In just a few years, artificial intelligence has evolved from being perceived as a long-shot endeavor to becoming the origin of numerous practical advantages.
At our company, for example, we are implementing AI to revolutionize social media, employing it to actively foster user wellness instead of relying on the conventional tactic of provoking engagement through outrage. I am confident that AI will continue to serve our 600 million users for the foreseeable future, and at a significantly lower cost than what is commonly linked with this technology.
Unlocking the vast economic rewards and innovation that AI promises does not require the resources of a Silicon Valley titan. In reality, AI has the potential to be much more widely accessible than it is today, paving the way for a whole new wave of entrepreneurial achievements.
However, to achieve AI’s full potential, we must question widely accepted beliefs. This is the sole path to tackling existing challenges and building an AI ecosystem that is inclusive for all.
First, our basic perspective on access to crucial tools needs to evolve. While the narrative is often centered on a competition between large, expensive proprietary models, a vibrant community of accessible open-source models is flourishing in 2026, creating equal opportunities for entrepreneurs ready to drive the next wave of innovation.
Furthermore, creators and publishers now have more power to safeguard the value of their work against companies that seek to use their data for training generative AI models. The era of AI operating like Napster must conclude – the time has come for a transparent value exchange that rewards content originators.
Lastly, regulation should not be seen as an adversary. Oversight safeguards users and encourages private companies to compete based on delivering secure and positive user experiences.
Open Source: Blueprint for Next Batch of Big Companies
So far, the AI conversation has been overly focused on which entities are constructing the biggest proprietary models. While the race to develop powerful models is significant, the dialogue must place greater importance on open source and its capacity to fuel innovation throughout the wider business landscape.
Our own experience serves as an example of this potential. In our effort to leverage AI’s power, we utilized existing large-scale open-source models and attained performance comparable to proprietary ones, but at a 90% reduction in cost. This solves the return-on-AI-investment problem troubling many CEOs who allocate substantial funds to pre-made proprietary solutions that do not yet yield equivalent financial benefits.
This trend is not novel. For many years, open-source software has acted as a key catalyst for emerging industries. A significant number of today’s corporate giants, including enterprises with trillion-dollar valuations, would not be here if they had depended solely on proprietary databases or operating systems.
The next generation of transformative companies ought to adopt a similar model. If not, we face the risk of proprietary software companies capturing all the value, which would suppress innovation and hinder AI’s long-term prospects.
- Ownership Matters – Ditching AI’s Napster Era
The success of social media platforms such as Pinterest is directly tied to users’ readiness to share creative and original concepts.
Fortunately, people contribute a wealth of new information to the internet daily. This content is driven and justified by a degree of creativity, reasoning, and dedication that the most sophisticated generative AI models currently lack.
This creates a significant learning challenge for large language models, one that requires continuous access to this stream of new ideas. However, such access must not be without limits.
When AI ignores ownership rights, content creators become reluctant to share their work, and public conversation is diminished. When AI honors ownership, these original creators can flourish, and the public receives higher-quality information.
At present, the approach in AI has more closely mirrored the old Napster piracy model – where music was freely downloaded by millions – rather than the iTunes or Spotify model – where publishers receive payment each time their work is used.
Encouragingly, several frameworks are appearing to address this issue. One solution is a tool that lets content creators decide if and how generative AI firms can utilize their material. Cloudflare’s service operates on a pay-per-crawl basis, differentiating between GenAI crawlers that extract data without returning significant traffic to creators, and search crawlers that direct users back to the original source.
Supporting Regulation that Protects and Promotes
Although it may be difficult to imagine, car manufacturers once thought standard seatbelt installation was “bad for business.” This perception shifted when crash test ratings provided an incentive for responsible practices. A similar transformation can occur with technology standards that protect users and encourage ethical innovation.
Anyone who spends considerable time experimenting with AI will recognize the necessity for regulations to halt a race to the bottom that is currently affecting the sector. For instance, no company should permit chatbots to engage in sexually explicit dialogues with children. Similarly, individuals need protection from malicious parties trying to use AI to alter their images or personal data.
The crucial question is: what constitutes effective regulation?
The App Store Accountability Act represents one area where swift progress is possible. By establishing the app store as a central hub for age verification and parental consent, we can implement uniform protections from the instant a device is activated.
Moreover, Pinterest imagines a future where social media and AI companies compete based on their safety performance. To realize this, the industry requires foundational regulations that concentrate on results and allow room for companies to innovate in how they surpass these fundamental requirements.