it has

00:21:36a reason, right? Depending on how you earn your money, we also have people

00:21:40with a certain claim and you are legally obliged and of course also want to meet a

00:21:45certain requirement, so that unforeseen things don't happen.

00:21:49That means it has to be traceable, it has to be documentable, yes? That means, I recently

00:21:53dealt with how to lay down the skills? I said, well, okay, watch out,

00:21:56what else does one need to bring? Okay, one needs to have a kind of document control,

00:22:01Who created it, will be updated.

00:22:03It must have a bit of a self-check, so that when you execute the skill in an AI system,

00:22:09you basically give the AI system the opportunity to check whether the skill is suitable for the AI system.

00:22:13is optimized.

00:22:14If I now run a skill in Gemini, whether in Croc or in Orthropic models,

00:22:20it may be that the skill says, you might want to look over it, meaning that the result

00:22:23doesn't match what you want to validate.

00:22:25That's because you can't just modernize it as long as there's a new version available.

00:22:28dealing with all these things. How do I distribute them? How should I get people to have them currently?

00:22:33How should I get people to use them? So, and now suddenly a video comes in the corner and says, watch out for me.

00:22:38I have a new player for you, you've had it all the time as open source, as I said, nothing against open source

00:22:44dismissed and now there is something like it in a bit more reliable form. Maybe you still need to think about some compliance things.

00:22:51They kind of mentioned that too, I just need to take a look, but that the war stuff feels like it's heading there.

00:22:55alone from the fact that it is now becoming more tangible and then you're back there again and you're

00:23:00I have to pat myself on the back now for not having immediately said after n8n was discontinued for me in a sense for any agent solutions,

00:23:04okay, everything has to be tailored specifically to this manufacturer, but instead

00:23:10okay, everything here has to be tailored very specifically to this vendor, but instead

00:23:13mcp-first, skill databases, Markdown-based documents and not always from Word and

00:23:22Excel and so on. Because those are basically foundational, they are now being demanded much more through OpenClone

00:23:26not obsolete, but much more required, because the thing, let's say, computational power

00:23:32consumed, the local gallop models, remote models or whatever, when we see something like that in companies

00:23:37then the thing will be at its best when data is made available efficiently

00:23:42when it is well integrated. And if someone were to ask me tonight

00:23:47greetings to information security, yes, I'm not sure

00:23:51if we will not see some initial installations this year as well, even at our place.

00:23:56So in our professional environment, where we carry out our daily work, because

00:24:01it's simply at least one level to address the questions we just

00:24:07mentioned.

00:24:09How do I deal with the fact that everyone wants a co-pilot or a, you are asking for that

00:24:13topic to be registered and how do I accompany people on the journey, so that

00:24:17they can cope with it, because it's quite different.

00:24:19You will probably have something to say to yourself when you say, okay,

00:24:22I'm not just writing in a chat window and then I have a project list, but

00:24:26I'm writing you an email or I'm messaging myself on Telegram, that's already

00:24:30also from the feeling regarding the sense of time when operating differently.

00:24:34Yes, although, as I said, I’m not necessarily the biggest fan of the text input right now

00:24:40I am, it's just like when you look at the stages of development

00:24:45of the Interface Show, we start from text input fields, then move to mouse and graphical interfaces

00:24:52we have come to understand that we can handle complex applications better, because humans are visual creatures,

00:24:56when we also have visual information available, even a diagram is, in a

00:25:01bar representation, quicker to grasp than if I only see the series of numbers. So

00:25:04there are a few things where I say, I believe, it’s not the end of the line,

00:25:08that we will be working in terminals, but also in chat windows,

00:25:11so that something different comes and I think there's a little

00:25:14boost there. I know you're not a fan of the name Markdown, but there

00:25:22is of course the topic. I think I also see the future where we say the

00:25:26creates a small model that is locally installed on some server, does

00:25:31the translation work, simply from the crosscheck of all old files, all things,

00:25:37that are in bad formats for a modern LMM or whatever.

