Future Says S4E5: Building AI Communities in the Automotive Industry
Until very recently, it was relatively easy to trace the roots of automotive manufacturing back to the days of Ford’s iconic Model T. Basic elements in automotive manufacturing such as the role of the driver and the ubiquity of the internal combustion engine remained virtually unchanged for a long time. What’s more, many manufacturing processes still echoed the ideas Henry Ford first introduced.
However, new and innovative concepts are replacing the principles that have long underpinned the automotive industry. Inevitably, megatrends such as electrification, autonomous driving, and connectivity are putting new technologies in the spotlight. At the same time, broader global forces are also exerting new pressure on the automotive industry and those within. So, while they’re facing a multitude of technical challenges, manufacturers are also simultaneously trying to manage unprecedented changes in the cultural, ethical, and legal landscape.
The Business of Ethics and Regulation
Nowhere is this more apparent than the fields of data science and artificial intelligence (AI). At the C-suite level, the talk is all about opportunities for business transformation. But in some cases, enterprises appear to be struggling to put the right people in the right places to achieve the right results. And the impact goes beyond the search for an acceptable return on investment. Organizations also face the daunting task of establishing appropriate ethical standards for the safe use of AI. What’s more, they must stay one step ahead of a complex, rapidly evolving regulatory framework for AI, AI communities, and data use that’s raising serious questions about how enterprises will comply with some of the proposals lawmakers around the world are currently considering.
These issues are central to Johan Kölhi, our guest on episode five of Future Says season four. As a partner at Accurate Consulting and the group’s head of innovation and digitalization, he’s engaged by blue-chip names in the automotive supply chain to support the roll-out of successful AI- and data science-powered solutions. He’s a firm believer in the importance of building flourishing AI communities within business.
The “Magic Mix” of Data and AI in Automotive
Kölhi brings an interesting perspective to the debate. A computer scientist by training, his career has taken him into fields that include innovation, patents, cyber and information security, and data protection. As such, he’s used to looking beyond the purely technical side of developing AI products to consider the ethical questions involved and the likely impact of legislation such as the European Union’s forthcoming AI Act.
In terms of the skills needed to drive successful projects, Kölhi is very much in tune with previous Future Says guests. Data scientists need to work side-by-side with subject matter experts to create what he calls the “magic mix.” Picking up another common thread that has run through recent Future Says episodes, he thinks data scientists must also appreciate the bigger picture. “Coming from the tech side, it’s super important to understand why you do this,” he said. “Ask yourself: What’s the business case? What’s the win for this manufacturing line or what is the win for the company as a whole?”
Keeping it Simple
Drawing on his experience working with a major global supplier of automotive components, Kölhi’s approach to fostering AI communities is informal, accessible, and organic. From an early stage with this client, Kölhi has scheduled short, but regular meetings that welcome anyone interested in AI and AI communities. Participants have the freedom to brainstorm, share ideas, learn from one another, demonstrate potential solutions, and explore new opportunities together. He’s found that these meetings have grown naturally and encouraged the emergence of AI champions and AI communities throughout the business. In terms of content, he’s a strong believer in running practical demos rather than simply sticking to conventional presentations.
Alongside data science and subject matter expertise, Kölhi says these AI communities should be looking to embrace ethical and legal know-how. In this respect, his views on the forthcoming AI Act in Europe are mixed. In principle, he believes its ideas and intentions are good. However, he expects significant problems in terms of how some of the current proposals might be applied in the real world. For example, when it comes to high risk systems, he says: “It’s very, very hard to show how the AI system came to this conclusion, based on the input that you have.” In its current form, the AI Act also demands the ability to show how a system performs in a sandbox. “It’s something that sounds really good on paper – but it’s almost impossible from a technology side to make it work.”
Even if it turns out to be less than perfect, organizations need to accept that legislation such as the AI Act will soon be a fact of life. For Kölhi, that means putting the necessary processes in place now. One example he cites is identifying proposals that might be classified as “high risk” under the forthcoming rules. By his own admission, keeping an eye on regulations “sounds horribly boring.” But he points out that sanctions for failing to comply with the AI Act could be even more punitive than those imposed for GDPR breaches.
The Future of AI in the Automotive Industry
Kölhi is well aware of AI’s potential risks and the need for safeguards against those risks. On the subject of ChatGPT, he recalls the tool was once compared to “having a friend with thirty doctorate degrees, who also lies a lot.” That means users must be conscious of the need to fact-check any results these tools generate.
For Kölhi, these risks emphasize the need for broad data literacy among people of all personas. Organizations should be looking to educate as many job functions as possible on the basics of AI, encompassing both the technology’s benefits and pitfalls. In this respect, interconnected AI communities can serve as bastions of knowledge.
Above and beyond any regulatory obligations, organizations also need to put measures in place that ensure the safe, ethical use of AI. And within these safeguards, Kölhi is keen to encourage people to explore and experiment with the potential of AI and data science. Overall, he’s excited about what lies ahead. In particular, he sees large language models (LLMs) rapidly evolving into accessible and universal digital assistants capable of producing stable results from raw inputs as basic as spreadsheets and images, and in response to prompts as straightforward as “How can I fix this?” or “Help me improve this.”
Conclusion
All this brings us to a core theme of both Future Says and Altair as an organization – democratization. At first glance, the world of highly intuitive generative AI might seem a long way from the era of the Model T. However, at the most basic level, Henry Ford wanted to put the freedom of the automobile within everyone’s reach. As a result, it’s safe to say there are plenty of parallels to the past, even if they appear very different on the surface. As AI continues to mature and evolve, hopefully it will exert a similar democratizing force around the world.
Click here to listen to the full episode with Accurate Consulting’s Johan Kölhi. To learn more about the Future Says series and browse previous episodes, visit https://altair.com/future-says.
Future Says Season 4 is proudly sponsored by Oracle. Oracle offers integrated suites of applications plus secure, autonomous infrastructure in the Oracle Cloud. For more information about Oracle (NYSE: ORCL), please visit http://www.oracle.com/.