DT wants to save the world from 'shit' algorithms

DT's Claudia Nemet spoke about the company's human-centric approach to AI and the opportunities presented by faulty AI algorithms.

Tereza Krásová, Associate Editor

June 20, 2024

3 Min Read
TM Forum's Nik Willetts and DT's Claudia Nemet in conversation at DTW.(Source: Tereza Krásová/Light Reading)

Deutsche Telekom's Claudia Nemat has highlighted the importance of ethics and a human-centric approach in the telecom industry during a keynote session at the TM Forum's DTW Ignite event in Copenhagen this week.

"Technology is not there for the purpose of technology," said the German operator's executive board member for technology and innovation, adding that "monetization in itself is not the purpose in itself. You want to be monetized. It is the result of being respectful to people, to communities and the world and offering something to them which is meaningful."

It is perhaps rare to hear a telco executive argue that monetization is not the be-all and end-all for the industry. Then again, DT is arguably performing better than many of its European competitors – speaking of the company's performance, Nemet said DT is currently growing in all its markets. She also called it "the only cross-Atlantic telco in the world" – a point that Telefónica, which operates in several Latin American markets, or TIM, with its Brazilian "cash cow," may disagree with.

Nemat argued that "it is up to us" to drive ethical and responsible AI and technology, noting that we live in an era when it is impossible to trust what one sees and hears – presumably referencing AI.

"Let's be the most ethical, the most human-centered industry in this world and what we did wrong decades ago when we over monetized our customers and others could disrupt us, now we disrupt the disruptors because many AI algorithms are just shit, because they actually increase engagement but not truthfulness," she said.

She added there is a large opportunity to "disrupt whatever is shitty and create opportunities that help us, as some of the CEOs said, to save mankind," referencing previous speeches on the day.

Human-centric AI

Asked by Nik Willetts, the CEO of the TM Forum, what human-centric technology means to DT, Nemat said it is determined by three principles, which she summed up as "be beneficial, don't create harm and be frugal." 

DT's first principle dictates that technology should have a practical benefit. Nemat pointed to a concept app-free smartphone showcased at MWC this year, in cooperation with Qualcomm and Brain.ai. It would act as a "concierge" that would take prompts from the user and execute them – e.g. book a flight – to save time, as Nemat argued.

One may wonder, though, how long it may be before such a phone makes its way to the market, and how much DT would contribute in technology terms – or indeed benefit in financial ones. 

The second principle is that technology should not create harm, and here Nemat brought up what may be seen as a more real-world example. She noted that the company introduced ethical AI guidelines back in 2018, on which basis it stopped using certain algorithms in Austria that would not recommend certain tariffs to elderly people. The grounds for this – as one might guess – were that it was considered discriminatory.

The third and final principle requires the company to be frugal – something that telcos are already pursuing with zeal on the capex side – but in this case it has to do more with saving resources and preventing wastage.

Expanding on this definition, she also noted that data is crucial to understanding customer needs, noting that each has different requirements. "The CEO of a logistics center said to me 'Claudia, I want the logistics center out of the box,' and the developer wants to have easy access to the network APIs."

Read more about:

EuropeAI

About the Author(s)

Tereza Krásová

Associate Editor, Light Reading

Associate Editor, Light Reading

Subscribe and receive the latest news from the industry.
Join 62,000+ members. Yes it's completely free.

You May Also Like