T-Systems-Claim-Logo
Search
Audio frequency made up of dots and blue lines

How do humanity and AI blend to create masterpieces?

The interaction between humans and AI is a decisive factor in whether sustainable AI projects are possible today and in the future

December 09 2021Pavol Bauer

Three components for AI

Faster automation, optimized processes and increased security – all of this is possible with AI. Nevertheless, according to Gartner, 50% of IT managers will have difficulty making their AI projects ready for production by 2023. This does not surprise me, as three success factors are often underestimated in AI projects: human expertise, the correct data, and their storage location, which must be compliant with data protection guidelines.

Algorithms revel in harmony

Robot hand on piano keys

The creative qualities of algorithms are not often a subject of conversation for customer projects. This made the Deutsche Telekom experiment “Beethoven X – The AI Project” all the more fascinating – a project where the Beethoven Orchestra in Bonn completed and performed the world premier of Beethoven’s 10th symphony with the help of AI. What particularly interested me was the quality of the results and the extent to which AI can support people in music and creativity. How human is artificial intelligence? The experiment shows that AI cannot exist without people. In fact, the recipe for success is the interaction between humans and technology. AI can only offer added value if it is trained and implemented correctly by humans. AI can handle lots of data better and faster, recognize connections, and find errors. But value judgments can only be made by humans, which shows the limitations of AI. It does not know what feels right. Or does it?

While AI can do a lot and learn very quickly, can handle a lot of data, recognize connections, form hypotheses, and find errors, it doesn't know what feels right.

Tim Höttges, CEO Deutsche Telekom

Does AI know what feels right?

Hand touches digital screen with dots

Opinions on AI projects often vary widely. Some people are scared of AI. But is this justified? I don’t think so, because ultimately human emotional intelligence is always involved. Therefore, AI could never take over the core tasks of care-givers, for example. It can correctly diagnose the symptoms of a patient but doesn’t know how to show empathy or observe ethical and moral principles. This is a pattern that we can apply to AI projects in general: without people who have comprehensive specialist knowledge of AI and therefore can fully understand the context of AI-based results, it isn’t possible to create and manage a robust AI pipeline at production level. Expertise is especially important in the scaling of small pilot projects across company-wide applications.

Can we trust AI?

Take the following use case: FUSE-AI offers an AI solution that identifies anomalies in MRI scans. But can we trust AI with sensitive data?

There are rules in place to comply with ethical and data protection guidelines:

  1. The proposed legislation from the EU Commission that should prevent discrimination or surveillance using AI, for example.
  2. Developers can check their AI applications and solutions against the AIC4 catalog.
  3. At T-Systems, we ensure on the one hand that AI systems and their use comply with our company values, ethical principles, and social conventions through our AI guidelines. We also make sure that data processing for AI applications is carried out in highly secure European data centers compliant with GDPR with our Sovereign Clouds, Open Telekom Cloud and our future Google Sovereign Cloud. This is our DNA. It is no mistake that our CTO Max Ahrens is a member of the board for the Gaia-X initiative, which is advocating for an independent European cloud infrastructure.

What makes AI successful?

Recently, I read that only 14.6 percent of companies have been able to implement their AI projects into production processes. Are most AI projects doomed to fail? Not at all: if you have the right combination of data, technologies and expertise, you have a very good chance of success. AI projects often fail because they don’t have these elements. In addition, the wrong type of questioning or a lack of reproducibility can threaten AI projects. We can help you overcome these hurdles and simplify the scalability of AI through our AI platform using the Open Telekom Cloud, which allows us to make completed AI models available directly from the cloud. And: as part of the AI Solution Factory, the T-Systems PaaS team wants to introduce a new AI platform in 2022 for customer-specific models that can then be used for all established cloud environments.

What can humans and AI achieve together?

Often the biggest hurdle for AI is a question of perspective. If we spin for instance this question “How human is AI?”, we get: “How artificial are humans?”. It is therefore clear that we already live with and trust AI today, whether it’s for help with operations, or in the form of a pacemaker or prosthesis. AI can add value that we would not achieve without it. Even in areas never imagined, such as the completion of musical masterpieces. So how much AI is allowed? Where does it need human regulation? Even if I do not know all the answers at the moment, one thing is clear: if we stick to consistent international rules for AI use, we can maintain the humanity it takes for humans and AI to work together to create more masterpieces, now and in the future. This is because with AI, we can build foundations for a more sustainable future – whether this is through the optimization of the circular economy, more efficient energy management, or even intelligent emissions monitoring. 

About the author
AU-Bauer-Pavol (1)

Pavol Bauer

Senior Data Scientist, T-Systems International GmbH

Show profile and articles

This may also interest you

Do you visit t-systems.com outside of Germany? Visit the local website for more information and offers for your country.