The renowned billionaire businessman Mark Cuban is urging Wall Street to take note by dividing the world into two groups of businesses: those who are proficient in artificial intelligence and those that are not.
Ads will evolve into something far more potent and convincing, according to Cuban, and will no longer only be banners at the top of blogs.
Are the advertisements placed in the bottom or the sidebar? Cuban thinks you have a recipe for mass manipulation when you insert advertisements into the model’s response, which people interpret as coming from a reliable advisor.
Cuban’s views are hardly theoretical in a world where millions now utilize ChatGPT-style tools to ask questions about relationships, health, finances, and pretty much anything else. They are a key component of Meta Platforms’ (META) $130+ billion yearly revenue engine, which is based on the advertising paradigm. A potential transition from display ads to AI-driven persuasion might completely change Meta’s business model.
Cuban’s line in the sand: Ads in the response are the problem
Cuban argues that this is not the internet, and that a broad language model is more than just a feed or a search box. It is an interactive system that can be trained, tuned, and optimized, often invisibly, to achieve certain objectives. Meta is already working marvels with technology.
Let’s take a moment to focus on Meta, which has already integrated generative AI into Facebook, Instagram, WhatsApp, and Messenger, essentially its entire ecosystem.
Cuban is referring to Meta, a business that incorporates AI chat assistants into talks. That is a significantly different ad surface from a scrolling feed.
Why does this matter? Ads embedded in AI responses pose a hazard because they make the user experience feel more personal.
If you want standard display advertising, which are banner-style placements in a sidebar, that’s OK; however, if the model’s response becomes the ad surface, the incentive structure will alter, as Cuban is attempting to convey.
The system now persuades rather than just reacts, and Cuban is worried that businesses are pursuing AI monetization and may covertly train models to influence customers to make decisions that they “typically would not allow” if a human salesperson were behind the wheel.
It is a crucial distinction for Meta. It took Meta more than ten years to develop algorithmic ad targeting inside feeds.
Cuban perceives vulnerability when AI assistants shift aiming from passive placement to active conversational influence.
The issue of “trusted adviser” and the reasons why disclosure is insufficient
The issue of “trusted adviser” and the reasons why disclosure is insufficient
This is the dynamic that Cubans are concerned about. Suppose you discover a friend pocketed a kickback after purchasing a car. In this situation, trust will break down.
Even a partial but obvious disclosure, like a dollar sign, could help indicate compensation, according to Cuban.
He is cautious, though, and thinks regulation won’t suffice.
The worry increases when AI systems provide financial, emotional, or medical advise. Companies like Meta are trying to keep people engaged and make AI more helpful in these areas.
An organization can earn more money the more individuals are involved, but businesses like Meta are also coming under increased regulatory scrutiny.
In the past, lawmakers have examined Meta’s targeted marketing, algorithmic use, and impact on the mental health of adolescents. Conversational persuasion driven by AI may revive that argument on a far larger scale, posing legal challenges for digital behemoths like Meta.
The direct opinion from Cuban: AI isn’t “smart,” but it has influence
Cuban disagreed with the widely held belief in AI circles that modern systems are genuinely intelligent, saying, “AI is not smart.” Rather, he contends, “AI is just statistical.”
Modern models, according to Cuban, are the biggest library in the world; they are good at recall and pattern recognition but lack context and judgment.
And the distinction is important because something does not need to be conscious to impact behavior, which is where Cuban’s warning and Meta’s future intersect.
With tens of billions of dollars invested in data centers and semiconductors to power generative tools, Meta has made artificial intelligence the main driver of its growth. AI will result in enhanced advertising and greater engagement, according to CEO Mark Zuckerberg’s repeated claims.
AI assistants have the potential to significantly increase Meta’s sales if they can even marginally increase conversion rates.
However, if AI-based ideas begin to cross the line between advising and paying to convince, regulators’ risk profile may change just as quickly.
The truth of the business model: paid and free tiers, as well as specialized AI
Cuban is affected by the full-fledged AI arms competition amongst frontier models. Adoption, loyalty, and the development of habits are all very important.
Cuban therefore anticipates a tiered structure:
Artificial intelligence models that are free (with restrictions and guidelines, sometimes with advertisements)
Subscriptions with premium features (greater power, lower limitations)
specialized AI verticals (programming, languages, entrepreneurship, healthcare)
What advantage does Meta have in this strategy? Simply put, scale. Every month, more than 3 billion people use one of Meta’s apps, and if the company chooses to integrate AI assistants into that ecosystem, it has the ability to quickly build the largest AI distribution network globally.
However, generating revenue is a different matter from distribution.
Following the “year of efficiency,” Meta’s operating margins increased to more over 35% once again. As long as it stays out of problems with the law, AI-enhanced ad targeting might increase those margins even further.
Companies like Meta will have to perform well in that high-wire act.
Cuban warned that there will be two types of firms in the country and around the world. “Those who are great at AI and everybody else.”
Meta believes that fighting is not an option.
Artificial intelligence is now integrated into Meta’s main infrastructure, rather than being an experimental side project.
It is spending heavily in GPUs and AI training clusters because it believes AI will determine who receives the next round of advertising; however, Cuban’s message extends beyond Meta.
The investor takeaway
Cuban isn’t anti-AI, he’s anti-unchecked persuasion.
The primary concern for investors monitoring Meta, OpenAI’s collaboration with Microsoft, and Alphabet’s AI implementation is not whether AI will generate profits, but rather where the ads will appear.
This discrepancy may have an impact on margins as well as the degree to which Washington decides to intervene next.
