In November 2022, the Californian company OpenAI presented its AI chatbot, which can hold conversations in natural language, to the public: ChatGPT. Just under two months later, Microsoft CEO Satya Nadella announced that Microsoft would be integrating AI models such as ChatGPT into all its products. Several thousand customers worldwide are already developing their own AI solutions based on generative AI models such as ChatGPT, Codex (creation of program code) and DALL-E (generation of images). They are provided to them by Azure OpenAI Service and combine the basic models of OpenAI with the regional availability of the Microsoft Cloud. This development is also supported by the Microsoft 365 tools by integrating AI as a Microsoft copilot in software for daily work.

Generative AI is thus becoming suitable for companies. It will provide support in day-to-day work, for example for creating texts and images as well as for meeting summaries or analyzing extensive reports. At the same time, generative AI can also deliver program code, for example for robotics applications, or sketches for new products and machines, or contribute to the improvement and automation of processes in companies.
The biggest success factor of generative AI is its accessibility: it allows us to communicate with machines in natural language without having to master program code or technical details. Management consultants McKinsey have calculated the opportunities that arise from this: The impact of generative AI on productivity could add between 2.6 and 4.4 trillion US dollars per year to the global economy. Just for comparison: Germany's gross domestic product in 2022 amounted to the equivalent of 4.3 trillion US dollars. According to McKinsey, the technology could help to compensate for "the overall slowdown in productivity growth in recent decades".

What can AI achieve in industry?

Generative AI has also sharpened the overall focus on the potential benefits of the technology: AI can accelerate the automation of production and value chains. One of the greatest strengths of AI is its ability to recognize patterns. It can do this much faster and more accurately than humans. AI can extract new insights from data and process this knowledge for forecasts - for example, about the utilization of production capacities, the sale of products or the condition of machines. If the underlying data quality is right and the models are trained for the intended purpose, the AI delivers reliable results.

The pattern recognition capabilities as well as the dialog and interaction capabilities of generative AI lead to exciting scenarios, for example in materials science. Here, AI can help to develop new valuable materials. In this process, known as 'inverse design', the required properties are first defined and only then are materials that have these qualities sought. This method can be used to find materials that are more conductive, have a higher magnetic attraction or are more resistant to corrosion, for example. Manufacturers can also use 'generative design' to develop lighter constructions and thus make cars more economical, for example.

Microsoft customers and partners are already testing exciting areas of application. For example, one user, a leading global automotive group, is integrating ChatGPT into its vehicles via Microsoft Azure OpenAI Service. This makes the built-in voice assistant even more intuitive to use and dialog-capable. In future, these voice assistants will be able to hold entire conversations, talk about interesting facts about the destination, suggest new recipe ideas or clarify knowledge questions.

The author: Gilbert Jacqué is Azure Solutions Go-To-Market Lead at Microsoft Germany.

© Microsoft

At the Hannover Messe, Microsoft showed how a German industrial company is using generative AI based on the Azure OpenAI service to accelerate development and maintenance, for example in error reporting in the product lifecycle management system, writing program code for machines or troubleshooting industrial robots in factory automation.
These use cases make it clear that AI interacts with other technologies and the industrial metaverse is emerging. For example, dialog-capable AI models together with solutions for the Internet of Things (IoT) are creating a real interaction space for people and machines in factories. Generative AI enables machines to speak our language for the first time.

AI helps companies to generate knowledge from data in order to make better decisions or solve tasks more efficiently. The better the available data basis, the more effective the application. However, this only works in an end-to-end digital environment - because analog processes do not generate digital data. Companies will have to evolve to achieve this. AI allows them to reach a new level of automation that covers a wide range of activities. It enables a new quality of corporate process design.

Artificial intelligence? Sure, but according to rules!

To ensure that AI is not only functional, but also safe and fair, the technology needs to be used responsibly. Microsoft demands this not only from itself, but also from its partners and customers.

© Microsoft

However, to ensure that AI is not only functional, but also safe and fair, we need a responsible approach to the new technology. Microsoft laid down the foundations for the responsible development and implementation of AI back in 2017. These principles come into play right from the start of every new project, as we have formulated six principles for 'Responsible AI by Design': Fairness, Inclusion, Reliability and Safety, Transparency, Security and Privacy, and Accountability.

Microsoft is not the only company to adhere to these principles: we also commit our partners and customers to the responsible use of AI. That's why all customers interested in Azure OpenAI Service must go through a special approval process for their applications. They define the planned deployment scenarios and these are checked by our specialized teams to ensure that they meet the principles for the responsible use of AI.
We also take concerns about data protection very seriously. As part of the Microsoft cloud, Azure OpenAI Service meets the same standards as all other Microsoft cloud services. This relates to data protection and security, the reliable availability of services and Microsoft's service promise regarding data sovereignty, which remains entirely with our customers.
The application examples mentioned above show how this works in practice: When using Azure OpenAI Service, companies retain sovereignty over their IT processes at all times. In addition, all data that is used is only accessible in the company's own cloud environment on Azure. Furthermore, no further training of the underlying AI models takes place with this data.

Work Trend Index - Relief through AI

AI can accelerate the automation of production and value chains. One of the greatest strengths of AI is its ability to recognize patterns.

© Microsoft

Every year, Microsoft publishes the 'Work Trend Index', in which we provide insights into the subjective well-being of working people. Among other things, the latest report provides valuable insights into the use of AI. According to the report, many employees are literally waiting for AI to take over their work tasks: 70% are willing to delegate as many tasks as possible to AI in order to reduce their workload. People also want AI to help them find the right information and answers, summarize their meetings and tasks and plan their day.
But as many as 49 percent also fear that AI could replace their job. For a responsible approach to AI, this means that the questions about jobs and qualifications must be answered in the interests of people - by companies and society. Everyone should see AI as a personal opportunity and be able to benefit from its advantages. This also includes the fact that these changes coincide with a demographic development that many companies are currently experiencing as a persistent shortage of skilled workers. AI is therefore also an opportunity to counter this shortage through automation.

Tackling opportunities and risks with determination

The first industrial applications in Germany and the enthusiasm for generative AI, which can be seen in the new 'Work Trend Index', show the many opportunities offered by the technology. However, part of the success story is also that we are concerned about the potential risks. Together, we need to focus on the challenges and risks with the same determination with which we are now exploring the new opportunities. Because AI is what we make of it together.