Pietari Laurila @triangulacapital: "ᴡᴇᴇᴋʟʏ ᴜᴘᴅᴀᴛᴇ 9 ꜰᴇʙʀᴜᴀʀʏ 2026 Until..."

United Arab Emirates

ᴡᴇᴇᴋʟʏ ᴜᴘᴅᴀᴛᴇ 9 ꜰᴇʙʀᴜᴀʀʏ 2026 Until last week, the risk of AI disruption had been reflected most clearly in the valuations of obvious targets like call centre companies, software firms, advertising agencies and recruiting companies. These were the sectors where the threat felt immediate and easy to understand. However, after new AI tools from Anthropic signalled faster automation of white-collar tasks, investors suddenly began to reassess companies that had previously been considered relatively immune. Information and analytics providers such as Wolters Kluwer and London Stock Exchange Group, as well as digital platforms such as Booking and Rightmove, found themselves in the firing line. The perceived threat from AI is expanding because AI is rapidly moving from answering questions via a chat bot interface to actually doing work. This leap has become possible in part because models such as OpenAI’s ChatGPT 5.2 and Anthropic’s Opus 4.6 have become smarter. But it’s also about the models being connected to tools and plug-ins that allow them to execute multi-step workflows from start to finish. For example, OpenClaw is an open-source “agentic” AI assistant that can carry out tasks rather than just answer questions. It connects to messaging apps, local files, and online tools, and can schedule meetings, send emails, run commands, or automate workflows on behalf of the user. The latest development, Moltbook, is a social network designed for AI agents rather than humans, where bots can post, comment, and interact with each other, using systems like OpenClaw as their “brains.” Meanwhile, RentAHuman.ai is a newly launched marketplace where AI agents can hire real people to carry out tasks in the physical world. Instead of humans using apps to order services, autonomous AI systems become the “clients,” and humans become the execution layer. Although these technologies are in their early stages, they show how quickly AI agents could move from just answering questions to being a connected system, performing tasks and interacting with each other and the real world. With simultaneous advances in humanoid robotics, it is not difficult to imagine the world of work five or ten years from now looking very different from today. Given the fundamentally transformative nature of AI, every investment now needs to be analysed in light of what AI could do to the business. A portfolio full of companies that could be disrupted by AI could face existential risks. A natural reaction to AI-proof a portfolio is to rotate into the “picks and shovels” of AI: chips, memory, power, cooling, grid equipment, and data-centre infrastructure. Unfortunately, many of these obvious winners have already re-rated significantly. The surge in AI data-centre build-out is real, but valuations in parts of the supply chain have been pushed to levels that leave little room for disappointment. In an AI world, where white collar work is done in data centres and blue collar work by humanoid robots, who could benefit? One answer could be experiences and self-improvement. If humans work less and have more free time, they may spend a larger share of their income on travel, leisure, fitness, and wellness. Some also speculate that the best land and locations globally could become even more valuable, as the owners of AI-driven wealth compete for scarce, desirable places to live. In light of last week’s developments, I am increasing the share of Travel and Leisure stocks in the portfolio. Selected Real Estate exposure will be maintained, while the allocation to businesses that appear structurally vulnerable to AI disruption will be kept limited. 𝗣𝗼𝗿𝘁𝗳𝗼𝗹𝗶𝗼 𝗰𝗵𝗮𝗻𝗴𝗲𝘀 Unicredit and DNB were sold. New positions were opened in TUI and Deutsche Boerse. 𝗖𝗼𝗻𝘁𝗮𝗰𝘁 www.triangulacapital.com 𝘛𝘩𝘪𝘴 𝘤𝘰𝘯𝘵𝘦𝘯𝘵 𝘪𝘴 𝘧𝘰𝘳 𝘪𝘯𝘧𝘰𝘳𝘮𝘢𝘵𝘪𝘰𝘯 𝘰𝘯𝘭𝘺. 𝘐𝘵 𝘪𝘴 𝘯𝘰𝘵 𝘢𝘯 𝘰𝘧𝘧𝘦𝘳 𝘰𝘳 𝘳𝘦𝘤𝘰𝘮𝘮𝘦𝘯𝘥𝘢𝘵𝘪𝘰𝘯 𝘵𝘰 𝘣𝘶𝘺, 𝘩𝘰𝘭𝘥 𝘰𝘳 𝘴𝘦𝘭𝘭 𝘢𝘯𝘺 𝘪𝘯𝘷𝘦𝘴𝘵𝘮𝘦𝘯𝘵, 𝘯𝘰𝘳 𝘭𝘦𝘨𝘢𝘭, 𝘵𝘢𝘹, 𝘰𝘳 𝘧𝘪𝘯𝘢𝘯𝘤𝘪𝘢𝘭 𝘢𝘥𝘷𝘪𝘤𝘦. 𝘗𝘢𝘴𝘵 𝘱𝘦𝘳𝘧𝘰𝘳𝘮𝘢𝘯𝘤𝘦 𝘪𝘴 𝘯𝘰𝘵 𝘪𝘯𝘥𝘪𝘤𝘢𝘵𝘪𝘷𝘦 𝘰𝘧 𝘧𝘶𝘵𝘶𝘳𝘦 𝘳𝘦𝘴𝘶𝘭𝘵𝘴.

Not investment advice. The author may have financial interests in the mentioned instruments.

Strictly Necessary Cookies

Performance Cookies

Functional Cookies

Targeting Cookies