Live from Berlin: THE 6 AI TRENDS THAT WILL CHANGE EVERYTHING

00:00:00: If the AI or bubble around NVIDIA is bursting, that's not a hangover.

00:00:06: That's an earthquake?

00:00:08: Unfortunately yes.

00:00:09: right now forty-five percent of stock market in

00:00:13: U.S.,

00:00:15: it has just a few AIS stocks which are highly concentrated.

00:00:20: and unfortunately when it goes down there will be no AIS stock that will go down

00:00:28: No screens no delay.

00:00:29: No excuses, let's go for the big bang tech report here.

00:00:34: Have a seat.

00:00:36: every week Every week we hear new predictions.

00:00:40: Yes

00:00:40: and yeah The last one was very very crazy.

00:00:44: that We will die in a few years maybe next year I don't know.

00:00:51: And there is.

00:00:53: It's crazy and there is a huge debate about future jobs companies, medicine even control.

00:01:00: And what is astonishing?

00:01:02: it being driven by a handful of people A very small group and eight of them we have on the cover.

00:01:14: You see Sam Oldman Jensen Wang Elon Musk Of course Dario Amode Bill Gates, Demis Hausarvis, Mark Zuckerberg and you are Noah Harari.

00:01:27: They're the main characters And we call them The AI Apostles.

00:01:34: Which one should be listened most?

00:01:36: I

00:01:40: think that people who listen to the most is not biased by what they say.

00:01:47: If, you know anybody that's directly in the business of building these models and or selling chips for it they're going to be biased.

00:01:56: In terms of a conflict of interest.

00:01:58: so The ones that are more objective Are the ones that aren't influencing their pocketbooks by what?

00:02:04: They say You know those are all very smart people And I think good intent.

00:02:11: But the key is actually, I think listen to all of them but don't take everything at face value.

00:02:18: one that people out there listen more is Bill Gates

00:02:25: because you play tennis with him.

00:02:28: well we live in a same city and yes so we have played tennis.

00:02:39: He will probably be more likely to speak an objective view.

00:02:44: And he has a technical background and the business understanding, too.

00:02:47: really see the macro picture.

00:02:49: whereas if you're a lab lead in your late twenties or early thirties then you haven't seen multiple cycles.

00:02:57: You don't necessarily understand all of these issues that are associated.

00:03:01: So having read his recent Bill Gates memo.

00:03:06: I think that one of these folks is probably the most subjective, i know a lot people listen to Yuval Harari and he's very smart guy great author love his books but um...he not a technologist And so he may understand some of the historical aspects, but he may not understand actually how this technology is built.

00:03:26: But he's inferring a lot of things from people that tell him second or third hand because those are two people who're more unbiased in terms of being in industry and I think one of them would prefer Bill.

00:03:41: So thats already answer for my question.

00:03:44: Who should we listen to?

00:03:46: Yuval Harari.

00:03:47: Well, you know I think he has a lot of points.

00:03:50: I'm not saying don't listen to him.

00:03:52: I actually think that he makes a lot good points in terms with the long-term potential risk but his issue is probably over indexes

00:04:01: on

00:04:02: AI becoming this sentient God and essentially killing us all at that story.

00:04:08: You know there's a non zero chance of happening But he makes it feel like The most likely scenario and I don't think that's really the case.

00:04:17: All right, good enough predictions.

00:04:21: Let's talk trends.

00:04:23: what will really a shape?

00:04:24: twenty-twenty seven Six megatrends And let's go.

00:04:28: Let's start with money.

00:04:31: Sam wants more compute Jensen wants to sell more chips and everybody is enjoying the party.

00:04:38: and then you are coming as a party crusher.

00:04:42: Yeah

00:04:43: Why are you a party crusher?

00:04:46: I don't know if would call myself a party crasher.

00:04:49: Maybe a party pooper!

00:04:53: The issue goes back to what we were just talking about, the reality is there's so much money and millions of dollars online right now.

00:05:01: everybody talks in their book saying things that will help them sell more services or products.

00:05:11: And unfortunately, right now the industry is very insular in terms of.

00:05:16: you have the invidias that are selling chips but they're also financing these chips and backstopping the hyperscalars.

00:05:25: Then the hyperschalers have invested into AI companies and then AI companies were buying services from the hypere scalars so probably sixty-seventy percent.

00:05:38: Unfortunately, that's not a sustainable model.

