- 1 day ago
Germany excels at invention but often fails to scale it. The SPRIN-D agency aims to speed up the path from research to market-ready technologies.
Category
đŸ¤–
TechTranscript
00:07There are ground-breaking inventions, and then there's the wheel.
00:11The very first wheel was likely made not for transport, but for a potter five and a half thousand years
00:16ago,
00:17although the inventor's identity remains a mystery.
00:20But that breakthrough set in motion a worldwide surge in innovation.
00:24The minimal friction from rolling translated into more efficient transport, boosting mobility and trade.
00:30As well as mills and trains, wheels also powered clocks and industrial machines.
00:36While we don't need to reinvent the wheel, innovations are as vital as ever,
00:41and the German economy could do with some right now.
00:49Also on this edition of MADE,
00:52Timu, Sheehan and the like, new customs on low-cost orders from China,
00:56plus Claude Mythos, how an AI model poses potential cyber risks.
01:12Inventing. Germany has pioneered or shaped technologies like the washing machine,
01:17the automobile, the telephone, the magnetic levitation train,
01:21the electric streetcar, the X-ray, aspirin, the computer, television, the chip card and much more.
01:28Inventors are people who tinker, often born of necessity.
01:33But at the moment, Germany seems to be asleep.
01:36Why is that?
01:39We've somewhat neglected creating new industries in the last 50 years,
01:43even though a lot has happened technologically.
01:47We have certainly become a little too complacent and comfortable in Germany.
01:54The automotive industry, the National Railway Deutsche Bahn, SAP, Airbus, Bosch and ThyssenKrupp
02:02receive millions in taxpayer subsidies for development projects.
02:06Many people have the feeling that it's not worth taking risks, building something on their own.
02:13The German government wants to change that,
02:16and it founded the Sprint Agency for Breakthrough Innovations in Germany in 2019,
02:22modelled on the American DARPA.
02:24It's the first of its kind in Europe.
02:28We need to invest the money in the up-and-coming industries
02:31to bring transformative technologies to the world
02:33and to solve the big questions of our time.
02:38Magnotherm is a start-up from Darmstadt.
02:40They've developed a magnetic drinks cooler that's already being used in supermarkets.
02:44It replaces conventional cooling by means of rotating magnets.
02:48The same technology will soon be used in data centers.
02:52Germany is actually an excellent location.
02:55In terms of know-how and the ability of the people, it's really very good.
03:00There are many things that we have to consider as a small company
03:03when funding projects like this.
03:06This start-up wants to improve cancer treatment with AI.
03:09The immune system cannot distinguish between benign and malignant cells.
03:13So the team has developed a method that can detect tumor cells.
03:17This could see the end of side effects such as hair loss.
03:22No start-up actually brings a drug to market on its own
03:25because there are very, very big hurdles due to the safety that's required
03:29because of the clinical trials, which often cost a billion or more.
03:33And we don't really have any opportunities of our own to develop something good
03:37because we don't have the capital.
03:41These founders left a large corporation.
03:43They've recognized a gap in the market.
03:46Many chemical residues from the chemical and pharmaceutical industries
03:50remain unused or have to be disposed of at high cost.
03:53Here, they're sold on.
03:57If I could wish for something, if you say we have a tax office,
04:01the biggest cost item for us is accounting.
04:05We would have loved to tap into European funds, for example.
04:08Many people would do that here.
04:11Everyone warns us that we have to hire a person for 9 to 18 months
04:14to even be able to tap into these funds.
04:19Start-ups struggle with bureaucracy.
04:21Permits that take two to three days or weeks in other countries
04:25can take over four years in Germany.
04:28Start-up rule mapping has set out to simplify bureaucracy with AI.
04:33Bureaucracy is basically a good thing.
04:35And as a society, we need rules for living together.
04:39And then above all, we need methods and tools to apply these rules digitally
04:42and as automatically as possible.
04:44And if that happens ten times faster,
04:47then we will no longer have this bureaucratic standstill.
04:51Too much bureaucracy, too long approval periods, too little capital.
04:56These are the problems that start-ups in Germany must contend with.
04:59Things are better in other countries.
05:04For me, the US is still the role model par excellence
05:07when it comes to innovation.
