The AI Hype Will Burst: It’s Our Responsibility to Find a Thoughtful Replacement | Mark Surman

It was December 1999, and the world of tech investments was at a fever pitch. The prevailing belief was simple: a flashy website paired with a Super Bowl advertisement could lead to instant wealth. In this frenzy, spending was mistaken for growth, and the vigor of marketing was seen as a viable business strategy. However, this prosperous façade was destined to crumble; within mere months, the dot-com bubble would burst, erasing $1.7 trillion in market value and inflicting a $5 trillion blow to the economy at large.

Yet, amidst this devastation, an extraordinary transformation began to take shape. The post-dot-com landscape wasn’t just mired in a sea of failed endeavors but instead saw the emergence of creativity and growth: welcome to the age of web 2.0, the proliferation of open-source software, and the creation of groundbreaking platforms like Firefox and Wikipedia. The key takeaway from this tumultuous period is clear: when bubbles burst, they pave the way for new opportunities—if we take the chance to reconstruct wisely.

Fast forward to today, and we find ourselves entrenched in yet another boom – this time centered on artificial intelligence (AI).

The current AI surge mirrors the dot-com euphoria with alarming resemblance. Reports indicate that nearly 80% of stock gains projected for 2025 are concentrated within a mere seven tech giants: Alphabet, Amazon, Apple, Meta, Microsoft, Nvidia, and Tesla. These companies are locked in a battle for control over the complete AI ecosystem, which encompasses hardware, software, data, energy, and infrastructure. This struggle transcends mere market share; it fundamentally determines how billions of people will learn, create, and navigate the world.

Such a concentration of power should raise alarms across industries.

Much like in the heyday of the dot-com bubble, valuations are skyrocketing without any evident route towards profitability. Companies are perpetuating the narrative that AI will inevitably supplant human jobs, despite the fact that 95% of AI experiments in organizations stumble before reaching production. Instead of prioritizing tools that enhance public welfare, the industry seems to be creating what Cory Doctorow describes as “productive residue”, inundating the market with synthetic media, misinformation, and deepfakes.

The crux of the issue lies not in AI itself, but rather in the economic framework guiding its development.

This situation is not unavoidable. It results from an economic model that approaches technology as an extractive venture—accumulating data, consolidating authority, and externalizing negative consequences. The ongoing race in AI isn’t fueled by innovative progress, but by a quest for control that prizes profit over human welfare.

An Alternative Economic Model Already Exists

On the positive side, an alternative economic model is already taking shape. Across the globe, open-source developers and mission-oriented companies are forging shared infrastructures for trustworthy AI—characterized by transparency, auditability, and local adaptability. They are demonstrating that innovation does not require monopolistic data control.

This reality is exemplified in the pioneering companies led by founders who prioritize ethical values alongside competitiveness. Organizations such as Hugging Face, which operates the world’s leading open-source machine-learning model hub; Flower AI, which promotes decentralized, federated learning to counter the predominance of centralized models; and Oumi, which offers an open-source platform for developing and deploying AI models on local infrastructure instead of closed cloud systems, are among those making strides in this direction.

These ventures are not shoots of speculation; they are building blocks of a more sustainable and inclusive technological ecosystem. This initiative aligns with what we envision as a double-bottom-line economic model for technology—an approach that values both mission and revenue.

Slop is Not Destiny

If history offers any insight, the current hype will conclude similarly to the dot-com bubble—inevitably culminating in a crash. However, this does not signify the end; rather, it signifies the dawn of a fresh narrative.

During the last bubble, the Linux stack emerged from the chaos, with open-source fundamentals that now support a significant portion of the internet, overpowering proprietary models like Windows. Over the past two decades, open-source components alone have generated $8.8 trillion in value, with research suggesting that tens of billions more could be realized if businesses pivoted from closed AI platforms to open-source alternatives.

The potential for value creation today is staggering.

When the AI bubble inevitably bursts, we will face an important decision. We can either rebuild the existing monopolistic framework, or we have the opportunity to reshape it into a model that empowers humanity and is centered on values. This means adopting open frameworks, ensuring transparent governance, and enabling equitable participation in the value generated by AI.

Moreover, we must prioritize what people truly desire from technology: their privacy, security, agency, and the joy of using it. The true promise of AI isn’t in reaching infinite scales but in simplifying our lives, enriching our experiences, and fostering creativity without compromising choice and dignity.

Signs of this shift are already evident. We are seeing advancements in privacy-focused, open-source technologies, such as for browser and email assistants, which are continually improving.

Now, envision a world where individuals and communities host small, localized AI models—optimized for energy efficiency and privacy, and finely tuned to their unique needs. Envision developers collaborating on tools instead of competing against one another, and a framework of innovation measured not by market share but by the public good.

This is not an unattainable dream. If we begin now to build AI grounded in openness, transparency, and shared principles, we can usher in an era of technology that enhances rather than restricts human freedom. The dot-com crash ultimately led to the development of the modern web. Similarly, the next economic correction could lead to a better, more inclusive technological landscape—provided we choose to rethink the economics of innovation.

Ultimately, the future is in our hands. We can choose to allow a select few companies to dictate our shared destiny, or we can collectively take ownership of the innovations we create.