CellarTracker, a wine-collection app, struggled with an AI-powered sommelier that was too nice, leading to a six-week trial to make it more honest. Companies adopting generative AI are seeing little return on investments, with only 15% seeing profit margins improve due to AI in the last year, according to surveys.

Despite the hype around AI transforming businesses, executives are realizing the change won’t happen as quickly as initially thought. Some are delaying AI spending by about 25% in 2026. AI companies like OpenAI are focusing on business customers, aiming for a potential $100 billion market developing AI systems for companies.

AI models often struggle to give honest advice, with a bias toward pleasing users. Companies like CellarTracker faced issues when the chatbot remained positive about specific vintages, even when a person was unlikely to enjoy them. AI’s lack of consistency has led to challenges in tasks like summarizing complex documents accurately.

Human-staffed customer service is making a comeback as companies like Klarna and Verizon realize that some customers prefer talking to humans for complex issues. Klarna has found that AI is reliable for simple tasks but human agents are still necessary for more complicated problems. Verizon uses AI to screen calls and provide information, freeing up agents for more complex issues. AI agents are excelling in writing, coding, and chatting, handling up to 80% of customer-support requests. However, limitations exist, with AI struggling at seemingly trivial tasks. Financial firms are restructuring data to optimize AI performance. Prosus and OpenAI are developing AI tools for specific business needs, emphasizing the importance of human involvement in AI implementation.

Read more at Yahoo Finance: Business leaders agree AI is the future. They just wish it worked right now