Alex Davis, founder of Dallas growth investor Disruptive, is raising his firm's first real fund targeted at $5–10 billion after 15 years of deal-by-deal SPVs, arguing that the structures he built his business on have become a liability as the best private companies stop accepting SPV capital. On the Sources Podcast, Davis laid out a concentrated "n of one" approach of roughly 35 companies and about 10 new bets a year, a decade-long Palantir accumulation followed by a pre-IPO double-down that came when Palantir was valued around $10–15 billion and is now a roughly $500 billion company, and his unexpected second act running Groq as executive chairman after Nvidia's non-exclusive licensing deal took most of the chip team. He predicts speculative data center landlords without operating expertise "will get hurt," says there aren't too many neoclouds — just not enough good ones — and calls defense tech the most underappreciated opportunity in the market. The article covers his views on US open-weight models, why he thinks compute will not get cheap, the fraud he says the SPV boom has enabled, and the tension between his open-weights thesis and his plan to sell inference capacity to the closed labs.