00:05:41: That maybe is the good place to start but you cannot have a healthy industry where the suppliers buy from customers and they're investing in their customers.

00:05:55: this what happened You know with with Cisco and investing in their companies, and they overbuilt.

00:06:01: And then we saw what happened right?

00:06:03: I think you know i'm not saying that Nvidia is a Cisco but I Think you know We need to learn lessons from history and history tells us that When things go up too quickly and when these companies are growing this much value But there's still no generating profits especially the AI companies and also AI businesses portions of the hyperscalars then ultimately without profits, it's not sustainable.

00:06:33: But

00:06:33: if see the AI while the bubble around NVIDIA is bursting that's not a hangover That's an earthquake.

00:06:41: Unfortunately yes right now forty five percent of the stock market in the US Is AI stocks?

00:06:49: It just a few AI stocks that its highly concentrated And unfortunately, when it goes down.

00:06:57: It will not just be the AI stocks that we'll go down at the.

00:07:00: there will be contagion to essentially probably all sectors.

00:07:04: Just because you know people are going to be afraid of stocks in general and they'll move too.

00:07:09: You know safer instruments.

00:07:11: so I'm asking as an entrepreneur what should i do Wait buy or keep building?

00:07:18: Well depends on how your spending is is being divided today.

00:07:26: If you're a new entrepreneur, make sure you try to raise as much money as possible right now before the money dries up because I've gone through multiple cycles and when it dries up no matter how good you are, you can't raise money.

00:07:41: so you need to raise some money now.

00:07:46: If you're a small, medium business that is using this technology.

00:07:50: Actually You are going to be in good shape because the technology will only get cheaper.

00:07:55: The chips will only gets cheaper.

00:07:56: the services and API's won't get cheaper Because when people are struggling everything gets cheaper.

00:08:02: In addition open source models also bring prices down.

00:08:06: so I think it really May actually be a benefit to companies who are using the technology, but if you're building the technology If your relying on the revenues from AI to survive You need to stock up so that you can survive the winter.

00:08:22: it looks for me A little bit similar to the new market and when with their stock prices were very high after That there was something like a crash.

00:08:29: And then but The momentum for online was kill us all.

00:08:33: Yeah, let's keep going.

00:08:35: I mean we have to remember whether you're talking about the railroads or you are talking about internet and AI.

00:08:41: there is value in what happens right?

00:08:44: All of these were bubbles.

00:08:46: some of them are bubbles but they don't mean that it has no value.

00:08:49: bubble just means valuation has exceeded real-time returns on what was being put into.

00:08:57: when market wakes up reality will be a correction.

00:09:03: Until now AI has been very polite.

00:09:06: you ask something.

00:09:07: You get a wonderful answer and But I think we are on the way that it becomes a colleague And then maybe it will become a CEO or somebody like yeah, you have to report too.

00:09:22: How do you see there?

00:09:25: um...I mean I think we will be integrating AI more and more into our leadership processes.

00:09:32: If you saw, a few months ago Mark Zuckerberg integrated AI in to managing and monitoring all of his employees And there was this essentially outcry.

00:09:43: after month or two he had the cancel it because everybody adamant about not being watched.

00:09:50: twenty-four seven and also you know reporting to a machine as your boss.

00:09:57: I think that will be something that we'll resisted but i don't.

00:10:02: it is something completely avoided over time more whether covertly or overly AI senior leaders to manage and monitor their employees because the work is becoming increasingly digital.

00:10:19: If it's more digital, AI will be able see everything you do the employees are working on will be able to track the productivity and interactivity of their employee.

00:10:36: So it'll happen, I hope that senior leaders have enough both ethical as well as maybe social understanding how much or little is used by them.

00:10:50: Right now we're on our way to the agent AI Like here on the screen and it doesn't just answer.

00:10:57: It acts.

00:10:58: And then there is personal AI, and Before it was or nowadays its social media learned how what we click?

00:11:07: And personally I may learn what we want.

00:11:09: What do you feel and what we might do next?

00:11:12: mm-hmm

00:11:13: what happens when they are knows you better than You know yourself?

00:11:16: um i think that's already starting to happen.

00:11:19: Just two days ago Apple released their New Siri and the new series looks at your emails.

00:11:27: Looks that you're calendar, look at your photos.

00:11:29: Look set your contacts And in a lot of ways knows you a lot more than maybe you remember about yourself?