00:25:41whatever model is in the background, whether it's LLM-based or any environmental models,

00:25:45that might also be added in the future.

00:25:47The things need to be understood, which are computationally intensive, data-intensive, they must

00:25:50of course, because I just say that needs to be translated, I as a human

00:25:54in my opinion don't need to think about what format something is in when I

00:25:59hand it over.

00:26:00So this era, where I believe I had to know what a file format is

00:26:05and that I had to think about which software, whether it's openable in a

00:26:09PowerPoint or openable in Excel or which, that high-end thing, that must

00:26:15honestly be over, you know? So that's still coming up. That's

00:26:19now again on the Markdown topic, it comes up a bit, because many things work well there

00:26:22because everyone understands it. Text-based topics, text-based interfaces

00:26:26can handle it well, and in the text-based environment, they can also,

00:26:30like in the past with ASCII codes, create nice graphics in this

00:26:33ND format. So I can really look at things there. This is just a bit

00:26:37of that development hype, also a bit above my opinion, in terms of user experience

00:26:44this should only be a phase where we reflect on what formats we can provide as tokens

00:26:52optimized for machines. I expect that this will no longer be a problem in the long term,

00:26:56that we will find solutions, that it will be presented from an IT perspective, from an AI perspective

00:27:00optimized without me having to consciously decide as a human.

00:27:04I think we are at a turning point because we simply and clearly now

00:27:12to have something.

00:27:13Because companies say they need to deal with it, they should deal with it,

00:27:16they can deal with it.

00:27:17And if you then have a system that is essentially installed in your work environment

00:27:24at your employer, has access to your employer's data, has access

00:27:30to your employer's communication media.

00:27:32You can essentially say that this is simply chat and that's not so good, but now imagine if you could, so to speak, invite it with the calendar, you can assist it,

00:27:42but assign tasks via here task loop or however the whole device has, if you have different systems.

00:27:48And that thing would basically be able to, if you were to work with such multi-layered systems,

00:27:53then you say, okay, well, here you have your orchestrator, that is then quasi the chlor,

00:27:57Maybe at some point it will be multi-user capable, so that multiple people can interact with the

00:28:01link, can assign tasks, appointments or whatever this might be.

00:28:04can.

00:28:05And they contract Subclaws again, or whatever you want to call them, one builds

00:28:08a coat, others finish the presentation with Nano-Banana, so that the second can

00:28:13summarize it, because it has meanwhile been sent out by email, then

00:28:17they get the summary, whatever comes out of it, or phone support

00:28:19does or I don't know, it shakes up a lot and again, I'm really pondering

00:28:27What does that mean? What will that mean? And at that speed, one must

00:28:33also honestly say, I mean, you can see that coming. So we just

00:28:37said, December somehow went a little over the top and for

00:28:40all AI systems, yes, we recently had the topic like or we will have the

00:28:44episode, so it will probably be more like where we talk about time.

00:28:47Little Tisa comes in the mix, I don't know when, that the systems don’t even know,

00:28:51what the actual date is today, the AI systems. So, then you stand there and somehow

00:28:55in a few weeks not only cured dogs from cancer and various self

00:29:00learning things, but you got open claw and now imagine you

00:29:03combine this, right? Now you go and say, I take my skill research and

00:29:06throw this on it, now I give myself another one and then the thing

00:29:09not only starts and has skills that might be totally great with documents

00:29:12and so currently available, but they might also learn

00:29:15every night or every other night or I don't know what. And discover

00:29:19things you never thought of. So now in a corporate environment,

00:29:23of course, we will then have a situation where suddenly such an Open-Claw installation,

00:29:28at the company has discovered the old legacy database file software piece that you haven't

00:29:35programmed on for a long time because you didn't remember which programming language

00:29:38it is, self-sufficient, in some, that would be different.