05:11Germany is slowly waking up, but still has some catching up to do.
05:20What do Jeff Bezos, Bill Gates, Steve Jobs and Elon Musk have in common?
05:27Before they got rich and famous, they each saw at least one big idea fail.
05:32From search machines to futuristic hovercraft and traffic data projects,
05:36new computers, rockets and cars, none of them took off.
05:41But those setbacks didn't stop their inventors from persevering and prevailing.
05:46But can the concept of failure as a catalyst for innovation work in Germany, too?
05:55The DESERTEC industrial initiative aimed to provide low-cost solar energy
06:00direct from the desert to Europe.
06:02A study by the German Aerospace Center deemed it feasible.
06:06But the entrepreneurs miscalculated in terms of both finances and demand.
06:11The project failed.
06:14Failure is inevitable, even when you consider that with the most successful venture capital
06:19companies, the clear majority of, say, 10 investments fails, and a totally natural process
06:25is taking place.
06:27These are innovations that involve breakthroughs, that involve risks, where you don't know if
06:33something's going to work or not.
06:36For years, air taxis were touted as the transportation mode of the future.
06:42The German Aerospace Center was researching the use of airways to reduce urban traffic congestion.
06:49This project also failed, and millions were wasted.
06:55You don't learn anything without failure.
06:58With building rockets and armaments, there are very experienced people around.
07:03Everyone benefits.
07:04It simply wouldn't have happened if they hadn't failed.
07:07Everyone was upset.
07:08But it's the right decision, because new things can come of it.
07:12If we hold on to things that are failing for too long, then this transformation won't happen.
07:19But how should companies and the state decide which technologies to focus on?
07:24When does investing make sense?
07:26Germany's Agency for Breakthrough Innovations, or SPRINT, was founded to answer that question
07:33and identify major technical opportunities at an early stage.
07:39You can do lots of good things with AI.
07:41It'll change our lives.
07:43We can have a before and after, a chat GPT moment that will go down in history.
07:48That's a breakthrough innovation.
07:51There is a catalogue of more than 90 parameters that determine which invention is worth investing in.
07:57What kind of people are inventors?
08:01We're looking for people who are geniuses, who are fundamentally curious,
08:05who don't have a work-life balance problem.
08:09They're often incredibly optimistic about the future.
08:11Everything can be solved.
08:14But they still have to be able to learn, to have that tenacity, that resilience.
08:21SPRINT has now introduced a law that lets it use venture capital without the usual restrictions
08:27on the allocation of public money.
08:31In this innovative kiln, materials such as ceramics, steel, glass or metal are not melted in a fire, but instead
08:39irradiated with LEDs.
08:42I was doing a PhD in something else and wanted to fire ceramics quickly, but there was no good way
08:48to do it.
08:49It needed blue or UV light.
08:51Five years ago, there were no semiconductors for it.
08:54It would have just stayed a fantasy.
08:58Instead of heating it to a thousand degrees, it stays cold.
09:02The energy goes right into the material.
09:04It melts in less than a minute.
09:07That was my first prototype.
09:10Initially, nobody believed that you could heat things up with blue LEDs, especially not over a thousand degrees.
09:17My first grant applications were rejected.
09:19And then, out of frustration, I threw my credit card on the table and just built this thing myself with
09:25stuff from the hardware store.
09:26And I spent my own savings on it.
09:32He has since filed seven patents.
09:34The first LED furnace should hit the market next year and go into industrial use.
09:42You can't succeed without failing.
09:45Eighty percent of my university projects were pointless.
09:48Or I messed them up or learned things in the process.
09:52And then I found something really interesting.
09:56And some of those were really big breakthroughs.
10:00Innovation drives economic growth.
10:02Failure is part of the process.
10:05Something newer and better replaces the old.
10:07This way, growth can continue and benefit humanity.
10:11You really can try out things several times.
10:14And that we shouldn't stigmatize failure.
10:17That we can take more risks.
10:19So get started with something new.
10:22And get back up again if you fail.
10:24Failure is worth it.
10:31Coming up with an invention is one thing.
10:33Bringing it to market, quite another.
10:38A German developed the technological foundations behind the Internet.