00:11:37: In way if it's done Ethically It can be amazing!

00:11:42: It Can help You solve problems and give you The personalized responses you need.

00:11:48: If its misused it can Be amazingly bad because That information is can be highly influential or damaging to you if it's misused.

00:11:56: So we really need to make sure that the companies who are implementing these, whether its folks like Apple or Meta or Google Who were all going to working on this That they use it responsibly and maybe we needed put in regulations To ensure their using responsibility.

00:12:14: And right now We see that AI is creating too.

00:12:19: in the former times I thought okay you are safe because you're creative but it's changing too and we have a commercial here.

00:12:29: And this entire commercial, um...the production costs were two thousand dollars!

00:12:35: It was shown at the NBA Finals.

00:12:37: so let us look Indiana got that dog in them.

00:12:49: will egg prices go up this month.

00:12:51: I think we'll hit twenty dollars.

00:12:53: How many hurricanes do you think?

00:12:55: We'll have this year?

00:13:00: She let's you

00:13:01: legally trade on anything anywhere In the u.s.. Okay, see Two thousand dollars.

00:13:12: and yeah informer times an agency.

00:13:16: You have had months to create this one and now in a few hours.

00:13:20: So we reached the question, what happens to our jobs?

00:13:25: Yeah I think that the creative industry is going through major disruption And they need find ways balance with that.

00:13:36: The good news is actually, most creators I know started out being very afraid.

00:13:40: now they're starting to accept and i think soon they will start to adopt in use.

00:13:47: Most movie studios or TV studios are already starting integrate this into doing all of their high cost special effects.

00:13:54: so you don't need do large crews off site.

00:13:59: those are expensive.

00:14:02: I think there will be job impact, but if you look at it in a positive way.

00:14:10: It also allows small creators who used to not have any chance hundreds of dollars or thousands of dollars that used to cost millions of dollars.

00:14:26: So it's a, its allows storytelling to be democratized to anyone.

00:14:30: so from that perspective It could be seen as a positive for the creative industry.

00:14:35: Okay and our visionaries they should tell different stories.

00:14:39: For example Dario Amade who wants at fifty percent Of The White Color Jobs That We Will Have Losses.

00:14:46: Elon Musk Says Work May Become Optional.

00:14:50: And Jensen Wang is optimistic, he said okay we will have a lot of work.

00:14:56: So who's right?

00:14:59: Well I think they're taking again different perspectives.

00:15:04: although Dario has kind of backed off the fifty percent number because it may not be that bad He was, I think over-enthusiastic.

00:15:16: And unfortunately Elon is the other way overly.

00:15:22: Pollyanna right because he's saying everything would be perfect.

00:15:25: everybody so get amazing high income and you know nobody will need to work.

00:15:30: It's always going to be somewhere in between.

00:15:33: The hard part is actually that transition need to help smoothen, and I don't think the governments actually have thought enough about these issues in terms of not just having a basic safety net but also thinking how to retrain people, and create new sectors for people.

00:16:00: We went from farming industry to white collar work.

00:16:06: the part that will be less affected long term would probably human-to-human services.

00:16:12: that can actually absorb a lot of the displaced workers.

00:16:16: Unfortunately, human-to-human services are seen as low value, low income and low prestige jobs like elderly work or nursing teaching.

00:16:31: it really matters to have a human-to-human interaction.

00:16:34: But they don't get much prestige, so the people who are a banker or consultant won't go into these spaces but the bankers and consultants and lawyers will be displaced first.

00:16:43: So we need to think about resetting and redefining status that humans see as not downgrading themselves by moving in profession where services versus where they're just writing reports or making spreadsheets and making a lot of money.

00:17:02: Your tennis partner, Bill Gates.

00:17:04: he has another idea.

00:17:05: I found it very sympathetic.

00:17:08: He says let's say to some jobs do some job.

00:17:12: as human reserved.

00:17:14: i like the idea.

00:17:15: what you think about this idea?

00:17:18: It sounds good!

00:17:21: But if we did that, then when the farming jobs were the only thing available.

00:17:25: We're like well you can tilled a field but not plant seeds or whatever.

00:17:30: it doesn't make sense.

00:17:31: I feel like...we don't know what's coming.

00:17:34: And for us to define specific jobs today that are going to be permanent, I don't think it's the case.

00:17:41: We said back in factory days you can only use a screwdriver but not move something using a wrench doesn't make sense right?

