Luke Harries ranks AI‑resistant roles

Luke Harries said on May 24 that some of finance’s most relationship-driven jobs remain among the least exposed to artificial intelligence, according to a post on X. Harries grouped “Capital allocators (VCs, PE, public market investors)” with “Deal making (M&A + sales)” in a ranking of roles he sees as highly AI-resistant. The post circulated across finance-focused accounts on May 24, adding to a broader online debate over which white-collar jobs AI can automate fastest and which still depend on judgment, trust and access. The post is published on X under ID 2058399861934854531. ### Which jobs did Harries single out? Luke Harries named two finance-heavy categories in the May 24 post: “Capital allocators (VCs, PE, public market investors)” and “Deal making (M&A + sales).” Those labels put investors and transaction-focused roles in the same bucket of jobs he views as relatively insulated from AI disruption. The wording matters because Harries did not point to a specific firm, strategy or software tool. (x.com) He framed the ranking around role types, with capital allocation spanning venture capital, private equity and public markets, and deal making spanning mergers and acquisitions work alongside sales. ### Why would capital allocators be harder to automate? (x.com) Capital allocators sit at the point where money, judgment and access meet. Venture capitalists, private equity investors and public-market investors all use data, but they also depend on manager selection, founder assessment, sourcing networks and conviction under uncertainty. Harries’ post did not spell out that reasoning, but his grouping suggests he sees the core task as more than pattern recognition. (x.com) In practice, allocators often compete on proprietary access as much as analysis. A firm can use AI to screen markets, summarize filings or model scenarios, but winning an allocation, backing a founder early or building trust with a management team still depends on human relationships and reputation. That is an inference from the roles Harries highlighted, not a direct quote from his post. (x.com) ### Why did deal making land in the same category? Deal making, as Harries described it, includes both M&A and sales. Those are jobs where persuasion, timing and negotiation often determine the outcome. A model can help prepare materials, surface comparables or draft outreach, but it does not by itself replace the person who gets a founder to sign, a board to agree or a buyer to improve terms. M&A advisers and salespeople also work in settings where information is incomplete and incentives are shifting. (x.com) That makes the job partly analytical and partly interpersonal. Harries’ inclusion of sales next to M&A suggests he was ranking roles by the share of work that still depends on live interaction rather than repeatable process. That is an inference based on the categories in the post. ### Was this a formal report or a social-media take? The May 24 ranking appeared as a social post on X, not as a published research note or employer study. Harries presented the view in short-form format, and the discussion spread through finance accounts rather than through a company filing or academic paper. That format matters because the post reads as an argument about labor durability, not as a measured forecast with disclosed methodology. (x.com) Readers looking for the exact wording can find it on X under post ID 2058399861934854531. ### Where does the debate go from here? The next step is likely to play out on X and across finance circles as users test Harries’ ranking against specific tasks inside investing, banking and sales. (x.com) The post published on May 24 remains the primary source for the claim, and the full wording is attached to X post ID 2058399861934854531.