HR edition
Employers want to see that people can use AI in practice. OpenAI and Google are releasing their strongest models carefully. At the same time, training in organisations is lagging behind.
Employers want practical AI skills, not just basic knowledge
According to Workday, AI changes jobs more often than it removes them. Employers are increasingly asking for practical AI skills.
On 5 October, Workday published its Global Workforce Report. Its main finding: AI changes jobs more often than it removes them. Of business leaders, 40% expect AI to give existing employees more work to do. By contrast, 28% expect their workforce to shrink. Demand for skills is changing too. Since its peak in January, demand for basic AI knowledge has fallen by 25%. Demand for practical skills has risen by 51%. These include building AI tools and automating work processes. Two other figures have a direct impact on your work. There are now 69 applicants for every vacancy that is filled. Previously, there were 58. Also, 57% of employers see fewer chances to progress within the company. This matters to anyone responsible for career paths and development. Why is demand for basic knowledge falling while demand for practical skills is rising? One possible explanation is that many people now know what AI is. That no longer makes them stand out. The difference lies with people who can truly redesign a work process. For HR, this means that 'AI-skilled' is too vague for a job description. It is better to describe a specific task. What work will someone do, with which tool, and with what result? You can check this in an interview. A line on a CV is not enough. This also applies to your current employees. Their work may be among the first to change when tasks are reorganised. HR often draws up the training plan too. But be careful about what these figures do not tell us. 'Jobs changing' sounds kinder than jobs being cut. Yet 28% expect a smaller workforce. And 69 candidates per vacancy points to greater pressure when hiring. Getting more from existing employees can also mean that each person has more work. So the discussion is not only about opportunities. It is also about workload and training. Ask your team leaders this week which tasks in their teams are likely to change first. Also discuss what training will be needed.
Background for subscribersThe reports show that employers are increasingly looking for practical AI skills. But use and training do not always keep pace with each other in organisations. So describe which AI task someone must be able to do, and what training they need for it. Also check whether providers explain clearly how their tools work. What data do they use, and who oversees them? Keep a person involved in choices that affect candidates or employees.
OpenAI reveals 'dots': AI agents that are always on
On 29 September, OpenAI presented more than 20 products at DevDay. The most important are the 'dots'. They are agents in ChatGPT that are always on. An agent is an AI that does more than answer questions. It also carries out steps on its own. According to OpenAI, dots can rearrange appointments, book flights and change code. They do this on your behalf. ChatGPT Space is also coming. It is a shared workspace where teams, ChatGPT and dots work together on projects. A Decisions API is coming too. This connection lets companies leave small, frequent choices to AI. Plugins (add-ons) will work as complete apps in ChatGPT. The announcement comes at a time of much debate about control over AI that acts on its own. An agent is given a goal and access to, for example, a calendar or project. It then chooses the steps it needs to take. In a shared workspace such as ChatGPT Space, people and agents work on the same documents. This can make it unclear who did what. With the Decisions API, the company decides what counts as a 'small choice'. That term can be interpreted broadly. Scheduling an appointment is a small choice. Rejecting a candidate is not. The benefits mainly come from routine tasks that are easy to undo, such as changing an appointment. The risks are greater when a choice affects someone's work or income. A person must be involved then. They must also be able to explain why the choice was made. Write down which HR decisions you will never leave to an agent. Do this before a colleague switches one on.
Background for subscribersThis is a taster.
Subscribers read the whole HR edition: every story, each with its background in plain language.
Subscribe →Not ready for that? Read the free letter first →
AI and tech, made simple.
This is the HR edition. Subscribers get the background to every story, in plain language.
Subscribe
inklaretaal · Amsterdam · © 2026