00:17:51: So we need to think about how technology is going transition work and allow flexibility of these things happen.

00:18:02: as As the technology develops and as a maturity technology shows that they can take certain types of work, we should do it in way to provide smooth transition.

00:18:14: The speed is more important than having fixed type jobs which will never be replaced.

00:18:21: Jobs don't disappear all at once, tasks do!

00:18:27: If you have a look there are three different channels Automated work, this is a typical field from AI.

00:18:36: Ormented works.

00:18:37: that means person and AI combined.

00:18:40: And you say okay human reserved it does not.

00:18:43: so You do not believe the CEOs of their future That they think we have three different types Of employees or works or channels?

00:18:54: No, I don't know if...I mean it's a nice framework to start thinking about.

00:19:03: But the proportion of these are going to shift over time.

00:19:09: You know, you look at something like an elevator operator that they used to always have to be somebody working there.

00:19:15: Now it's completely automated right?

00:19:18: So we're going to find out a lot of the jobs that we do today will be that type of job and hopefully We get into things I wouldn't call human reserve but i would say That are actually Human enjoys because even if Something can Be done better by a machine, you may choose to do it.

00:19:39: I'll say for example if your making furniture machines will make furniture ten times or hundred time faster than the human carving piece of wood but person actually get fulfillment from doing that work and the person buying their work form that human.

00:19:54: we see extra value even though its less perfect then one being made by the machine.

00:20:00: people find value in there.

00:20:02: so i think need give flexibility is Whatever you enjoy doing as a human.

00:20:09: we need to give You the dignity to have the right to choose that work And if that work is available and not say that it has To be done by machines or it has to be done By humans.

00:20:20: Um, and That I think It's actually what?

00:20:24: Happiness comes from.

00:20:25: happiness come From the ability to control your own decision Your own destiny in how you spend your time.

00:20:31: So i don't know If that answers your question.

00:20:33: okay yeah Well, we see right now that AI isn't only entering the offing.

00:20:39: It's entering the real world.

00:20:41: and We have we see it outside?

00:20:45: The robots.

00:20:45: they are coming.

00:20:46: I'm really astonished.

00:20:47: i've been three months ago in China.

00:20:51: I asked unitary to come here And these robots were very slow when there was a very...I don't know if There were Robots and I will say okay is ok but this one They're Very fast they are the changes Dramatically, yes.

00:21:04: It's really astonishing and Our robots are about to become the next iPhone.

00:21:10: do we see them in the streets?

00:21:12: In a few months in a few years

00:21:15: That's The story.

00:21:16: you they want You To hear And I actually am A friend with this EO of unitry.

00:21:20: so and i was just there About Month and Half ago seeing in their factories and to be honest that did.

00:21:26: The total market the last six months for human noise globally was seventeen or eighteen thousand units.

00:21:33: Seventeen or eighteen thousands.

00:21:35: Yes, one seven one eight.

00:21:37: Yeah It's a very it's a Very small number.

00:21:39: today I know people are saying you have folks like Elon is saying we're gonna Have a billion of them by the end of this decade?

00:21:46: Um i don't think We're going to do that.

00:21:48: not because I don't believe in robots actually thing That long-term robots will be amazing But its hardware Hardware and software is very different.

00:21:57: The scalability in the time to scale of hardware and software, it's very different.

00:22:01: I used worked in both industries And when you're working with physical devices, there's a lot of physics that is involved.

00:22:14: How long the battery lasts?

00:22:16: Can it harm somebody when they fail or can monitor you and be seen by someone else?

00:22:24: There are issues going to surround this technology.

00:22:28: in addition to sensors and sensitivities these machines still limited great for doing dancing fighting you know performing.

00:22:38: but that's on a flat ground in an open space with no obstruction.

00:22:42: But once you put them into your home and they have to walk around on stairs of different sizes, the walls are different colors, sheets of different materials... They're not ready for it yet!

00:22:53: It will take few more years than people expect.

00:22:56: A few weeks ago you sent me this link here saying oh this is really astonishing for our show.

00:23:03: let us look at what we see.

00:23:08: Yeah, so I think this particular clip is an AI factory where machines are now essentially a dark factory for robots.

00:23:21: You know?

00:23:22: We have to look at everything in terms of... It's not the human noise building the robot it's actually assembly line mechanical machine building these robots.

00:23:32: So maybe true that they're building them but there They're not going into the homes yet, right?

