Insights Into the State of AI From Web Summit 2023

Kelly Rutherford bei der Web Summit 2023

Improving Productivity with GenAI

The number of companies showcasing AI solutions, such as virtual assistants and GenAI tools, clearly demonstrated the power of AI to improve productivity and drive business growth. Jan Rezab, Founder and CEO of Time is Ltd, talked about how AI is a relief to the many companies struggling with unproductivity by minimizing the time and cost it takes to complete a project. This sentiment was echoed by the "Will GenAI transform marketing and advertising?" panel, which talked about how while AI can be used to cut costs, the much more exciting possibility is to use existing resources in combination with new tools to accomplish projects that were previously impossible due to cost and resource constraints. However, Data Scientific CEO Cassie Kozyrkov clearly pointed out that AI is just a tool to improve execution. The human remains the artist behind the scenes. She makes this point by distinguishing between thinking and thunking. Thinking is how you want to see the world, imagining the solution - the creative part. Thunking is the boring execution - the mindless tasks it takes to go from idea to execution. AI can shorten the thunking time, but the thinking remains 100% human. Keozyrkov ended the talk by making the point that in AI, we often talk about taking humans out of the loop, but it still requires human judgment to know if the output is good. And when you automate at scale, you still have to test, check, and maintain the systems. And because everything is happening at a faster pace, it may even require more labor.

Staying Unique

What was clear throughout Web Summit 2023 is that we are currently at the peak of the AI hype cycle, and we need to be more intentional about how we use these technologies going forward. Many speakers, including Kelly Rutherford, actor and investor at Whyzzer, and Lars Silberbauer, CMO at Nokia Phones, talked about the digital fatigue we're collectively feeling. Going forward, we need to find a way to integrate it into our culture and work in a way that adds value rather than creating a sea of sameness. People are increasingly fed up with quick-fix content, and the rise of AI-generated material only exacerbates this problem. That's why injecting the human experience into the process remains critical. As people increasingly seek more meaning and community online, brands will need to be more intentional in their use of emerging technologies. One opportunity here seems to be custom-training AI systems based on your own data, as demonstrated by Burkhard Müller and Heinrich Paravicino of Mutabor, who showed how they use AI to create imagery that is unique to a brand.

Regulating AI

When it comes to regulating AI, we are entering uncharted territory. Andrew McAfee, Principal Research Scientist at MIT, emphasizes the importance of regulation for high-risk systems and advocates upstream governance in these sectors, where requirements need to be met before the technology is implemented. However, he acknowledges the trade-offs involved, as these requirements can stifle innovation in other areas by requiring more financial resources, legal support, engineering effort, and time. McAfee suggests a balanced approach by advocating permissionless innovation in all other areas, although emphasizing that permissionless innovation doesn't mean absolute freedom from regulation - it's just a timing issue. To illustrate, he cites the example of "a couple of idiots" in America who when smartphones came out, used them for invasive photography. Although there were peeping tom laws, they did not explicitly prohibit this behavior at the time. However, when the incidents occurred and were reported, legislative action was taken within two days to correct the oversight. Echoing Jimmy Wales' philosophy, in his talk, McAfee advocated for a system that allows mistakes to be corrected rather than making them impossible in the first place.

In Short