To avoid having their core inference services commoditized, frontier labs like OpenAI and Anthropic will inevitably move up the stack. They will build applications that compete directly with their largest customers, such as those in legal tech or design, posing an existential risk for any startup building on their platform.

The primary threat for companies dependent on frontier AI models isn't the expense. It's the scenario where providers like OpenAI decide their compute is more valuable for training AGI and abruptly cut off customer access, crippling dependent businesses overnight.

The most powerful AIs may never be released publicly due to their dangerous capabilities. As they are used internally, they pose significant risks that current transparency laws, which focus on public models, do not cover.

The assumption that startups can build on frontier model APIs is temporary. Emad Mostaque predicts that once models are sufficiently capable, labs like OpenAI will cease API access and use their superior internal models to outcompete businesses in every sector, fulfilling their AGI mission.

The decision to restrict powerful but dangerous AI models like Claude Mythos to a select group of large corporations for safety reasons risks creating a massive centralization of power. This gives these entities an insurmountable technological advantage over smaller players and the public.

By restricting its most powerful model, Mythos, to a consortium of large companies, Anthropic is creating a two-tier economy. Smaller companies are left without access to the same advanced offensive and defensive AI capabilities, ending the previously democratic access to cutting-edge models and creating a significant competitive disadvantage.

Contrary to the idea of AI for all, the most powerful models will likely be restricted to a few high-paying clients to prevent distillation and maximize revenue. This creates a future where competitive advantage is defined by exclusive AI access, potentially allowing large incumbents to crush smaller competitors.

Venture capitalists are hesitant to fund new AI labs ('Neolabs'), even those with superstar talent. The primary concern is that any meaningful breakthrough can be quickly replicated by frontier labs like OpenAI, which possess the scale and distribution to capture the value, leaving the startup with acquisition as its only viable exit.

While AI alignment gets attention, the risk of AI concentrating immense power in the hands of a few actors (corporations or states) is arguably more neglected. This could enable unprecedented surveillance or create a single company with the economic power of a nation, posing a distinct and severe threat.