This Fortune article outlines the risk of using general purpose LLMs for CX. Leading AI startup Cursor CX agent hallucinated a key policy causing confusion amongst its users. General purpose LLMs lack determinism, and are prone to consequential errors. AI agents, especially for transactional applications like CX, require different LLMs, designed with deterministic controls that guarantee policy following and eliminate hallucinations around critical workflows. Until very recently, LLMs with deterministic controls for complex and consequential applications like CX didn’t exist. Now they do. Sharon Goldman
NEW: AI startup Anysphere has been riding high over the past two months, thanks to the skyrocketing popularity of its AI-powered software coding assistant, Cursor. The company, which was eyed for acquisition by OpenAI and has reportedly been in funding talks for a valuation of nearly $10 billion, hit $100 million in annual revenue since Cursor launched in 2023. But this week, Cursor went viral for all the wrong reasons: Its customer support AI went rogue, triggering a wave of cancellations and serving as a cautionary tale for other startups betting big on automation. Read the full story: https://lnkd.in/eqQujxYn
