How gen AI is changing our work

It is to be expected that the capabilities of AI will continue to increase. So, what might the working world look like in a few years after the breakthrough of generative AI (GenAI) based on large language models? "First of all: As fast as the technology is developing, even the near future is very hard to foresee. However, there is a trend that we can definitely see: If AI is introduced sustainably and intelligently in the company, then a clear added value often becomes apparent quickly," says Dr. Björn Bringmann, managing director of the Deloitte AI Institute.

When talking about AI in the company, it is very important from Bringmann's perspective that AI is not only seen as a technical issue but as an overarching C-level topic: "You can handle the technology if necessary, but AI must be strategically thought of from technology to compliance to talent - and is therefore a top priority. Keeping people in mind is crucial." The biggest change in the working world arises because generative AI with large language models can now also analyze data, the analysis of which was previously mainly the responsibility of humans. This includes the (statistically based) understanding of texts, such as notes written in customer service or maintenance - but also of images, audio, and video.

“Humans are no longer the entity that produces the result alone from a blank sheet, but will in many cases develop into a superior corrective and control entity that verifies and builds on the preliminary work of AI,” says the AI expert. Additionally, the written output is consistently and person-independently written in a high-quality language style. “We will experience that we as humans can always have a digital assistant with us that supports us in a variety of tasks - sometimes taking them on alone, sometimes interacting with us,” believes Bringmann.

AI empowers employees at all levels

Above all, the overview changes: Using the example of a sales employee, an AI solution can help find the right solution for individual customer requirements from a very large product range - even if it is not possible for humans to analyze thousands of products in seconds. For everyone new to a company, it will be much easier: Their AI assistant will explain exactly how the processes work and what the individually used abbreviations mean. They no longer need to know in which system information is located; instead, they ask a question and receive the answer in natural language.

Power users from specialized departments, who may only have limited IT skills, suddenly acquire entirely new abilities that are career-relevant. For example, no one needs to master the intricacies of Excel anymore: GenAI tools translate instructions into code, and AI helps interpret complex data relationships. A combination of low-code development platforms and LLM-based tools enables specialized departments to develop their own solutions and close process gaps beyond limited IT resources.

“I want to fully exploit AI as a transformation turbo”

In the interview, he talks about his strategic priorities, the importance of AI as a company-wide force, and the cultural change in the organization. His goal: Get AI out of the pilot phase - into widespread use.

Changed leadership roles

The requirements in leadership positions are also likely to change. It will be essential to sensibly balance the human-machine relationship, believes Bringmann: "Leaders will sooner or later manage not only people but also digital employees in their teams. They need to understand how projects can be planned or processes redesigned in this combination - which work steps can be intuitively completed faster or with higher quality by the machine, and what that means for management and time management," explains the AI expert.

Many companies have already started to make their accumulated data knowledge available through their own GPTs (General Pretrained Transformers). "The start is still ahead for many, but the pioneers are already a step further. They are looking at which areas they can use AI for their employees, their processes, and their product development," says Agnes Heftberger, CEO of Microsoft Germany. If, in the future, natural language - as in all our human communication - becomes the starting point, the question arises of how soon software solutions will be used, for which comprehensive training was previously necessary. "Perhaps a small look ahead: In the not-too-distant future, AI could become our most important user interface. You no longer log into individual applications. The starting point is then a prompt with which tasks are completed together with AI," postulates Heftberger.

The future belongs to multi-agents

Whether there will be an interface in the future from which interactions with downstream applications occur in natural language, or whether individual applications like CRM, ERP, or PLM will have such an AI interface, is currently open from Bringmann's perspective. "There will be a race because some providers already recognize that they could be cut off if they don't offer the best interface, similar to what happened with the platform topic," states Bringmann.

Especially for recurring tasks where many, even imprecise, pieces of information need to be processed, AI can increasingly take over in the future. Bringmann cites the example of responding to supplier inquiries about the payment process at OEMs. For an OEM, an AI agent was created that works "artificially intelligently" through countless systems to find out the respective cause. This AI agent provides an explanation of how it proceeded and automatically suggests a response email - for example, because a part of the delivery is still missing or when the amount will be specifically transferred. In the future, there will be many more such (networked) multi-agent systems that interact with each other and perform highly complex tasks. Mercedes CIO Kathrin Lehmann sees agentic AI, or "AI systems that perform specific actions," as one of the most important IT trends as early as 2025.

Above all, collaboration in globally operating companies becomes significantly easier. Training documents, training or customer videos are not translated time-consumingly or created multiple times with country-specific aspects. AI assistants handle the translation and are increasingly trained to automatically include country specifics. All information, recordings, or transcripts from meetings are immediately available in one's own language. Even for those who could not attend, their to-dos become directly visible. In the future, it might also be possible for everyone to speak in their own language and have it directly translated into the language of the respective participant.