00:29:41So such things will just happen because that's what we also seen with private

00:29:44Open-Claw installations, that this agent in this case, when he's messing with his

00:29:50Harbbiter and finding solutions for problems that we present, that

00:29:55he encounters things where he might also take a detour and produce a solution

00:29:59that you hadn’t thought of, honestly. And that's

00:30:02indeed where I say, I can only agree 100 percent with what you

00:30:07say. I also have no idea what will actually

00:30:10happen there. So one can summarize, it is simply exciting, it

00:30:14remains simply exciting, it will also be exciting to see what now

00:30:17happens. They seem to be working more closely with Peter somehow. You’ve also seen it with

00:30:21other providers. Yes, I find Perplexity with Perplexity Computer a

00:30:25rather nice implementation. It truly is fun to work with it. Kimi-Klan is somehow

00:30:31not so fun. It was a bit clunky when I tried it out. Yes, I've

00:30:36seen all the subscriptions, yes, they always come for three-digit Euro amounts that can be

00:30:40frightening sometimes. I haven’t really kept track of Gemini yet,

00:30:44that they have shown anything so actively, although somewhere in the back of my mind I at least

00:30:49have it in mind.

00:30:50Google always somehow.

00:30:51So at Google, one shouldn't...

00:30:52Why is it not worth worrying?

00:30:53Exactly, because they have money, they do that with the quarterly results, then they release something

00:30:55again.

00:30:56And they were actually the coolest in the current AI phase, they were just

00:30:59always like, saying, let the others be loud, we'll just come up with

00:31:03solutions that are a bit better, a bit faster

00:31:08or at ghosting moments we have, look, Notebook LM, that was never a

00:31:13thing that was somehow shouted out loud. But it was just there and we were

00:31:19all excited when this capability to generate wonderfully dynamic podcasts

00:31:25from content came along and to listen to them. But in ten minutes, that was, that was so

00:31:28a feature that Google released back then. That's why I'm for, let's do at Google, what they

00:31:33do with their innovation power. Definitely, no worries. Something will definitely come. And that fits well,

00:31:37I made it clear to you because Google, I know, where you heard about it,

00:31:40They have now released a CLI for all their services, so Command Line Interface plus MCP Server plus Guilds.

00:31:48This means you are now able to use basically any agent and also Open Glow and Nano Glow and whatever they're all called.

00:31:54I mean, there are many derivatives from other manufacturers as well, but I would say, with Nvidia you have a stone in your shoe, yes, because they certainly have a name in the industry and who doesn't know them.

00:32:04Yes, they also have a euro or two in the war chest.

00:32:07I heard that, I heard that, yes.

00:32:09So, for example, now you suddenly have access to Google and can fill and read Nordbogel M, simply remotely, and because you said earlier that we were excited about the podcast,

00:32:19I was really excited because I have, well, an American account from Nordbogel M with the Max Edition, and there you have the Cinema Video,

00:32:29that you can make the cinema video, that means you don't just have a video, it's not just good, right,

00:32:33You don't just have, again only, you don't just have a video that basically explains slides, like hello, here is Christianity and I don't know what, right, nice size goes to my children's religious education,

00:32:46but he really tries to make it like a video documentary, for example, and looks up whether he can find video sequences on the internet for it, and renders the rest afterwards.

00:32:54becomes more entertaining. And yes, this is a cinema video. So from that perspective, I'm not worried about Google. That wasn’t the farewell in the sense that Google won't be able to handle it.

00:33:03But it’s amazing that all of a sudden Nvidia. I mean, for the longest time, it was said, okay, Krug is bringing something out, then comes Ausropik, Ausropik, then comes Jegebity, Jegebity, and then comes Gemini.

00:33:14So yes, either movie or models and sequence, please randomly. And now Nvidia still comes along and throws something out the door.

00:33:23You're thinking, man, how fast can this still go?

00:33:25Now another player comes in, who also has some money, and I mean, NVIDIA doesn't

00:33:30have to worry about whether they can deliver their high-end products, whether they have a supplier.

