Context 

IW Consult has conducted a comprehensive study on behalf of the Eco Association on the relevance of AI data centres for Germany’s economic development. The result: data centres are the backbone of Germany’s digital transformation. The use of AI generates additional value creation potential of 330 billion euros. Companies that use data centres are more innovative and more successful. Germany as a whole could increase the annual rate of increase in labour productivity from 0.8% to 1.3%.  

Data centres and the use of AI therefore play a key role in productivity, competitiveness and overall economic growth.      

About 

Dr. Henry Goecke 

Bela Waldhauser   

Interview  

Gregor: In the first part of our interview, you described how investments in AI infrastructure fundamentally lead to economic growth. In this second part, I would like to understand how this effect works in practice. If politicians and businesses spend a few billion, how can this be turned into sustainable economic growth?hat?  

Bela: In order to build a vital AI ecosystem in Germany, we need investment in several areas. It all starts with the availability of cheap energy. Above all, we need more power plants to meet the growing demand for it. As this supply is becoming increasingly volatile due to the switch to solar and wind, transmission grids, substations and storage technologies also need to be expanded. Furthermore, the data centres themselves and the development of local waste heat consumers, usually in the form of hot water, are also important.

Incidentally, the demand for water in modern data centres is generally overestimated. We operate here with closed circuits that do not draw any water during operation.  

What is often forgotten, however: Although the costs for a data centre are high, the majority of the investment is spent on what is inside. This can be explained quite well using an investment by Microsoft. In the press release, the colleagues spoke of three billion for a data centre project. I then did a rough estimate of the framework data and realised: The actual data centre could have cost a maximum of 1 billion. In other words, the rest of the money went on IT hardware and software.     

Gregor: But how does the economic miracle come about? 

Henry: At its core, the AI transformation is a continuation of digitalisation. Because we have been conducting studies on this for years with recurring questions, we can clearly describe these effects. 50% of companies use some form of the cloud, 25% of companies cannot operate their business model without it. Companies that use the cloud and AI are growing faster with new products and are therefore more innovative. In total, 125,000 well-paid jobs have been created every month in the last 3 years as a result. In total, around 6 million jobs in Germany now depend on cloud use. This is all happening at a time when traditional sectors such as banking and automotive are cutting jobs.  

If you look at how much revenue successful digital companies such as Microsoft, AWS and Google, as well as German software companies such as SAP and Celonis, generate per employee, you can see how much more productive these jobs can be. This effect is further amplified by artificial intelligence.  

Gregor: So the most important economic effects don’t come from you, Bela, but from the ecosystem around you? 

Bela: Yes, exactly. Of course, we also employ staff, such as electricians, air conditioning technicians, security experts and some traditional employees. But the economic added value of digital infrastructure is primarily generated by the companies that use this infrastructure. One example of this is Daimler, which has relocated its data centres from Stuttgart to Frankfurt, while the IT infrastructure continues to be used for the Group.  

As a Frankfurt resident, I can confirm Henry’s analyses in practice: banks and automotive suppliers have been cutting jobs here for years. Most recently, for example, Commerzbank cut 4,000 jobs. But the digital industry is desperately looking for staff and pays above-average wages.   

So it’s about the whole cosmos surrounding the cloud and AI, digital parks, start-ups and the digitalisation of our economy.  

Gregor: So even the best AI infrastructure is useless if German companies don’t follow suit?  

Henry: That’s why we’re proposing a special depreciation allowance for AI investments.   

This would primarily support small and medium-sized enterprises in their digitalisation efforts. We see a need here for the entire range of IT modernisation: switching to flexible and scalable cloud infrastructures, modernising applications, improving cybersecurity, automating internal processes, more agile organisations, predictive maintenance, software and platform-based business models and new sales and marketing channels. All of this means an enormous effort for SMEs. In contrast to many large companies, they lack both the experts and the capital.  

With the special depreciation allowance, we are supporting them financially and giving them the right incentives. Cloud and artificial intelligence then help them to operate more competitively with fewer staff. In this way, we are accelerating the pace of our digital structural change. 125,000 new digital jobs per month could then become many more.  

Bela: As data centre builders and operators as well as the energy industry, we would then be left with the task of satisfying the ever-increasing demand for our services. Cross-industry coordination of investments by those involved and less or better harmonised regulation would help here.  

Gregor: Bela and Henry, thank you very much for your time!