AI “bottleneck trading” or is coming to an end; early SpaceX investors say the market focus will shift to long-term value

BlockBeats News, June 17 — The hot trading activity surrounding artificial intelligence supply chain shortages may be nearing its end, as some investors shift from searching for the next shortage link to screening companies that still have long-term competitiveness after the AI infrastructure cycle.

Altreides Management Managing Partner and early SpaceX investor Gavin Baker said in an interview with TBPN that over the past year, the market has been keen on chasing bottleneck assets in AI development, including DRAM, storage chips, and key material suppliers. But he believes that these "AI bottleneck trades" are approaching their conclusion.

Baker pointed out that Japan’s Ajinomoto refused to raise the price of a key chip packaging insulator material, indicating that the pricing power of some supply chain bottlenecks may have already begun to ease. This material is used in the packaging layer connecting processors and chips, which previously attracted investor attention due to surging AI chip demand.

He said the previous market game was to find the "next bottleneck," but the more important question at the next stage is which companies can maintain a lasting franchise value after these bottlenecks recede.

This view has cooled the recent surge in storage and AI material stocks. Storage stocks like Micron and SanDisk have risen sharply this year, driven by AI data center capital expenditures, HBM demand, and long-term procurement agreements that have led to a supply-demand re-evaluation. But as stock prices rapidly climb, the market has begun to discuss whether the trades are becoming overly crowded.

Baker also stated that the next main focus of AI infrastructure might shift toward the actual deployment of computing power and electricity. He mentioned SpaceX’s potential in ground and orbital AI data centers and said the market will pay attention to how quickly it can increase "gigawatts" of computing capacity.

This means AI trading may shift from simply betting on shortages to a broader infrastructure competition: whoever can acquire land, electricity, GPUs, and data center capacity faster may hold a more advantageous position in the next AI investment cycle.