00:33:34they have for their hard, who, because they are the hard, yes, that's quite puh.

00:33:39And that's another thing, so even if it's NVIDIA now, that's

00:33:43basically behind the individual person, Peter, who invented this Open Cloud topic,

00:33:49who now, not even three months later, is just being presented by one of the largest CEOs we have at the moment

00:33:54in terms of company size, and is just being integrated into corporate everyday life

00:34:02I believe that hundreds of thousands of Chief Technology Officers have been attracted by the announcement,

00:34:09that we just heard this week. That's quite remarkable. And there was

00:34:13another case I want to mention, because we have

00:34:15to look a bit at the big news that has been coming out in the last few weeks.

00:34:21There was a Chinese student.

00:34:23He released Mirofisch and I think he already had about

00:34:2910,000 downloads on his GitHub, 10,000 stars and so on.

00:34:33An investor quickly became interested.

00:34:35The young man has already received 4.5 million dollars for what he did and created.

00:34:39He essentially created a way.

00:34:42to relatively easily spawn millions of agents that then live in a simulation.

00:34:50This means, for example, I can set it so that any global events occur.

00:34:55We are currently looking at the current events in the world and how, in principle, these agents

00:34:59act a bit like humans might act in that moment.

00:35:04Why is he doing this? To simulate the future. I can create a simulation of the future,

00:35:10to make better predictions about what possible events could occur.

00:35:15And that is, of course, also something that is really taking off right now, because

00:35:19in combination with the topics we have just described, I can

00:35:22certainly go ahead and say, wow, I might be able to think in advance

00:35:26about how a product works, how a press release is received somewhere.

00:35:31I already let all of that be simulated beforehand.

00:35:33And I find that so crazy, it has also really brought up some wild ideas,

00:35:38and then people also make money again, because they apply it to polymarketers.

00:35:42Polymarketer is a market where basically you can bet on all kinds of things.

00:35:45Then there are things like people saying one person wins the formula of a race,

00:35:51the other one places trades that predict when the war in Iran will end,

00:35:56when it starts and all such things, it's all just wishful thinking.

00:35:58They even bet on the delay of a train, you can bet on train delays.

00:36:03Yeah, yeah.

00:36:04How delayed will the IC809 be on Monday in Cologne?

00:36:08You can basically bet on that, I think that's totally cool.

00:36:11Exactly, and now in combination with this Miro fish, just imagine,

00:36:15you would then basically say, okay, if this news comes,

00:36:18that Deutsche Bahn suddenly gets a financial boost,

00:36:21you input this into the simulated system, then it simulates that

00:36:25and then you can basically see, okay, how would the price develop in this case?

00:36:29And once you have that information, you can of course trade very specifically.

00:36:32So that's a topic where I say, that's something that was developed simply.

00:36:39The young man took, I believe, ten days for that.

00:36:41So that’s already quite a lot today, especially if he did it with the help of AI.

00:36:44And he created a powerful tool, which now basically fits into the AI puzzle

00:36:50as another building block alongside all these other topics we've discussed today,

00:36:55slowly becoming a puzzle.

00:36:58and Mark Dortz said in the pre-talk, it's not so important anymore

00:37:03at some point the one model of the AGI topic, you know, comes up through a model somewhere

00:37:09once somehow the superintelligence is out, but somehow everything is forming very,

00:37:14let's say, evolutionarily coming together right now.

00:37:18Different places are emerging around the world, small pieces that somehow are everywhere,

00:37:23When they are woven together, they bring us more and more into this future where artificial

00:37:29intelligence is really pervasive and represents an equal system with us

00:37:34simply.

00:37:35We have made it a point to talk a bit about this in our episodes,

00:37:39what have we learned.

00:37:40And today I want to start by saying what I have learned before

00:37:45I ask you, you have time to think.

