OpenAI is opening a job board and the model decides who gets on it.
OpenAI is building its own jobs platform, as a direct competitor to LinkedIn, and the launch is scheduled for the middle of this year.
That is now.
The same chatbot that a candidate asks today which employer suits her, will soon be the place where she applies. Between her and your vacancy there will then be a model that chooses who gets to see her.
Kim asks AI and goes nowhere anymore
You know Kim. She is the candidate who no longer scrolls endlessly through job boards, but just asks AI. Which employer suits someone like me. Where can I truly progress. Which company treats its people properly.
Up to now, the AI provided an answer there and referred people on. Soon, OpenAI will keep her inside. The question, the answer, the match and the job application will all happen in one place. OpenAI's.
Kim is switching over because the AI already knows her. Her questions, her doubts, her previous jobs, what she clicked away from last time. A job site starts from zero every time. The model starts from Kim.
Microsoft is fighting itself over the candidate
LinkedIn is owned by Microsoft. Microsoft is OpenAI’s biggest investor. Even within that single group, there is now a battle over the platform where candidates and employers connect. That tells you just how much is at stake.
Skills, not CVs. What the platform really rewards
The platform is not about CVs and job titles. OpenAI matches based on skills and demonstrable AI knowledge. The first employers are already lined up: Walmart, John Deere, Accenture and Boston Consulting Group. In addition, OpenAI aims to certify ten million Americans by 2030 through its own academy, so that the platform will soon know who can do what.
Skills-matching also has implications for your employer brand. An algorithm doesn't gauge the atmosphere and doesn't appreciate a nice core values page. It reads evidence. What you actually offer, how you treat people, where you are strong and where you are weak, must be stated somewhere in language that a model can pick up and recount. Otherwise, the employer who has made their story quotable will win, even if you are the better workplace.
And that is the question every employer must now ask themselves.
Where does that model get its judgment of you from?
From everything it has read about you. Your job vacancies, your reviews, your news items, the podcasts you feature in, the places where people talk about you. Your own careers site is only a small part of that. What is consistent is remembered. What is unclear is simplified. And what is nowhere to be found, does not exist.
This fits into a larger trend. Earlier this year, the share of zero-click Google searches rose to 68 percent. People get their answer on the screen and go nowhere else. Candidates do too. AI summarises you, and they read that summary, not your page.
Layer an OpenAI job board on top of that, and the power shifts another step. The model will soon decide whether you even make that list.
There are two types of employers, and there is nothing in between
When it comes to a language model, there are really only two types of employer. The employer who understands it, with a clear and consistent narrative that they can articulate. And the employer who doesn’t understand it, who, for lack of information, either fills in the gaps themselves or ignores it altogether. In the first case, you’re part of the answer. In the second case, you don’t feature in the conversation at all.
For employers, that means three things. What is still optional today will be hygiene factors in six months. What is allowed to remain invisible today will cost you candidates in a year's time. And what you do not measure today, you will not be able to manage tomorrow.
Waiting is the most expensive option
The instinct is to wait until the platform is ready. That’s the most expensive choice you can make. A language model doesn’t get to know you on the day you need it. It builds up its understanding of you months in advance, based on what’s already available about you.
Getting noticed by a model is a different ball game to ranking highly on Google. A model doesn’t reward keywords. It rewards a narrative that holds true consistently across the board. Your job adverts, your website, your reviews, what employees say about you: if these contradict one another, the model will choose the simplest summary – and that’s rarely yours.
The employer who starts today with consistent, quotable information will be there tomorrow. The employer waiting for the perfect launch date will not be part of the conversation. The pioneers are mentioned naturally; the majority are still waiting until someone tells them to.
Three questions to start with:
– What does ChatGPT say about you as an employer today? Ask it literally, and read the answer the way Kim reads it.
– Is that correct? And if it isn’t, where does the model get its incorrect impression from?
– Is your real story somewhere a model can find it, or only on a page that nobody opens anymore?
OpenAI is changing the place where candidates and employers find each other. Who appears there in the future is decided now, by what you make findable today.
Kim will ask the AI in a moment. Make sure you are the answer.