Recent discussions about OpenAI’s financial runway and its exploration of an advertising model have sparked debate about a potential AI bubble. However, insights from key figures suggest a different narrative focused on tangible value and long-term strategic evolution.
The predicted trajectory for 2026 centers on the maturation of AI agents, particularly multi-agent systems. These systems are expected to handle complex, end-to-end tasks in both enterprise and consumer contexts. For businesses, this means automating entire workflows like ERP management. For individuals, it could mean seamlessly planning a trip by coordinating flights, restaurants, and schedules. A significant “capability gap” currently exists, where most users only scratch the surface of an AI’s potential. The goal is to transition AI from a conversational partner to an active task executor, bridging this gap over what is seen as a decade-long adoption journey.
The healthcare sector exemplifies both AI’s high potential and its regulatory hurdles. Millions already use AI for health inquiries, and many doctors employ it as a diagnostic augmentation tool. The value for consumers lies in regaining health autonomy—researching symptoms or analyzing a restaurant menu for dietary restrictions. The core challenge is not capability but navigating legal and institutional frameworks that restrict AI from performing certain tasks, like prescribing medication, despite its potential proficiency.
Financially, the argument against an AI bubble is built on metrics like API call volumes, not stock prices, indicating real, growing demand. Internally, AI tools are already creating measurable efficiency, such as automating contract review for finance teams, which boosts productivity and morale. The business model is described as a multi-layered “cube.” The foundation is a flexible, multi-cloud infrastructure. The product layer expands from ChatGPT to specialized versions for health, work, and creative tools like Sora. The top layer diversifies revenue through subscriptions, enterprise SaaS, high-value scenario pricing, and future models like licensing fees from drug discovery breakthroughs. This complex combination aims to generate the funds needed for massive compute investments, which are planned years ahead to meet anticipated demand.
Regarding the new ad model, the stated principles are user-centric: always prioritize the best answer, ensure ads provide inherent utility (like a helpful travel ad), and maintain an ad-free subscription tier. Data privacy, especially for sensitive areas like health, is emphasized as non-negotiable.
For the enterprise market, success is claimed to be already underway, fueled by a “consumer flywheel” effect where personal use drives workplace demand. The strategy involves deep vertical industry solutions, moving from light customization to transforming core business operations. Despite this, significant opportunities remain for startups, particularly in areas requiring deep domain expertise, access to proprietary data behind firewalls, or management of complex, unique workflows.
Looking further ahead, a profound economic shift is anticipated. As AI and robotics drastically reduce the cost of labor and expertise, a deflationary environment could emerge. This raises critical questions about social adaptation, with suggestions that substantially enhanced government safety nets and access to near-free basic services, including AI-powered education and healthcare, may become necessary. The transformation extends beyond technology to the very structure of economy and society.