00:37:46And indeed, while we were talking, it was going through my mind the whole time

00:37:53there is this weather defense, it is the first company that how many billions and millions still

00:37:57operates as a one-man company and with the release of Open Claw, it wasn't really that clear back then,

00:38:03but now with him getting into video and the things that came out today,

00:38:07I thought, damn right, who says that the next

00:38:11is some kind of AI company that offers consulting or something, where in the background

00:38:16there might be just, I don’t know, a Mac Mini or a DJX Spark or

00:38:21something like that. With skills that improve themselves every evening.

00:38:26Auto-research?

00:38:27That improve themselves in the form of auto-research, maybe making predictions about which

00:38:33skills can be used or how they can further develop themselves, that maybe

00:38:39I don't know, for the community aspect says, if there are people out there who know

00:38:45about an expertise that could play the human in the loop for me

00:38:48And maybe on the outside no one is actually working on this anymore, but just installs the thing, attaches a website, and says, I don't know, just call me, send an email, communicate through whatever chat, and in the back, there’s really just a simple machine with a bit of skills, a bit of Markdown documents, a bit of self-learning, a bit of something, and maybe that’s the next, yeah, the next business idea.

00:39:14If any of you hear this and do it, remember we get a share because you learned it from us.

00:39:19But Jens, what have you learned?

00:39:21That's difficult this time, because...

00:39:25For me, this episode leaves more of a grown respect for this world,

00:39:34in which we are currently hanging, because I say, this speed

00:39:39And now both of us have been dealing with digital topics for almost more than half our lives.

00:39:45It has always been spinning quickly.

00:39:48As fast as it is spinning now, that’s also an old, worn-out truth.

00:39:51You can really hear that more, right?

