Jeff Bezos's new AI startup, Project Prometheus, is focused on engineering and manufacturing for computers, aerospace, and automobiles. This is a strategic move to create vertically integrated AI for industries where he has massive existing investments (AWS, Blue Origin, Rivian), signaling a focus on physical-world applications over competing in the crowded foundation model space.
Related Insights
Jeff Bezos's post-Amazon focus isn't on space colonization but on offshoring Earth's polluting industries, like manufacturing and data centers. This "garden and garage" concept treats space as a utility to preserve Earth's environment, not just a frontier for human exploration.
When evaluating AI startups, don't just consider the current product landscape. Instead, visualize the future state of giants like OpenAI as multi-trillion dollar companies. Their "sphere of influence" will be vast. The best opportunities are "second-order" companies operating in niches these giants are unlikely to touch.
Instead of selling software to traditional industries, a more defensible approach is to build vertically integrated companies. This involves acquiring or starting a business in a non-sexy industry (e.g., a law firm, hospital) and rebuilding its entire operational stack with AI at its core, something a pure software vendor cannot do.
Current M&A activity related to AI isn't targeting AI model creators. Instead, capital is flowing into consolidating the 'picks and shovels' of the AI ecosystem. This includes derivative plays like data centers, semiconductors, software, and even power suppliers, which are seen as more tangible long-term assets.
Jeff Bezos's co-CEO role at AI startup Prometheus mirrors Steve Jobs's NeXT venture. This strategy allows a high-profile founder to build relevant new expertise and prove current capabilities, creating a smoother path to potentially return to the helm of their original empire via acquisition or board appointment.
The true economic revolution from AI won't come from legacy companies using it as an "add-on." Instead, it will emerge over the next 20 years from new startups whose entire organizational structure and business model are built from the ground up around AI.
SpaceX is reportedly targeting a $1.5 trillion IPO to raise $30 billion. This capital isn't just for rockets but to fund a new AI infrastructure business: data centers in space. This represents a significant strategic shift, leveraging its launch dominance to compete in the AI compute market by acquiring massive quantities of GPUs.
The true commercial impact of AI will likely come from small, specialized "micro models" solving boring, high-volume business tasks. While highly valuable, these models are cheap to run and cannot economically justify the current massive capital expenditure on AGI-focused data centers.
Permira's AI strategy uses a clear framework: invest in the 'picks and shovels' of compute (data centers) and in applications with unique, proprietary data sets. They deliberately avoid the hyper-competitive model layer, viewing it as a scale game best left to venture capital and strategic giants.
Investing in startups directly adjacent to OpenAI is risky, as they will inevitably build those features. A smarter strategy is backing "second-order effect" companies applying AI to niche, unsexy industries that are outside the core focus of top AI researchers.