Generative AI – how big is the wave?

Word has spread that we are about to enter a new age of work with Generative AI. Word has also spread that this time it won't hit the jobs that require little prior training, but rather, above all, those that are done with the head. What many of us still don't realize is that it's not the future.

According to a study by CAIS, about one-third of Germans were aware of ChatGPT in February of this year. Around 50 percent were already worried that AI tools like this chatbot will impact the job market. More recent figures of comparable quality do not yet seem to exist. I assume that awareness has increased, probably also job concerns. But too little is happening.

One reason, cited by Ethan Mollick, associate professor at the Wharton School of the University of Pennsylvania, on the McKinsey Global Institute podcast: so far, change has taken longer to announce itself. He says that he sat out the NFTs and Web3 trends. They are typical examples to which the Gartner Hype Cycle might apply. That is, the initial hype is over, followed by disillusionment. Whether and when we will reach the path of enlightenment and reach a plateau of productivity remains to be seen. Let's take a very prominent example from the past where we can be sure that the Gartner Hype Cycle applies: the Internet. There was enormous hype in the 1990s, followed by the bursting of the dot-com bubble in the early 2000s. After that, the Internet established itself as a transformative technology that now permeates almost all aspects of business and life.

Why shouldn't the initial hype of Generative AI be expected to be followed by disillusionment? I see three reasons:

1. The number of experts. In his GatesNotes "The Age of AI has begun" Bill Gates writes that at the beginning of personal computing, the software industry could meet at an onstage conference. That is, there were only a limited number of experts. Today the software industry is one of the big industries. There are many experts who will help Generative AI to break through quickly. But not only that - when I look at the progress being made with ChatGPT, I see that it's not just IT professionals who are bringing the topic into companies, but also everyone who is already recognizing the potential for their own work.
2. Product maturity: here I revisit Ethan Mollick. "AI is here now. ... It's available to billions of people. It literally can write code for you. It can literally do reports for you. It can pass the bar exam." Since ChatGPT has been using the GPT 4 language model, the topic of hallucinations has been quiet. Yes, mistakes still happen. But they have become fewer. And word has spread that we can't blindly trust the language models. The product has evolved in a very short time, and so have the users.
3. Efficiency gains: Even early tests using ChatGPT as a texting assistant saw efficiency gains of around 30 percent. Estimates for improvements are as high as 80 percent, depending on the activity. What was that like with the steam engine? It produced an increase of about 20 percent, according to Ethan Mollick.

The efficiencies will come quickly and will give an immense advantage to those companies that understand how to use Generative AI. But they will only be an intermediate step. They will give us the breathing room to develop new business models that no longer rely on charging for time or brainpower. Because that's where we've got unbeatable competition.

How big is the wave that is rolling toward us? It's not a tsunami that will sweep away and destroy everything. It's more like one of those waves that the pros surf along so beautifully. Why can they do that? Because they have acquired the skills for it and have practiced a lot. It's time to do that with Generative AI.