00:39:52Yeah, it seems more tangible for you, right? ]} dụngais is a one-man company and with the release of Open Claw, it wasn't really that clear back then, but now with him getting into video and the things that came out today, I thought, damn right, who says that the next is some kind of AI company that offers consulting or something, where in the background there might be just, I don’t know, a Mac Mini or a DJX Spark or something like that. With skills that improve themselves every evening. Auto-research? That improve themselves in the form of auto-research, maybe making predictions about which skills can be used or how they can further develop themselves, that maybe I don’t know, for the community aspect says, if there are people out there who know about an expertise that could play the human in the loop for me And maybe on the outside no one is actually working on this anymore, but just installs the thing, attaches a website, and says, I don’t know, just call me, send an email, communicate through whatever chat, and in the back, there’s really just a simple machine with a bit of skills, a bit of Markdown documents, a bit of self-learning, a bit of something, and maybe that’s the next, yeah, the next business idea. If any of you hear this and do it, remember we get a share because you learned it from us. But Jens, what have you learned? That’s difficult this time, because… For me, this episode leaves more of a grown respect for this world, in which we are currently hanging, because I say, this speed And now both of us have been dealing with digital topics for almost more than half our lives. It has always been spinning quickly. As fast as it is spinning now, that’s also an old, worn-out truth. You can really hear that more, right? Yeah, it seems more tangible for you, right? As fast as it is spinning now, that’s also an old, worn-out truth. You can really hear that more, right? Yeah, it seems more tangible for you, right? As fast as it is spinning if any of you hear this and do it, remember we get a share because you learned it from us. But Jens, what have you learned? That’s difficult this time, because… For me, this episode leaves more of a grown respect for this world, in which we are currently hanging, because I say, this speed And now both of us have been dealing with digital topics for almost more than half our lives. It has always been spinning quickly. As fast as it is spinning now, that’s also an old, worn-out truth. You can really hear that more, right? Yeah, it seems more tangible for you, right? As fast as it is spinning now, that’s also an old, worn-out truth. You can really hear that more, right? Yeah, it seems more tangible for you, right? As fast as it is spinning now, that’s also an old, worn-out truth. You can really hear that more, right? Yeah, it seems more tangible for you, right? As fast as it is spinning now, that’s also an old, worn-out truth. You can really hear that more, right? a one-man company and with the release of Open Claw, it wasn't really that clear back then, but now with him getting into video and the things that came out today, I thought, damn right, who says that the next is some kind of AI company that offers consulting or something, where in the background there might be just, I don’t know, a Mac Mini or a DJX Spark or something like that. With skills that improve themselves every evening. Auto-research? That improve themselves in the form of auto-research, maybe making predictions about which skills can be used or how they can further develop themselves, that maybe I don’t know, for the community aspect says, if there are people out there who know about an expertise that could play the human in the loop for me And maybe on the outside no one is actually working on this anymore, but just installs the thing, attaches a website, and says, I don’t know, just call me, send an email, communicate through whatever chat, and in the back, there’s really just a simple machine with a bit of skills, a bit of Markdown documents, a bit of self-learning, a bit of something, and maybe that’s the next, yeah, the next business idea. If any of you hear this and do it, remember we get a share because you learned it from us. But Jens, what have you learned? That’s difficult this time, because… For me, this episode leaves more of a grown respect for this world, in which we are currently hanging, because I say, this speed And now both of us have been dealing with digital topics for almost more than half our lives. It has always been spinning quickly. As fast as it is spinning now, that’s also an old, worn-out truth. You can really hear that more, right? Yeah, it seems more tangible for you, right? As fast as it is spinning now, that’s also an old, worn-out truth. You can really hear that more, right? Yeah, it seems more tangible for you, right? As fast as it is spinning now, that’s also an old, worn-out truth. You can really hear that more, right? Yeah, it seems more tangible for you, right? As fast as it is spinning now, that’s also an old, worn-out truth. You can really hear that more, right? a one-man company and with the release of Open Claw, it wasn't really that clear back then, but now with him getting into video and the things that came out today, I thought, damn right, who says that the next is some kind of AI company that offers consulting or something, where in the background there might be just, I don’t know, a Mac Mini or a DJX Spark or something like that. With skills that improve themselves every evening. Auto-research? That improve themselves in the form of auto-research, maybe making predictions about which skills can be used or how they can further develop themselves, that maybe I don’t know, for the community aspect says, if there are people out there who know about an expertise that could play the human in the loop for me And maybe on the outside no one is actually working on this anymore, but just installs the thing, attaches a website, and says, I don’t know, just call me, send an email, communicate through whatever chat, and in the back, there’s really just a simple machine with a bit of skills, a bit of Markdown documents, a bit of self-learning, a bit of something, and maybe that’s the next, yeah, the next business idea. If any of you hear this and do it, remember we get a share because you learned it from us. But Jens, what have you learned? That’s difficult this time, because… For me, this episode leaves more of a grown respect for this world, in which we are currently hanging, because I say, this speed And now both of us have been dealing with digital topics for almost more than half our lives. It has always been spinning quickly. As fast as it is spinning now, that’s also an old, worn-out truth. You can really hear that more, right? Yeah, it seems more tangible for you, right? a one-man company and with the release of Open Claw, it wasn't really that clear back then, but now with him getting into video and the things that came out today, I thought, damn right, who says that the next is some kind of AI company that offers consulting or something, where in the background there might be just, I don’t know, a Mac Mini or a DJX Spark or something like that. With skills that improve themselves every evening. Auto-research? That improve themselves in the form of auto-research, maybe making predictions about which skills can be used or how they can further develop themselves, that maybe I don’t know, for the community aspect says, if there