Seligman Investments has doubled the deployable capital of its venture business to $1 billion less than a year after launching it, betting that the physical constraints of the artificial-intelligence buildout will generate a new group of major technology companies.
The expansion of Seligman Ventures reflects growing investor interest in the hardware and infrastructure required to support increasingly compute-intensive AI systems.
The strategy targets a fundamental constraint in the AI boom.
Artificial-intelligence development depends not only on increasingly capable software models but on semiconductors, networking, storage, power infrastructure, cooling and other physical systems capable of supporting enormous computing workloads.
Those bottlenecks have become increasingly important as hyperscalers and AI developers commit substantial capital to expanding data-center capacity.
Seligman’s decision also illustrates a broader change in technology investing.
Many high-growth technology companies are remaining private longer, limiting the opportunities available to public-market investors seeking exposure before businesses reach very large valuations.
Venture investment offers one route into that earlier stage of development, although it carries materially different liquidity and valuation risks than publicly traded securities.
The expansion comes as public investors are becoming more selective about the economics of AI infrastructure. Higher interest rates increase financing costs for capital-intensive projects, making revenue visibility, customer concentration and eventual cash generation increasingly important.
That does not necessarily weaken the infrastructure thesis. It may instead increase competition for companies solving specific bottlenecks that can demonstrate pricing power or strategic importance.
For markets, the significance of Seligman’s move lies less in the size of a single venture fund than in where sophisticated capital is being directed.
AI investment is increasingly spreading outward from foundation models and headline chipmakers toward the physical systems required to deploy artificial intelligence at scale.
The next test will be whether that private-market enthusiasm produces sustainable businesses capable of justifying valuations once they eventually approach public markets.
