Where we are, where we’re headed, and what’s important
Among companies that put out quarterly reports, half mention artificial intelligence within them.
Tech giants are investing billions in this area, and management consultancies are already generating significant revenue by helping clients implement AI.
But while some companies are making headlines with impressive increases in efficiency thanks to AI, many others are still scratching their heads.
One thing is certain: Generative AI will transform almost all industries in the near future. With its ability to understand, translate and generate language, AI can usher in huge boosts to productivity and quality in companies. That’s assuming human and machine are working hand in hand.
I believe the cost of getting to know AI – really getting to know AI – is at least three sleepless nights.
says Ethan Mollick, Associate Professor of Management at the University of Pennsylvania’s Wharton School, where he researches and teaches in the areas of innovation and entrepreneurship. He is a leading expert on the impact that AI has on work and companies. To use the popular cliché of the moment: AI is a “jagged frontier,” with many unknowns left to explore. This requires courage, perseverance, and the commitment to embark on a steep learning curve.
High time to pick apart the status quo and jump on opportunities.
From resource scarcity to simulated consciousness
Generative AI has been on an impressive winning streak – and at the same time, it’s unearthed a number of issues that we’ll continue to grapple with as a society.
On the one hand, it’s burning through resources like crazy: Training large language models such as GPT-4 consumes massive amounts of electricity and water – and that in turn has a huge impact on the climate and environment.
We’re also seeing a bottleneck with access to necessary training data: Either the quality is lacking, or there’s not enough of it, or data-protection regulations make in inaccessible.
And then there are the ethical challenges, not to be underestimated. How do we ensure that AI systems don’t take on the prejudices and blind spots of their programmers and make biased decisions?
How do we combat deep fakes and stem the spread of false information? The German digital association Bitkom recently conducted a survey that shows how urgently clarification is needed here: One in three people in Germany have never heard of deep fakes, and a further third have at most a rudimentary knowledge of the subject.
That should unsettle us. And what if the illusion of consciousness eventually makes communicating with a machine indistinguishable from talking to a human?
Humanizing the systems is a huge problem.
warns AI expert Fabian Westerheide, organizer of Rise of AI, one of the largest AI conferences in Europe, which just met for the eighth time in Berlin.
“We talk about AIs understanding things, but this is inaccurate, because they have no knowledge of the world and no values of their own.”
We always have to keep in mind that even the most sophisticated AI is still not a sentient, moral actor. It’s a statistical model that operates on the basis of historical data.
Generative AI sneaking its way into everyday corporate life
The pace of development is picking up. It feels like every week there’s a new breakthrough being announced by research labs and tech companies, from multimodal language models and AI-generated videos to complex planning and reasoning capabilities.
Many companies have already started testing the technology in specific scenarios. The investment bank Morgan Stanley, for example, uses GPT-4 to provide wealthy clients with personalized investment insights and recommendations. CRM specialist Salesforce uses the same language model to help its support team respond to complex customer queries.
Overall, though, most German companies are still lagging behind it terms of what AI promises. According to that survey, only 3 percent of German companies are already using generative AI, but for the vast majority, the technology is still uncharted territory.
Obstacles to becoming an AI-driven company
As AI continues to weave through more and more aspects of our world, it’s crucial that companies retain control over their data, processes and values. Most of the systems in use today, however, were developed in the United States and therefore don’t always meet Europe’s standards for privacy, transparency and ethical responsibility.
All the more reason for German companies to focus more on European AI solutions. Solutions developed here on the continent tend to understand our languages and cultures and also meet our high standards of data protection and transparency.
The good news is that many European start-ups and research institutions are already working on trustworthy AI systems like that.
Now we just need to work together on refining and expanding these approaches and introduce them into the broader user base. In that way, Europe will be able to keep up the pace technologically speaking in the age of AI and promote its values worldwide. But time is of the essence.
What companies need to do now
How can you approach the use of AI in your company systematically?
You’ll need to first ask yourself two crucial questions: “What do we produce, and how can we improve it with AI or create completely new products?” and “How do we work, and how can AI make us more efficient so that we can reduce costs and increase quality?”
Start with manageable use cases, learn as you go and gradually expand your scope. One thing that’s crucial here is a positive learning culture that leaves room for experimentation.
How man and machine can collaborate successfully
Another important factor, especially when using generative AI, is how man and machine actually interact. In his latest book, Ethan Mollick outlines four guiding principles for successful collaboration that I fully support:
1. Always invite AI to the table.
Using AI in every area of your life and work helps you learn about what it can do as well as its limitations. By doing so, you can begin to understand where it can be helpful and when it can pose a threat.
2. Be the human in the loop.
Even though AI can do more and more things, humans still need to retain control and responsibility. It’s important to remain vigilant and involved to avoid becoming too dependent on AI.
3. Treat AI like a person (but tell it what kind of person it is).
In many ways, AI behaves more like humans than another piece of software. It makes sense to assign the AI clear roles and specifications so that you can leverage its full potential. But don’t forget: You’re not dealing with a real person.
4. Assume this is the worst AI you will ever use.
AI is just going to keep on getting better, evolving at a rapid pace. Cutting-edge today; obsolete tomorrow. The key is to remain open to new, improved AI systems in the future.
Moving forward: Don’t shrink away from technology; dive in head-first
Just one more thing: Generative AI is a powerful force, and it’s not going away. Even now, it’s changing the way we communicate, plan, design and program. It’s already clear that, sooner or later, almost every business and workflow will be impacted and transformed by this technology.
For companies, this means two things: On the one hand, they must rise to the challenge, get out in front of it and develop strategies for wrangling AI for themselves. This is no time to drag your feet, because the competition never sleeps, and technological progress is going to keep on accelerating. Those who are actively engaging with AI now will most likely have a say in where their industry is headed in the future.
On the other hand, companies also need to take responsibility for any system they use. After all, generative AI is neither all good nor all bad – it depends on how we humans develop and use it.
Companies have a responsibility to deal with ethical issues related to how data is protected, the fairness of the models and to what degree the generated results can be explained. And since they’re driving this technological change, they also need to keep societal impacts in mind. That includes the future of work.
AI is just another a mirror of humanity, reflecting all our strengths and weaknesses. A blindly devoted techie is just as reckless as an uninformed Luddite. We hold the reins when it comes to where this tech can take us. Without us, technology doesn’t exist. But pretty soon, there will be no us without them.
In Fabian Westerheide’s words: “We actually need a digital revolution.” That’s his plea to companies to waste no time in exploring the potential that technology holds, to fully exploit what AI can offer in terms of value, growth and prosperity. All the while, working hand in hand with regulators, researchers and the general public. Let’s get to work – with curiosity, prudence and confidence!
—
About the author: Jörg Müller is an experienced digital expert who lived through the early days of another disruptive technology – the internet – both as a witness and by actively guiding the impact it had on companies. During his many years at a media conglomerate, he held various roles, including Product Manager, Agile Coach and Innovation Manager. It was in these roles that he first began to realize the enormous potential of generative AI during the winter of 2020/21, when it was still in its early stages. This initial exposure prompted him to research possible applications.
These days, Jörg Müller advises companies on how to find suitable use cases for AI and helps them to develop pragmatic solutions that will work for SMEs. He strongly believes in explaining both the capabilities and limitations of the technology.
E-mail contact: joerg.mueller@co-intelligence.ai