Artificial Intelligence overcomes our human limitations in sharing knowledge. And that changes everything.

This year, 2025, has been incredibly intense whilst leading Digital Projects, GOL Labs, and GOL’s innovation team, particularly on Artificial Intelligence project fronts. I had to learn in practice that there’s a fundamental difference between developing innovation for processes and for products. And, unsurprisingly to anyone, here lies the challenge: we’re not perfect communication machines! In process innovation, there’s heavy dependence on knowledge transfer amongst human colleagues scattered throughout the company. Unlike humans, AI doesn’t suffer from the cognitive limitations that affect our capacity to articulate and absorb knowledge. Whilst product innovation involves technological knowledge that can be developed autonomously and asynchronously, process innovation is more associated with complex knowledge requiring understanding of the company’s unique system and is systemic, connected to other innovations or processes. And that’s when 2025 made it clear to us: AI overcomes human limitations in knowledge transfer and, when implemented at scale, becomes a strategic ally of human collaboration (let’s always reinforce this: collaboration!) to unlock process innovation.

AI Requires Redesigned Processes to Deliver Results

Because of this specific nature, process innovation depends more on exchanges between different company units than product innovation does. That’s where the real problem lies: even having the ability to transfer knowledge, without the necessary motivation, people retain valuable insights, whether due to lack of incentives or fear of losing power or status. They’re genuinely afraid of being discarded. Although companies can offer tangible and intangible rewards to influence and incentivise this “motivation”, the effectiveness of these rewards is limited because individuals don’t always want the same incentives companies offer.

This is the point where AI technologies overcome human limitations for knowledge transfer, unlocking more and greater process innovation. More specifically, the scale at which companies implement AI has a positive impact on the probability of introducing process innovations. This relationship is stronger when there’s less opportunity for interpersonal knowledge transfer, especially when companies have larger numbers of employees, struggle to offer synchronous face-to-face training for everyone, cultivate a “silo” culture with excessive verticalisation, don’t disseminate an effective collaboration culture amongst teams, and have high staff turnover.

I took the opportunity to research some studies and articles (sources at the end of the text), and empirical data from 2,268 Belgian companies confirm: companies implementing AI at greater scale have higher probability of developing process innovations, and this effect is stronger in companies that don’t offer face-to-face training, don’t have a collaboration culture, or have high turnover. Belgians aside, studies with German companies reinforce: there’s evidence and facts of positive and significant associations between AI use and productivity, both for sales-based measures and for added value.

And a warning to the impatient amongst us: German data shows that productivity gain isn’t instant magic. There’s a natural delay. The study points out that the real leap happens when the company understands that AI doesn’t run by itself on top of old processes. AI requires redesigned processes to deliver results. Those expecting miracles the following month will be disappointed.

Knowledge Barriers, Not Just Motivation

With data in hand, we have a new challenge to common sense: it’s not about AI generating pretty images or replacing human work simplistically with its thousands of agents. AI was and is revolutionary as an ally in knowledge transmission, overcoming limitations that previously seemed insurmountable. Whilst conventional wisdom almost exclusively blames motivational factors for the difficulty of this transfer, empirical studies suggest that knowledge-related barriers are the most important blockers.

Obviously, AI doesn’t replace human collaboration; it amplifies it. In environments with low collaboration culture amongst teams and geographically dispersed teams, AI acts as a bridge, transferring complex and systemic knowledge through data-based processes, where the capacity to access and assimilate information depends purely on the scale at which AI is implemented. AI systems increase the company’s capacity to identify patterns in large volumes of data. Using structured data, machine learning algorithms recognise complex relationships between tasks and outcomes. Analyses that would take hours, days, and months for a little human—AI can extract this real knowledge and distribute it systematically throughout the entire corporation. In the same way, for everyone, systemically.

Scale Matters More Than Isolated Experiments

In summary, looking at the entire company, AI implementation strategy must be comprehensive and global, not based on isolated agents. AI benefits for process innovation are more powerful when technologies are implemented this way throughout the corporation. If your company faces constraints in knowledge transfer due to limited collaboration amongst teams, it’s precisely in this scenario that AI can help you most. In this direction and thinking strategically for 2026, our innovation policies must shift focus from experimental and restricted AI adoption to supporting scalable and multifunctional AI systems. The good news is that transparency in these studies’ results leaves no room for doubt: AI technology adoption has a positive and significant impact on companies’ productivity. Both the use and the intensity with which companies explore AI’s potential significantly increase knowledge transfer, collaboration, sales, and added value.

2025 proved that AI isn’t about technological fad. It goes beyond that. It’s not merely a new Product. It’s a paradigm shift in processes, collaboration, and learning. It’s about recognising, based on facts, that AI represents a fundamental change in how knowledge flows within organisations, overcoming barriers of our extremely limited human capacity to share what we know!