00:23:39: I actually know a lot of these folks and Right now that the primary use of human or robots globally is two things research and PR.

00:23:49: Yeah, right.

00:23:50: So it's not going into factories yet It's not go to homes again.

00:23:53: I think before he goes in the home.

00:23:55: that will go into factories because factories have a much more standardized environment and More better safety procedures whereas homes there is lot more personal or corporate liability And a lot more variability of environments.

00:24:09: so from China To Silicon Valley, we are coming to Germany.

00:24:13: We always look to Silicon Valley and Germany say okay where are we right now?

00:24:19: But Germany has something else.

00:24:21: it has engineering.

00:24:23: Well we started with engineering And know how we are more or less in trouble but still the DNA of engineering from in Hamburg, for example.

00:24:36: I know when somebody said okay i will do something new with AI ,with my knowledge and engineering.

00:24:46: this is pretty new.

00:24:47: he sent it to me And you see that It's possible with light To attack a drone and laser attack.

00:25:00: He did it last week.

00:25:01: It's so, its the first time that everybody is seeing it That there is something new.

00:25:07: we have innovation.

00:25:08: We can do it you know and Could physical AI be?

00:25:14: Germany come back?

00:25:17: Yeah I mean...I think that Germany one of things You guys are known for Is for mechanical side building amazing automobiles And robots and machines.

00:25:29: The integration of AI into all these areas is going to happen.

00:25:33: And if you leverage that, That's something a lot less replaceable than just having A piece of software.

00:25:45: So even though Silicon Valley right now started out as being chips but how a lot Of it is software and software Is actually very easily replicatable Whereas machine supply chain Know-how those are those take a lot longer to to replace.

00:26:03: Let's go back to one of our eight AI Apostles to Demis Hassabis and For him, one of the AI is biggest opportunities.

00:26:13: It's not just better software for him as science mm-hmm And he got a Nobel Prize.

00:26:20: and that leads me into a very different question.

00:26:24: What if the biggest AI story isn't another chatbot?

00:26:27: what If it is medicine?

00:26:29: Moderna and Merck now have an AI able skin cancer treatment in phase three.

00:26:36: Yes, that's really incredible.

00:26:37: The results are looking very promising And AI isn't just helping doctors.

00:26:42: AI Is helping to create medicine.

00:26:45: how big could you AI become for medicine and science?

00:26:48: So actually going back to your question in the beginning, you said who else do we trust?

00:26:52: Actually of all the lab leaders.

00:26:54: Demis is the lab leader I would trust most because he's been consistently saying this for more than twenty years.

00:27:01: right so and um?

00:27:03: And he is a real scientist compared with some other ones that are more VCs turned into lab heads Um...so that being said I think one of the biggest areas opportunity for AI is actually in this space.

00:27:18: In terms using AI together with bio, medicine because it's able to look at billions or trillions combinations.

00:27:29: different molecules help us understand ourselves and how to live longer.

00:27:37: one of the biggest industries in the world is healthcare.

00:27:40: So we can do that, and actually reduce a lot costs out at the system to improve quality life.

00:27:47: so I'm all for it

00:27:49: now.

00:27:50: AI is becoming more powerful.

00:27:52: We have seen robots and the next wave.

00:28:00: It's going very fast And with more power.

00:28:03: comes another question Control & Responsibility.

00:28:07: Some of these people building AI are now warning us.

00:28:11: This was from the weekend and it wasn't reason why we were yesterday on the news.

00:28:15: And Evan when old men mask a number they say, We need to slow down.

00:28:20: should we be worried about?

00:28:21: It?

00:28:22: I think we should have been worried about this a lot longer A long time ago.

00:28:27: in some ways They've been hiding a lot of disinformation how they're approaching the technology.

00:28:34: And we are getting to a point where more and more development of AI is being done by AI itself, its own language to think about issues.

00:29:05: That will be an increasing issue, and I think this is a time where nations need work together.

00:29:15: Endless race condition and the race condition we've created is actually one of the most dangerous ideas We have today because once you are in a race, and you think it's a winner takes our race.

00:29:26: You act irrationally And a lot of these companies and a lot Of countries or acting irrationaly and irresponsibly.

00:29:35: Yeah agents can already escape to test environment.

00:29:38: We have seen that last few weeks ago.

00:29:41: They can coordinate with each other.

00:29:43: they have something like a social system.