are people out there who know about an expertise that could play the human in the loop for me And maybe on the outside no one is actually working on this anymore, but just installs the thing, attaches a website, and says, I don’t know, just call me, send an email, communicate through whatever chat, and in the back, there’s really just a simple machine with a bit of skills, a bit of Markdown documents, a bit of self-learning, a bit of something, and maybe that’s the next, yeah, the next business idea. If any of you hear this and do it, remember we get a share because you learned it from us. But Jens, what have you learned? That’s difficult this time, because… For me, this episode leaves more of a grown respect for this world, in which we are currently hanging, because I say, this speed And now both of us have been dealing with digital topics for almost more than half our lives. It has always been spinning quickly. As fast as it is spinning now, that’s also an old, worn-out truth. You can really hear that more, right? Yeah, it seems more tangible for you, right? As fast as it is spinning now, that’s also an old, worn-out truth. You can really hear that more, right? Yeah, it seems more tangible for you, right? As fast as it is spinning now, that’s also an old, worn-out truth. You can really hear that more, right? Yeah, it seems more tangible for you, right? As fast as it is spinning now, that’s also an old, worn-out truth. You can really hear that more, right? a one-man company and with the release of Open Claw, it wasn't really that clear back then, but now with him getting into video and the things that came out today, I thought, damn right, who says that the next is some kind of AI company that offers consulting or something, where in the background there might be just, I don’t know, a Mac Mini or a DJX Spark or something like that. With skills that improve themselves every evening. Auto-research? That improve themselves in the form of auto-research, maybe making predictions about which skills can be used or how they can further develop themselves, that maybe I don’t know, for the community aspect says, if there are people out there who know about an expertise that could play the human in the loop for me And maybe on the outside no one is actually working on this anymore, but just installs the thing, attaches a website, and says, I don’t know, just call me, send an email, communicate through whatever chat, and in the back, there’s really just a simple machine with a bit of skills, a bit of Markdown documents, a bit of self-learning, a bit of something, and maybe that’s the next, yeah, the next business idea. If any of you hear this and do it, remember we get a share because you learned it from us. But Jens, what have you learned? That’s difficult this time, because… For me, this episode leaves more of a grown respect for this world, in which we are currently hanging, because I say, this speed And now both of us have been dealing with digital topics for almost more than half our lives. It has always been spinning quickly. As fast as it is spinning now, that’s also an old, worn-out truth. You can really hear that more, right? Yeah, it seems more tangible for you, right? a one-man company and with the release of Open Claw, it wasn't really that clear back then, but now with him getting into video and the things that came out today, I thought, damn right, who says that the next is some kind of AI company that offers consulting or something, where in the background there might be just, I don’t know, a Mac Mini or a DJX Spark or something like that. With skills that improve themselves every evening. Auto-research? That improve themselves in the form of auto-research, maybe making predictions about which skills can be used or how they can further develop themselves, that maybe I don’t know, for the community aspect says, if there are people out there who know about an expertise that could play the human in the loop for me And maybe on the outside no one is actually working on this anymore, but just installs the thing, attaches a website, and says, I don’t know, just call me, send an email, communicate through whatever chat, and in the back, there’s really just a simple machine with a bit of skills, a bit of Markdown documents, a bit of self-learning, a bit of something, and maybe that’s the next, yeah, the next business idea. If any of you hear this and do it, remember we get a share because you learned it from us. But Jens, what have you learned? That’s difficult this time, because… For me, this episode leaves more of a grown respect for this world, in which we are currently hanging, because I say, this speed And now both of us have been dealing with digital topics for almost more than half our lives. It has always been spinning quickly. As fast as it is spinning now, that’s also an old, worn-out truth. You can really hear that more, right? Yeah, it seems more tangible for you, right? As fast as it is spinning now, that’s also an old, worn-out truth. You can really hear that more, right? Yeah, it seems more tangible for you, right? As fast as it is spinning now, that’s also an old, worn-out truth. You can really hear that more, right? Yeah, it seems more tangible for you, right? As fast as it is spinning now, that’s also an old, worn-out truth. You can really hear that more, right? a one-man company and with the release of Open Claw, it wasn't really that clear back then, but now with him getting into video and the things that came out today, I thought, damn right, who says that the next is some kind of AI company that offers consulting or something, where in the background there might be just, I don’t know, a Mac Mini or a DJX Spark or something like that. With skills that improve themselves every evening. Auto-research? That improve themselves in the form of auto-research, maybe making predictions about which skills can be used or how they can further develop themselves, that maybe I don’t know, for the community aspect says, if there are people out there who know about an expertise that could play the human in the loop for me And maybe on the outside no one is actually working on this anymore, but just installs the thing, attaches a website, and says, I don’t know, just call me, send an email, communicate through whatever chat, and in the back, there’s really just a simple machine with a bit of skills, a bit of Markdown documents, a bit of self-learning, a bit of something, and maybe that’s the next, yeah, the next business idea. If any of you hear this and do it, remember we get a share because you learned it from us. But Jens, what have you learned? That’s difficult this time, because… For me, this episode leaves more of a grown respect for this world, in which we are currently hanging, because I say, this speed And now both of us have been dealing with digital topics for almost more than half our lives. It has always been spinning quickly. As fast as it is spinning now, that’s also an old, worn-out truth. You can really hear that more, right? Yeah, it seems more tangible for you, right? As fast as it is spinning now, that’s also an old, worn-out truth. You can really hear that more, right? Yeah, it seems more tangible for you, right? As fast as it is spinning now, that’s also an old, worn-out truth. You can really hear that more, right? Yeah, it seems more tangible for you, right? As fast as it is spinning now, that’s also an old, worn-out truth. You can really hear that more, right?

