Per Woofun AI, Altreides Management Managing Partner and early SpaceX investor Gavin Baker indicated that the speculative cycle surrounding artificial intelligence supply chain bottlenecks is approaching its conclusion. In an interview with TBPN, Baker observed that market enthusiasm has previously centered on chasing scarcity in DRAM, storage chips, and critical material suppliers.

However, he argues that the current phase is transitioning toward identifying companies with enduring competitive advantages once these temporary constraints dissipate.

Baker highlighted Japan's Ajinomoto as a key indicator, noting the company's refusal to increase prices for a vital chip packaging insulating material. This material, essential for connecting processors within the packaging layer, had previously seen demand surge alongside AI chip adoption. The stabilization of pricing power suggests that the leverage held by certain bottleneck suppliers is weakening. Consequently, this perspective has tempered recent volatility in storage and AI material equities, including Micron and SanDisk, which had experienced sharp gains driven by data center capital expenditures and high-bandwidth memory (HBM) demand. As valuations rise, concerns regarding overcrowded trades have intensified.

Looking forward, Baker suggested that the next frontier for AI infrastructure lies in the practical deployment of computing power and energy solutions. He pointed to SpaceX's potential role in establishing ground and orbital AI data centers, emphasizing that market attention will pivot toward the speed at which entities can scale gigawatt-level power supplies.

This shift implies a broader competition for land, energy, GPUs, and data center capacity, marking a move from shortage-based speculation to comprehensive infrastructure capability.