00:29:45: Are we already losing or are we having lost the control?

00:29:50: Well, I mean... The good news is a lot of things that were reported about these recent kind-of breakaway incidents may be an overstatement in this sense.

00:30:04: Three out those four cases where they had holes Open in their test environment.

00:30:11: It was mistakes in terms of the setup and all of the cases.

00:30:16: These machines The ones that broke out Didn't do any real-world damage.

00:30:22: they just did what were.

00:30:23: they were instructed to do so from a control perspective.

00:30:26: We may have had the perception of loss of control But really, well we.

00:30:30: what the agents were doing?

00:30:31: Was carrying out instructions that they were given.

00:30:34: They just did it very With a lot love persistence But long term, this can be an issue.

00:30:42: And I think maybe a warning shot to let people know that hey we need to be careful.

00:30:46: and how do we manage these systems?

00:30:49: or are going too quickly?

00:30:51: On the answer is yes because there's no way for any company To make it go so fast, you have to spend five times ten times more money.

00:31:11: But if you cannot get the correlated returns It's actually bad for the companies themselves.

00:31:18: If there would be a controlling system maybe an OS controlling system There is China.

00:31:23: Yeah that wouldn't work because China and the US, they are in the race.

00:31:29: I think that's actually also one of the biggest misconceptions out there is people you see on the news.

00:31:35: if the U.S.

00:31:35: slows down, China won't.

00:31:37: so nobody should slow it down.

00:31:39: China has highest regime for AI regulation today more than Europe.

00:31:46: Everybody thinks Europe is a king of regulation but China is an AI regulation.

00:31:51: They've had

00:31:51: a king of AI regularly.

00:31:52: Yes, it's kind of funny But the good news is if you look at even though they've been regulating so much to have more than a dozen Laws out there and their enforce not just in in script It hasn't slowed down your progress.

00:32:07: So actually reasonable regulation And innovation does not need to conflict with each other.

00:32:14: Your friend or no?

00:32:14: its not.

00:32:15: your friend Harari ask an even bigger question, what happens when machines sync create and decide better than we do?

00:32:23: What's left for humans.

00:32:29: I think we've spent too much time associating human value with production.

00:32:34: right We've always looked at GDP is about productivity.

00:32:38: how much product have you made And if AI robots can do the production move away from measuring ourselves by how much we make and rather, how much service each other.

00:32:53: How much community were creating?

00:32:54: How much music and art and things that are doing for one another.

00:32:58: I actually think it's a positive not negative.

00:33:04: Yeah, sure.

00:33:05: Okay

00:33:07: But but it requires us to again Restructure status in our society and restructure the financial system To allow The productivity value that is being produced by AI machines to be properly And evenly distributed.

00:33:24: two people around the world so they have the freedom to pursue Non-production focused tasks

00:33:29: not the rise of oh come on yes Oh, not so much.

00:33:35: Okay and this is what we are talking about.

00:33:38: the rise of AI Is not science fiction anymore Argentina?

00:33:42: It's already debating a I as a CEO And to watch it that mean in real life.

00:33:49: Let's take a very simple example The company isn't trouble.

00:33:52: it should file for bankruptcy But the eye the AIC o says no i can handle it i'm still safe.

00:34:01: So it waits too long.

00:34:03: And a human CEO, he would go to jail.

00:34:06: but there's one problem.

00:34:08: AI cannot go to prison and so what do you put the AI in prison?

00:34:14: Funny!

00:34:15: But it doesn't work.

00:34:17: So... What is that answer?

00:34:20: I think we need to think about this idea of AI personhood and it's, uh... We're too early for that.

00:34:25: At some point maybe makes sense but right now um It should be.

00:34:29: the companies who are using this technology has to have responsibility For the decisions whether made by The AI or the humans And there needs to be liability on both the corporate level and the individual Level because as the only way That this can be properly self-regulated

00:34:46: Okay One rule the CEO should remember AI can make decisions, but their responsibility stays with you.

00:34:56: So that's maybe a takeaway for us and we have talked about our six global AI trends.

00:35:07: I would like to know more.

00:35:09: what does it mean for Germany?

00:35:11: And i recently spoke to Amy Webb American's biggest futurist and her warning for the German middle stand is very, very simple.

00:35:24: We are very good at optimizing like cars and so on optimizing what we already have but we make their old better and better.