10:42But it became a global phenomenon via the US.
10:46The same goes for Apple, making the MP3 format mainstream and search machines that again involved German research being commercialized
10:54by Google.
10:55But does it always have to be that way?
11:00California's Silicon Valley in the US is home to universities and over 7,000 companies.
11:06It's the global hub for tech and Internet startups, known for rapid knowledge exchange, a high appetite for risk and
11:13abundant venture capital.
11:18I spent half a year studying at Stanford, observing how things work in Silicon Valley and learned they're not so
11:24different after all.
11:27That's why, along with BMW shareholder Susanne Klatten, he co-founded the startup center Unternehmer TUM in Munich in 2002.
11:35Here, students and young founders receive both financial support and entrepreneurial know-how.
11:43Germany has exceptionally strong research, but far too little focus is placed on turning these insights into real products.
11:57Here, young founders can develop and test their ideas using equipment they couldn't otherwise afford.
12:04Today, it's one of the largest startup centers in Europe.
12:09One of the many successful companies that started here is Isar Aerospace.
12:16Some argue that really large, successful companies can only be built in Silicon Valley.
12:21But it's also possible in Germany.
12:24Following the Munich model, similar startup factories are set to be established in Berlin, Hamburg, Dresden and seven other cities.
12:33Universities, founders, companies and investors will work together,
12:36and more than 300 startups are expected to emerge from these centers each year.
12:40We come up with many innovations, but instead of building industries around them, we keep selling them to the U
12:47.S.
12:49The German Federal Agency for Disruptive Innovation, Sprint, aims to help young founders gain a foothold in Germany.
12:59Per capita, Germans even invent twice as much as Americans.
13:03We're very good at inventing, but less good at turning ideas into market success.
13:09In Germany, startups typically need 10,000 to 500,000 euros in initial funding, and several million later on.
13:17Marvel Fusion has raised over 380 million euros in venture capital, backed by Sprint and Siemens Energy.
13:24Its goal, green electricity from nuclear fusion, fusing atomic nuclei.
13:29But the necessary laser systems and fuels are still under development.
13:33They're also exploring whether existing nuclear plants can be adapted.
13:41Germany clearly plays a leading role in Europe.
13:43If we get it right, the first fusion power plants could once again be made in Germany.
13:51Patents are especially important for startups to protect their technologies.
13:58Patents give their holders exclusive rights for up to 20 years.
14:02After that, the invention enters the public domain and can be used and further developed by anyone.
14:10Germany leads Europe in patent applications, with more than 62,000 filings per year.
14:18Patent protection makes companies more attractive to investors.
14:21As a rule, there's no funding without it.
14:25This startup has already filed more than 60 patents.
14:29Supported by the Leibniz Supercomputing Center and the Federal Ministry for Research, Technology and Space,
14:34it's developing a quantum computer.
14:38We've grown rapidly, reaching 120 employees in just three years.
14:43That's only possible with the right investors.
14:47We now have more than 20 paid collaborations with a wide range of partners,
14:51from large companies and SMEs to research institutes.
14:57Unlike conventional computers, quantum computers can explore multiple solutions at the same time.
15:04With their vastly greater computing power,
15:07they could be used to develop new materials for aerospace,
15:11as well as in pharmaceuticals, logistics, and encryption.
15:19The technology we rely upon comes from Germany,
15:22making it an excellent environment for us
15:24and giving us confidence we can be among the global leaders.
15:29It remains to be seen whether Germany's young startups
15:33can regain their place on the global market in the years ahead.
15:41Global markets for all manner of products are under pressure,
15:45not least from import tariffs.
15:48They're meant to protect domestic economies,
15:51but they also lead to inflation.
15:55Donald Trump's aggressive tariff policy has shaken the global economy,
16:01while the EU tries to slow the flow of cheap goods from China.
16:07Bargains are just a click away.
16:10Online shopping is booming,
16:12and platforms like Alibaba, Temu and Shein lure buyers with low prices.
16:17European airports are flooded with parcels from Asia,
16:20but from July 2026,
16:22the EU plans a new customs charge on low-value imports.
16:26Will it make a difference?
16:29Wool and colorful beads.
