Before many people think about investing, they think about whether they will still have a job. The worry is immediate because AI is already strongest at work that is structured, repetitive and easy to turn into a clear process. An OpenAI study found that around 80% of US workers could see at least 10% of their tasks affected by AI tools such as ChatGPT and Claude. Around 19% could see at least half of them affected.
The first rule follows from that. Risk rises quickly in any role built around routine drafting, admin, formatting, coordination or other repeatable digital tasks. In separate research, customer support agents using AI became 15% more productive on average, with the biggest gains among less experienced and lower-skilled workers. When AI brings average workers closer to stronger ones, routine work loses value. Companies need fewer people for it, because the same work gets done faster and more consistently.
Task type, whatever the title, is the dividing line, which is why trades look safer, though not untouchable. A working paper from the Czech National Bank argues that jobs done mainly on a computer will feel AI’s effects earlier. Jobs built on physical labour or social interaction will feel them later. So the earlier pressure falls on office-based service roles.
Inside the office, safer ground lies in work built around communication, ownership, judgement and decisions that do not follow a script. The support study suggests experience still holds value in that work, where context, responsibility and accountability carry more weight than speed.
Closeness to the business is the second rule. People closest to clients, revenue, delivery and decision-making are harder to remove than those further from outcomes. The person producing routine internal material is easier to replace than the person carrying the client relationship.
The third rule is trust, along with the network and reputation behind it. As output gets cheaper, trusted relationships gain value by comparison. People still work with people they know, trust and believe will stand behind the result when something goes wrong. The same rules apply to firms. Those built around repeatable service work face more pressure on pricing, staffing and margins than those built around trust, relationships and direct ownership of outcomes.
