Is AI Really Intelligent? It Doesn’t Matter—What Matters Is Solving Real Business Problems!

When I read many posts, interviews, or reactions about AI, I’m struck by the focus on AGI (Artificial General Intelligence) or Super Intelligence—the so-called singularity point when humans might no longer be the most intelligent species on earth. It’s easy to see why this vision is captivating. Much of the conversation is shaped by influential leaders who, while certainly knowledgeable, are also driven by their own interests. When Sam Altman from OpenAI says intelligence will continue to grow at an exponential rate and completely transform the world, he’s likely right—but he’s also making a case for the billions of dollars invested in OpenAI to become the dominant player (the “Google of Intelligence”). Conversely, when Yann LeCun argues that current generative AI models (like transformers) will never fully replicate basic human or even animal abilities, he’s also right—but he’s promoting his own research in vision, physical world understanding, and symbolic reasoning, which complements language models.

And then there are others: some claim the sky’s the limit for AI due to its exponential growth, while others dismiss it entirely because it can’t do something as simple as counting the letters in a word (a limitation that was by design in many generative models, though newer versions of models are improving in this area). But using an LLM to count letters is missing the point—it’s like trying to use a hammer to drive in a screw. The debate is passionate, and it’s easy to get swept up in extreme views.

To business leaders, I urge: don’t get caught in this whirlwind of opinions, especially about AGI. Instead, focus on what’s happening right now and the opportunities AI presents over the coming years. The truth is, the real breakthrough with generative AI has been up to now mainly in language. These models ingest and structure language, building an artificial understanding of our world that allows them to read, analyze, and generate content. This capability alone is transformative.

In my experience working with corporations to improve their business processes, especially in B2B interactions like source-to-pay or order-to-cash, the foundation of these workflows is documents—emails, meeting notes, presentations, quotes, questionnaires, contracts, purchase orders, invoices. These documents form the bedrock of business relationships, and generative AI has the potential to revolutionize how we handle them.

For over 20 years, I’ve worked with companies around the world on process optimization, and I can confidently say generative AI is a game changer. Imagine a single assistant that holds the collective knowledge of every type of business document in every language. Suddenly, reading, analyzing, extracting structured data, assessing risks, or generating polished documents becomes far easier and smarter. The kind of process automation and optimization that was previously difficult or uneconomical to implement is now within reach.

Take the concept further: these AI assistants can not only read and generate documents, but they can also perform actions—fetching data from the web, interacting with APIs, much like a human employee navigating multiple online tools. This is already happening, and the technology is advancing faster than most of us realize. OpenAI’s latest model o1, for instance, is improving its ability to think through multi-step problems before responding, adding even more value for business users. All of this will undeniably boost employee productivity and automate many tasks—freeing people to focus on adding value in new ways and reducing costs in labor-intensive areas.

At Corcentric , for example, my team and I have used the latest AI technologies to automate accounts payable (AP) processes. Corcentric platform receives documents from suppliers, analyzes them, extracts key data, and matches it against company records, all the way through to automatic reconciliation with purchase orders and receipts. The goal is to reduce friction and errors in processing invoices and to improve supplier relationships by enabling timely payments and early issue identification. We’ve applied similar techniques to contracts, purchase orders, and quotes, learning firsthand how generative AI is making previously challenging tasks not only possible but cost-effective. Building real-world applications means combining several foundational models, adding mechanisms to route tasks efficiently, setting up safeguards for compliance, and designing application logic that reflects the business needs. This opens up years of opportunity for improving business processes, boosting productivity, and enhancing quality and insight for companies of all kinds.

While others are racing to build the most advanced AI models, we’re focused on helping companies get better—leveraging this technology as it emerges. Now is the time for corporate leaders to learn about AI’s practical applications to increase efficiency and tackle their existing pain points. Predicting the far future is difficult, but understanding what we can achieve today with the incredible technology already available is the best way to be ready for whatever comes next.s