16:31Sabine, who lives in a small German town,
16:34orders her craft supplies online,
16:36mainly from Chinese sellers.
16:41When it comes to beads,
16:43it's so much cheaper on these platforms,
16:45incredibly cheap.
16:47And once you spend above a certain amount,
16:49there are no shipping costs,
16:50so you really just pay for the materials.
16:53It's so much cheaper that it's worth it.
16:56If I could buy these things locally, I would,
16:59but you can't find them in our small town.
17:05In many smaller towns,
17:07local shops have disappeared.
17:09But even in major European cities,
17:12retailers say fewer customers are coming into stores.
17:15Like Sabine,
17:16millions of consumers across Europe
17:18now order directly from Asia.
17:21Online retail in Europe has grown rapidly.
17:24Amazon leads with about 52 billion euros in annual revenue,
17:28while Otto and Zalando keep expanding.
17:31But with sales up nearly 300 percent in just a year,
17:34Chinese platform Temul is the fastest growing.
17:38Germany's retail association says 460,000 parcels from China
17:43arrive in Germany every day.
17:45And there's another problem.
17:48Our criticism is based on the fact
17:50that tests carried out by consumer organizations,
17:54including Stiftung Warentest,
17:55as well as various authorities,
17:58repeatedly show serious problems
17:59with many of these products.
18:01They include unsafe goods,
18:03missing electrical certifications,
18:05and in some cases,
18:06hazardous substances in toys
18:08or toxic chromium-6 in children's shoes.
18:11It's unacceptable.
18:13It's utter junk that should never be permitted
18:15to enter our market.
18:20European policymakers are responding.
18:22From July 2026,
18:25low-value imports from outside the EU
18:27will face a new customs charge.
18:30Previously, goods under 150 euros were exempt.
18:33The proposal adds a flat fee
18:35of about 3 euros per parcel.
18:38In 2022, about 1.4 billion
18:41of these shipments arrived in Europe.
18:43A year later, it was 2.3 billion.
18:46By 2024, it had doubled again
18:48to roughly 4.6 billion.
18:52Airports and logistics operators
18:54are overwhelmed with shipments.
18:56Liège Airport in Belgium,
18:58a major European hub for Chinese e-commerce,
19:01is uneasy about an EU-wide customs charge.
19:04France and Italy have already imposed similar duties,
19:08disrupting the air freight market.
19:11A similar tax in Italy shifted traffic
19:13from Milan to other airports.
19:17In France, it moved traffic away from Paris
19:20to other airports.
19:21That's very bad for jobs in those countries
19:23and at those airports.
19:26People who work in sorting, handling, and logistics
19:29can lose their jobs from one day to the next.
19:35The EU says the goal is a uniform system
19:38across all member states.
19:40Customs authorities are also expected
19:42to increase spot checks on imported goods.
19:45But since final EU rules are still being negotiated,
19:48German customs decline to comment.
19:50We're calling for stronger targeted customs inspections.
19:54France is already doing this
19:55and Germany should too.
19:57Our message is we have to pull the plug
19:59on anyone who systematically defrauds
20:01the state, consumers, and the economy.
20:06But the airport spokesperson understands
20:09why European customers are turning
20:11to cheap Chinese goods.
20:12He says the new tariffs are unlikely to change that.
20:19If you look at how the market is evolving,
20:22companies like Amazon, Zalando, Alibaba, and Temu
20:25are all trying to deliver faster and more efficiently.
20:28They're doing that because the market
20:30and customers demand it, not only in Europe,
20:32but around the world.
20:37Sabina isn't sure whether the new customs charge
20:40will stop her from ordering from Temu.
20:42And she understands why these platforms
20:45can be so addictive.
20:49Even if I'm being sensible
20:51and I know I've just ordered new craft supplies
20:53and don't really need anything else,
20:56you click and suddenly all these beautiful images pop up
20:59with lovely materials and beautiful colors.
21:04Your brain starts thinking,
21:06oh, I'll just have a look.
21:07And once you start scrolling,
21:09it's incredibly hard to stop and tell yourself
21:11that's enough.
21:13It's endless, almost addictive.
21:16It's incredibly hard to counteract that
21:18and tell yourself, no, that's it.