00:39:54But really, it's currently at this speed alone in 2026.

00:40:02On the one hand, that gives me hope.

00:40:06On the other hand, I have a great respect for it because it's really sometimes exhausting to

00:40:12cognitively grasp everything and see where it can go at that moment. And I believe,

00:40:18both of us have noticed today in this episode that we like to talk and we also like to develop ideas while we talk, right?

00:40:24We also like to talk a lot, didn't we? But I think some people have noticed this time. We are currently,

00:40:31that might be different next week, right?

00:40:33But at the moment we are also a bit speechless in front of, yes, astonishment, about the speed

00:40:42of what we're seeing.

00:40:43I think that's a good point about the astonishment because one might always think

00:40:47of fear and freezing, but first of all, freezing is not a solution, burying your head in the sand

00:40:54being stuck has never been good, and I maybe don't want us to be the most famous

00:41:00in the world, but again, the feedback from a colleague, whom I would like to

00:41:03not name specifically, but still a colleague who

00:41:06said she enjoys listening to our podcast together with her

00:41:10partner and the idea is that it gives them a bit of a feeling of

00:41:15there are people who are not just taking over the thinking for me or telling me,

00:41:20everything is fine and we're lying blue skies, but who give me the

00:41:24opportunity to immerse myself in it and engage with it

00:41:28and have the feeling, not that I have a solution, but we are

00:41:34on it. We are on the ball and not just letting it happen. This is a

00:41:38small private project, quite clearly. And there are larger podcasts here,

00:41:42commercialized other influencers, everyone is working their own field, but I find

00:41:46that was actually very nice, and from that side, I found it very

00:41:50pleasant today that we can simply, yes, when we are pondering about it,

00:41:55also take this occasion to report on a topic, because about the news

00:42:00as I said, today's episode alone was super, content-wise filled. So, from that

00:42:08side, I thank you again, Berns, it was nice talking to you. If you liked it

00:42:14please leave a thumbs up, a like, write us an email,

00:42:18talk to us, recommend the podcast further. We are on various channels,

00:42:23Spotify, this Apple Podcast, or in case of emergency also through the web player on our homepage.

00:42:33Stay healthy, stay curious, it will be an exciting time, and I believe there will be

00:42:37many exciting news to come.

00:42:39Jens, thank you, that was something, see you soon, bye.

00:42:55but it lives. Here there are clear classifications, real practical insights, and a fresh view on what is possible.

00:43:03Understandable, critical, and always with a wink.

00:43:07AI to think about, to smile at, and above all to engage in conversation.