In a recent Town Hall, OpenAI CEO Sam Altman sat down with a room full of builders to pull back the curtain on GPT-5, the "Jevons Paradox" of coding, and why human attention is becoming the world's most valuable commodity. For value investors, the takeaway was clear: we are moving from a world of software scarcity to a world of radical abundance, and the traditional moats of "execution" are being drained in real-time.

Investors often worry that if AI can write code, we’ll need fewer engineers. Altman argues the exact opposite using the Jevons Paradox, the idea that as a resource becomes more efficient to produce, we actually consume more of it. He envisions a future where software isn't a static product you buy, but a fluid service constantly being rewritten and customized for a "market of one."

For the retail investor, this suggests that the "Total Addressable Market" (TAM) for software is essentially infinite. We are moving toward "micro-apps" and hyper-personalized tools that solve problems too small for a traditional SaaS company to bother with today.

If building a product becomes trivial, where does the value go? Altman, drawing on his days running Y Combinator, notes that the "hard part" has shifted entirely. The technical barrier to entry is collapsing, meaning a "great product" is no longer a competitive advantage, it’s the baseline. The real bottleneck is now Go-To-Market (GTM) and human attention.

Altman dropped a bombshell regarding the cost of intelligence: he expects a 100x reduction in cost for GPT-5-level reasoning by the end of 2027. This is the "deflationary" force of AI. In an economy where the cost of cognitive labor drops by 99%, the business models that rely on "charging for seats" or "billable hours" are at extreme risk. Value will instead flow to those who can orchestrate these cheap "agents" to solve complex, multi-step problems autonomously.

As the cost of execution plummets, the "Quality of Ideas" becomes the primary differentiator. Altman teased the development of tools, essentially a "Paul Graham Bot", designed to help humans brainstorm and iterate on ideas. He believes that models capable of scientific discovery will soon be able to suggest high-value product insights that humans might overlook.

Perhaps the most insightful observation for investors was Altman’s take on human-centric branding. He noted that while AI (affectionately dubbed "clankers" by some) can produce incredible art and text, humans lose interest the moment they realize there is no soul behind it.

For investors, this means the "Human Premium" is real. Companies that lean into community, storytelling, and human connection will likely maintain pricing power, while "faceless" utility companies will be commoditized by the falling cost of AI.

Altman didn't shy away from the risks, specifically in biocurity. He admitted that "blocking" access to dangerous info won't work forever. Instead, the world must move toward a "resilience" model, similar to how we built fire codes and flame-resistant materials rather than banning fire. This signals a massive emerging sector in AI-driven security and resilience infrastructure.

The "so what" for the value investor is this: Do not invest in companies whose only value is "we use AI to do [X] faster." That advantage will be gone by the next model update. Instead, look for:

The era of "building software" is being replaced by the era of "directing intelligence."

For more of my insights on this topic, be sure to follow me.