01/06/2026
📰 In the latest interview for My Company Polska, CEO of COGITA AI and our member Adam Gabriel Dobrakowski shares practical insights from supporting analytics and hashtag , to why many -generated applications collapse the moment they hit production scale.
As Adam Gabriel Dobrakowski highlights in the interview:
💬 “AI agents are excellent at detecting narrow bugs and vulnerabilities, but they still struggle with deep architectural redesign. This stems from the nature of these models themselves. (...) Moreover, if we use the same AI model to search for errors that originally generated the code, there is a high probability it will not notice them at all – because it will reinforce its own flawed reasoning.”
💬 At the same time, he notes that AI is dramatically lowering the barrier to :
“What once required months of development work and budgets of €200-300K to build an MVP can now be achieved by a technical team in just a few days with the support of AI tools, significantly reducing the initial entry threshold for startups and founders. Two years ago companies were still debating copyright issues around AI-generated code – which, by the way, remains unresolved in some ways. Today nobody questions it anymore: if a senior developer doesn’t use AI, they will simply fall out of the market.”
Big thanks to Wiktor Cyrny for a substantive and well-conducted conversation.
🔎 Read the full interview here: https://lnkd.in/dFFv6sgn