The next phase of AI - five key perspectives
08 | 2025 Jari Huilla, Kipinä CTO & Partner
Artificial intelligence is a hot topic in both technology discussions and corporate strategies. AI strategies are playing an increasingly critical role in companies' visions for the future. Few topics are as promising and at the same time as misunderstood. Over the summer, Kipinä's experts published a blog series presenting five different perspectives on what the development of next-generation artificial intelligence solutions really requires.
One of the most important findings was that building artificial intelligence does not start with code, but with the question of where it really fits in and what problem it solves, as Kipinä's CTO Jari Huilla explains in his blog. Applying artificial intelligence to all possible situations sounds appealing to many, but in reality, the best results are achieved with precisely defined and understood problems. However, understanding requires the ability to look under the hood, from model architectures to their mathematical constraints.
High-quality, clearly structured, and controlled data is the cornerstone of artificial intelligence, and it must be included at the design stage, not added later, says Kipinä's partner and Data Architect Antti Rintala. A good AI solution always starts with good datais the idea on which the foundations of next-generation AI solutions are built. Only solid foundations can create scalable solutions that stand the test of time.
The third insight is provided by Case PlanetAI, where Kipinä's development team helped turn an ambitious vision into a production-ready service that helps global companies verify their sustainability efforts. The project demonstrated that a successful AI solution requires a combination of technological expertise, business acumen, and the ability to build a solution that scales and stands up to international comparison.
Despite its almost limitless possibilities, technology never exists in a vacuum. Information security and responsibility are an essential part of the design and implementation of every AI solution, emphasizes Senior Software & Cloud Engineer Erkka Korpi. The more systems are automated and interconnected, the more important it is for humans to maintain control—and understand when to stop and ask: “As coders and customers, are we sufficiently aware of the race between AI and information security in software development? Do we recognize the risks of haste and ever-increasing speed?”
When discussing artificial intelligence, its future, and its possibilities, we cannot ignore the role of humanity and empathy. Psychological safety, openness, and the ability to view one's own decisions from different perspectives are prerequisites for AI to truly support collective thinking, rather than merely accelerating existing practices. Artificial intelligence can be a powerful tool to support thinking and decision-making, but its potential will remain untapped if we do not dare to be incomplete and ask difficult questions.
By combining the five most important perspectives, we can form an overall picture of the “second phase” of artificial intelligence:
Precise problem definition
Strong data foundations
Implementation that serves business
Data security and responsibility
Humanity alongside technology
Developing artificial intelligence is not just about the latest technology. It is about the ability to build solutions that are simultaneously effective, safe, and meaningful. It is a journey that begins with questions—and continues with the courage to look more closely at both the data and ourselves.