There are countless studies predicting the destruction of millions of jobs due to generative AI, which would put an end to many professions in the process. Klarna's recent change of heart doesn't mean a complete reversal of the situation, but demonstrates a more pragmatic understanding of the limits of automation, particularly in sensitive interactions where empathy, judgment, and nuance remain difficult to replace.
AI is not enough
Klarna, one of the world's leading specialists in deferred payment, was one of the first companies to collaborate with OpenAI in 2023. For a year, the Swedish startup had simply stopped recruiting to focus on developing its AI capabilities. By 2024, Klarna had claimed that generative artificial intelligence was now doing the work of 700 customer service agents. An announcement that caused the stock price of French group Teleperformance, a customer service giant, to tumble.
Sebastian Siemiatkowski, CEO and co-founder of Klarna, acknowledged that this strategy was no longer appropriate. "Unfortunately, cost has been too predominant an evaluation criterion in this organization, which inevitably leads to lower quality," he explained at the Stockholm headquarters. "Really focusing on the quality of human support is the way forward for us." A spectacular about-turn!
"I think it's essential that the customer knows they can always speak to a human if they want," the executive added. To achieve this, Klarna intends to follow the "Uber model": agents can connect and work remotely whenever they want. The thousands of humans employed by subcontractors will eventually be replaced by these new types of agents.
Ultimately, Klarna is adding precarity to precarity, but cloaked in a narrative of flexibility and "community engagement." The company is targeting profiles such as students or rural residents, and even its own most loyal users. Two pilot agents are already in place, proof that the transition to this hybrid model between AI and decentralized humans is well and truly underway.
This hybrid system also demonstrates the risks financial services companies face when they replace humans with technology that is still largely unproven. In another sector, that of airlines, Air Canada was found responsible for the hallucinations of its AI bot… This may be food for thought for companies that are diving headfirst into AI at full speed.
Source: Bloomberg
0 Comments