Mark Cuban says AI's 'biggest challenge' for businesses is consistency

Billionaire entrepreneur Mark Cuban recently raised an intriguing point about the challenges faced by AI technologies. In a post on X, he expressed his belief that inconsistency is currently the most significant hurdle for both enterprise AI solutions and consumer applications. Cuban noted that when users pose the same question multiple times, they often receive different answers. "It's still impossible to guarantee that everyone gets the same answer to the same question, every time," he stated. "AI doesn't understand the implications of its responses." This observation underscores a fundamental distinction between AI systems and traditional software. Historically, enterprise tools have been designed around deterministic rules, ensuring that the same input yields the same output consistently. In contrast, generative AI technologies, such as OpenAI's GPT-5.5, Anthropic's Opus 4.7, and Google's Gemini 3.1, function on probabilistic frameworks. This means they generate responses based on likelihood, allowing for a variety of potential answers rather than adhering to a singular response. As a result, the unpredictable nature of AI can create challenges in business environments, leading to what some refer to as 'AI hallucinations.' However, some critics of Cuban's perspective argue that this variability is a necessary tradeoff, especially for creative tasks where multiple answers can be valid. They contend that enforcing strict consistency may stifle the tools' utility, particularly in scenarios that require creativity or ambiguity. In his discussion, Cuban emphasized the increasing importance of human judgment in working with AI outputs. He remarked, "Judgment and the ability to challenge AI output is becoming increasingly necessary and valuable," suggesting that domain knowledge is more critical than ever. This perspective reflects a broader trend he has previously highlighted regarding the divergent ways individuals are leveraging AI technology. On the Big Technology Podcast, Cuban elaborated on this divide, describing two distinct groups: those who depend on AI as a crutch and those who use it as a tool for learning. He warned that those who treat AI as a "drunk intern" may find themselves at a disadvantage, while those who engage actively with the technology can cultivate valuable skills. Cuban has also provided practical advice for users looking to harness AI effectively, including three initial questions to pose to Anthropic's Claude for business development.