Sri Lanka, June 29 -- AI reasoning models were supposed to be the industry's next leap, promising smarter systems able to tackle more complex problems. Now, a string of research is calling that into question. In June, a team of Apple researchers released a white paper titled "The Illusion of Thinking," which found that once problems get complex enough, AI reasoning models stop working. Even more concerning, the models aren't "generalizable," meaning they might be just memorizing patterns instead of coming up with genuinely new solutions. Researchers at Salesforce, Anthropic and other AI labs have also raised red flags. The constraints on reasoning could have major implications for the AI trade, businesses spending billions on AI, and even t...