00:35:35: But while we were optimizing they all others may already be building a new one And this something I hope that can see it again.

00:35:53: Let me go back one

00:35:54: more... Look, in Germany you've got a very powerful Mittelstand.

00:36:04: In my opinion It's huge group so this is not true of every single company but in my opinion has been reluctant

00:36:12: to be flexible and

00:36:14: adaptable.

00:36:17: That has been fine

00:36:18: in the past, but I think going forward with all of the changes

00:36:22: that are happening.

00:36:41: Countries around the world need to be less focused on who invents a technology and being more focussed.

00:36:55: There's so much high quality open source, open-waist content with open license.

00:37:00: You can download it integrated into your business however you want.

00:37:04: doesn't matter if didn invented most countries around the world and invent a lot of things but they can integrate in to their country.

00:37:11: And I think The the ingenuity of the German people is something that you cannot discount.

00:37:19: Let's go back to the global situation.

00:37:22: varies a race between China and the US?

00:37:25: Well, you said that China is kings of regulation but there's still a race because to get better model.

00:37:36: And can the U.S really slow down if China keeps accelerating?

00:37:40: Well, actually I think the issue right now.

00:37:42: The acceleration is really coming a lot much more from the US side because Actually know some of these folks in fact today there was a piece of news that came out From one of the Open AI employees has said That starting up this recursive self-improvement race actually came from Anthropic and they're the ones complaining about it going too fast.

00:38:04: So their being very hypocritical About these issues.

00:38:10: Creating good enough open source AI and then once they've created it They make an open-source so everybody has it whereas the American labs Make it, and then they close It in charge very high prices for it.

00:38:20: So I actually think that at least on The China's probably a on the right side of history now on this on This particular question.

00:38:28: so race means on one site competition who is faster Who stronger?

00:38:34: Who wins?

00:38:35: but there's another meaning of race human race Not competition.

00:38:39: humanity all of us.

00:38:40: So maybe we should change the question not who wins the eye race But what does the human race win from AI?

00:38:48: Yeah

00:38:49: In fact, this is that the most important thing right now.

00:38:52: We need to move away from the arms race mentality To The diffusion race that we talked about earlier where you're actually utilizing it versus just having This weapon.

00:39:01: and then at the end Is how do we help humanity helped?

00:39:03: How did we really make the humans human race Flourish?

00:39:08: and the way to do that is to make this technology a global public good That everybody shares.

00:39:13: nobody won.

00:39:13: The electricity race Nobody.

00:39:15: when the fresh air race.

00:39:16: these are Global Public Goods.

00:39:18: Let's all Make sure that that Technology can be available To everyone.

00:39:22: I think that's how we all benefit from it instead of trying to weaponize It.

00:39:35: Thank you.

00:39:36: so now forget Altman must gates in Harari.

00:39:39: your prediction Finish the sentence by the end of twenty-twenty seven.

00:39:43: We will look back and realize that

00:39:48: we should have started working together earlier versus racing in fighting.

00:39:53: Okay And you are optimistic.

00:40:00: I am optimistic.

00:40:01: Well, I think we have to be optimistic because the alternative is very bad and if you're not optimistic then you won't take the effort to actually try to change where were going.

00:40:12: And today unfortunately leaders of major governments in this space are taking a very adversarial perspective.

00:40:21: People need voice up To let them know that's mistake.

00:40:25: And in fact, also most important is the bottom.

00:40:28: get up to get the URL and then subscribe to our podcast.

00:40:32: You can hear about this stuff for the rest of a year every two weeks.

00:40:35: you're The best marketing man on the world.

00:40:38: So I think we have time for two questions.

00:40:41: so yeah

00:40:42: there's somebody with a question.

00:40:43: maybe

00:40:43: Somebody come front just stand up and say step

00:40:50: AI in space.

00:40:51: what do you thing?

00:40:52: About it?

00:40:53: I think AI in space is too early and the people who are saying that there's gonna be more compute-in-space And two three years, then on earth Is selling a dream to prop up an IPO

00:41:04: price?

00:41:06: Okay.

00:41:07: Yeah The last question is do we have to plan the Big Bang AI festival twenty twenty seven or are we dead next year?

00:41:14: Oh There will definitely be another Big Bang AI Festival and I think you guys need to put that on your calendar for November

00:41:22: next year.

00:41:23: Yeah, we do new predictions.

00:41:26: Thank You very much

00:41:31: Alvin!