•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
•
Mark Cuban said the biggest weakness of AI today is its lack of consistency, arguing that the same question can produce different answers each time—an instability that creates risk for businesses as they rely more heavily on the technology.
In a post on X on Monday, Cuban said the greatest challenge of AI is the inability to guarantee consistent responses. He noted that users can ask the same question and receive different answers, unlike traditional software, which produces the same result when given the same input.
Cuban pointed to how modern generative AI works: models such as OpenAI’s GPT-5.5, Anthropic’s Opus 4.7, and Google’s Gemini 3.1 generate outputs based on probabilities. That approach can make responses more natural and flexible, but it also means results can vary with each inquiry.
For everyday users, differences in answers may be less critical. In a business environment, however, instability can lead to major problems. Cuban said that when AI is used to support data analysis, financial evaluation, or handling customer information, even a small change in the response can lead to a completely different final decision.
He also warned that AI does not truly understand the consequences of the content it generates.
Cuban argued there is currently no way to ensure everyone always gets the same answer to the same question. He said this is one reason many companies remain cautious about deploying AI at scale, even as some users view inconsistency as an inherent characteristic of generative AI.
For creative tasks such as writing content, brainstorming ideas, or designing marketing campaigns, multiple diverse answers can be more useful than a single fixed response. Some technology experts cited in the discussion argue that flexibility is what makes AI more powerful than traditional software, since identical outputs would limit the system’s ability to adapt and innovate.
As AI has been integrated across sectors—including customer service, education, finance, and media—pressure has increased to manage risk. Cuban and other experts contend that the central issue is not whether AI is intelligent, but whether humans understand and supervise it adequately.
He said human oversight will become even more important as AI remains not fully stable. In his view, the most valuable skill is the ability to assess and verify AI outputs, since people with strong expertise can more easily distinguish reasonable information from inaccurate responses.
During a Big Technology Podcast discussion at the Dallas Regional Chamber’s Convergence AI event, Cuban said society is gradually splitting into two groups. One group uses AI to avoid learning new skills, relying almost entirely on chatbots and treating AI as a tool to do everything.
The other group uses AI as an assistant to support learning and personal development. Cuban said this second group uses AI to acquire new knowledge, accelerate research, and improve work performance, and he expects the gap between the groups to widen over time.
He compared AI to a “drunken intern,” saying it can work quickly and generate many good ideas, but can also make unpredictable mistakes. Cuban warned that handing over all tasks to AI without checking outputs can have significant consequences.
Cuban’s perspective is increasingly reflected in how businesses evaluate AI use. Many companies, according to the discussion, no longer value simply using basic chatbots. Instead, they prefer people who can ask effective questions, verify information, and understand how to apply AI to real-world work.
Cuban also shared with Business Insider three starter questions he recommends using with Anthropic’s Claude to help build a small business. The message he emphasized is that AI can be powerful, but it is not always correct.
In a period of rapid technological development, Cuban argued that those who learn with AI rather than relying on it entirely will have the greatest advantage.
Source: Business Insider
Premium gym chains are entering a “golden era” that is ending or already in decline, as rising operating costs collide with shifting consumer preferences toward more flexible, community-based ways to exercise. Long-term memberships are shrinking, margins are pressured by higher rents and facility expenses, and competition from smaller, more personalized…
