At the EIB Innovate Day 2024, Elin Hauge, AI and business strategist, emphasized the importance of opting for subscription-based services rather than relying on free models, highlighting that when a service is free, it’s often user data being used as payment.
EIB Innovate Day 2024 brought together global thought leaders, industry experts, and innovators to explore advancements in technology and their transformative impact on society. With a spotlight on artificial intelligence, the event featured compelling keynotes, interactive workshops, and real-world case studies showcasing AI’s potential to drive innovation, solve pressing challenges, and create a more equitable future. “So I think it’s a combination of the experiences we’ve shared with colleagues about what we do within the EIB Group, along with examples from other industries, even if they are not in the financial sector. Today’s event with many stands and our keynote speaker has enabled them to talk to experienced people, attend workshops and learn something new. So, I think that’s a valuable learning opportunity,” said Kaisu Christie, Group Chief Digital Officer at the EIB.
Combining expertise in physics, mathematics, and business strategy, Eline Hauge, delivered a presentation that demystified the complexities of artificial intelligence while inspiring the audience to think beyond technology and focus on the human story. “I think what we have in society, right now, is everybody is trying to grapple with AI. On one hand we have those people that are kind of completely against it or they see doomsday scenarios, and this criticism is so strong that it doesn’t really get us anywhere,” said Hauge.
AI can elevate human potential, solve global challenges, drive innovation, and pave the way for a more equitable future, when approached correctly.
“Let’s try to do things smarter. And that also requires that we do software development and algorithmic development smarter because code can be written in good code and daft code.”
Elin Hauge, AI and business strategist
AI not a silver bullet
AI has been described as a powerful toolbox with compartments that serve different purposes, from interpretation to prediction to generation. So while technologies like ChatGPT have gained immense popularity, they represent only a small corner of what AI encompasses. “It is mathematics, not magic. AI is a toolbox, and ChatGPT is just a tiny part of it,” said Hauge.
AI’s applications, like the early days of image tagging by Google and IBM, evolve from “wow” moments to becoming mundane tools embedded in our everyday lives. Leaders need to focus not on individual tools like ChatGPT but on foundational models and systems that address specific organizational challenges.
From customer service to development
Customer service transformation is seen through Klarna’s adoption of GPT-4, which replaced 700 full-time employees by handling over 2.3 million customer service conversations monthly across 35 languages. The result was an expected first year financial improvement of $40m. This example demonstrated how AI can deliver efficiency. “What’s better than excellent customer service? No customer service. Let’s solve problems before they even appear,” said Hauge.
AI in software development is a clear example, as developers now use AI driven tools like GitHub Copilot to draft code, saving time and effort. However, there is a growing issue of “code rot,” and the risk of accumulating poorly reviewed or outdated code, a call for responsible AI use in software engineering. Legacy system challenges pose untapped potential for AI in modernizing legacy code, which often becomes obsolete as programming languages and developers fade into history. Large language models could play a critical role in reviving such systems.
“So let’s at least not scale the problem too fast. Let’s try to do things smarter. And that also requires that we do software development and algorithmic development smarter because code can be written in good code and daft code. And daft code requires a lot of Power, a lot more power than good power,” said Hauge.
“96% of executives expect productivity gains from the use of AI.”
Elin Hauge, AI and business strategist
Challenges: bias, cybersecurity, and strategic missteps
Bias in data and models is apparent and inherent in data today because it reflects human behavior and decisions over decades. Organizations must identify and address bias to ensure ethical and effective AI solutions, particularly with impending regulatory requirements like the EU AI Act. “What happens when we’re wrong? We need to understand that so that we apply these technologies in ways that minimizes the risk of consequences when we’re wrong, because we know it’s going to be wrong. Humans also make mistakes, but that’s much more random,” said Hauge.
Generative AI is taking cyber threats to a new level, making phishing scams and data pollution increasingly sophisticated. “I urge vigilance, especially regarding seemingly harmless actions like clicking links or scanning QR codes,” said Hauge.
Many leaders equate AI strategies with acquiring ChatGPT licenses, which is viewed as a strategic misstep. Instead, organizations are advised to build tailored applications on top of foundational models to meet their specific needs.
Beyond the AI bubble
Looking ahead, there will be a less hyped, more integrated relationship with AI, where the technology becomes seamlessly embedded into everyday digital tools, akin to how Spotify uses recommendation engines. However, it is predicted that there will be job losses, as routine tasks are automated, prompting individuals to redefine their roles. “We need to train board members and C-levels on what these technologies are really about,” said Hauge.
Adaptability and learning are crucial. Education might become optional in an AI dominated world, asserting that the most successful users of AI will be those who are highly competent in their fields, and capable of mastering new tools.
The human-Centric AI vision
AI reflects the data, and, by extension, the values of its creators. The future of AI will depend not just on technological advances but on human intention and responsibility. Leaders are urged to focus on role understanding rather than specific tasks “It’s not about which activities you do but about the purpose of your role,” said Hauge.
There is an urgent call to think critically about the ethical, strategic, and human dimensions of AI. This highlights a striking disconnect between executives’ expectations and employees’ real world experiences with AI tools. It raises critical questions about whether decision makers fully understand the practical challenges and usability of these tools, especially given that many executives themselves rarely use them. “96% of executives expect productivity gains from the use of AI. Whilst at the same time, 77% of employees experience reduced productivity. What is this really about? How many of you are actually using these tools on a daily basis? And moreover that is the question,” said Hauge.