21:21No more shopping.
21:25Even with new customs charges,
21:28thousands of shoppers are likely to continue
21:30buying low-cost goods from China,
21:32just like Sabina.
21:39The global market for cyber security
21:41is also fiercely competitive
21:43and worth an estimated $258 billion.
21:47More sophisticated attacks
21:49are driving the rise of specialized cyber defense providers.
21:54It's ideal for new AI systems,
21:57but could the AI itself become a threat?
22:06A laptop hacked in minutes,
22:09or the IT systems of a power company,
22:12a stock exchange,
22:13or any other computing network.
22:16That's the fear surrounding a new AI model
22:19called Claude Mythos.
22:22This particular tool seems to be more powerful
22:25than anything else we've seen before
22:27and is able to identify vulnerabilities
22:30that others, including humans,
22:32have overlooked for decades.
22:34The developers say
22:35it's already found countless vulnerabilities.
22:38We found a number of vulnerabilities
22:40where, as a user with no permissions,
22:43I can elevate myself to the administrator.
22:46For the U.S. military,
22:48that creates both a threat
22:49and a potential advantage.
22:52So what makes Claude Mythos different?
22:55And how can these risks be contained?
23:00Claude Mythos was developed
23:01by U.S. company Anthropic.
23:05We haven't trained it specifically
23:07to be good at cyber.
23:08We trained it to be good at code,
23:10but as a side effect of being good at code,
23:13it's also good at cyber.
23:14The Pentagon pushed for unrestricted use
23:17of Anthropics AI,
23:18but the company rejected uses like surveillance.
23:21Governments worldwide are watching closely.
23:24We believe that AI models like Claude Mythos
23:27and others that may follow
23:29could significantly change
23:30the cybersecurity landscape.
23:35What worries experts
23:37is not just the model's speed.
23:39It's the model's ability
23:40to combine multiple weaknesses
23:42in a single attack path.
23:44It has the ability
23:45to chain together vulnerabilities.
23:47So what this means
23:48is you find two vulnerabilities,
23:50either of which doesn't really
23:51get you very much independently.
23:53But this model is able
23:54to create exploits
23:55out of three, four,
23:56sometimes five vulnerabilities
23:58that in sequence give you
23:59some kind of very sophisticated end outcome.
24:02That's why Anthropic
24:04has kept Mythos tightly restricted.
24:06The company has launched a project
24:08called Glasswing.
24:11It allows a select group
24:13of U.S. partners
24:14to use the model
24:15to identify vulnerabilities
24:17before criminals do.
24:20The fact that this is one company
24:24that has the power to decide now
24:26which of its selected partners
24:27has access to this
24:28really illustrates how powerful
24:31these tech companies are.
24:32And at the moment,
24:34it's really unclear
24:35who has access to this.
24:37Access outside the U.S.
24:39is limited.
24:40Countries like India
24:41worry this could undermine
24:42their digital sovereignty.
24:44Without access to advanced systems
24:46like Cloud Mythos,
24:47they could fall behind
24:48in cyber defense.
24:52Countries and companies
24:53need ongoing access
24:54to advanced AI tools
24:56to identify and secure vulnerabilities
24:58in their products and systems.
25:01That means reaching agreements
25:03with international manufacturers
25:05to ensure that success.
25:10For now, there's no clear roadmap.
25:13The U.S. government says
25:15it wants to work with Anthropik
25:16and other tech companies
25:17on new standards and safeguards.
25:22At a major AI summit
25:24in February 2026,
25:26India's Prime Minister Narendra Modi
25:29called for global cooperation
25:30on AI governance.
25:32And Anthropik's founder
25:34was among the speakers.
25:36But agreeing on global rules
25:38is difficult,
25:39especially when cybersecurity
25:40and intelligence are involved.
25:42Meanwhile,
25:43the race to control powerful AI
25:45is accelerating
25:46and governments
25:47are struggling to keep up.
25:51That's all from Made.
25:52We looked at how Germany
25:54might renew its innovative potential,
25:56the EU protecting itself
25:58from discount Chinese goods
25:59and the pros and cons
26:01of cyber defense.
26:02Stay safe.
Comments