I spent five years of my life working on this problem, and even got involved with the standards-making process. First, there already is a standard for resumes and has been for 20 years. It's called HR Open Standards Candidate (it is the successor to the old HR-XML standard, which in turn was based on OASYS). There are also five or six other attempts at making a standard CV / Resume too. The problem with all of them is that every time one gets adoption, another standard, usually created because the author was unaware or did not like the existing standard for some use case. In some cases the standard is de-facto, like the XML format Indeed uses, and in others it is a formal spec (like the HR Open Standards JSON family).<p>The reason this problem is hard is that a resume is an intersection of multiple kinds of data: education, licenses, contact information, work history, and narrative text. Business can't even agree on how to best organize this data, or even how to represent it. Individuals? Well, the CV is all about me, and I want mine to uniquely be me -- even in that means using comic sans. Also, people don't look for jobs every day, so many make a new resume and a new profile whenever they start searching. So you have consumers of data (businesses) and creators of data (job seekers) who really are misaligned.<p>Finally, business want perfect candidates, and people are not perfect. There's actually an incentive in job hunting for people to stretch the truth, which makes CV data unreliable regardless of format. In some cases, the automated screening is so tight that only a lie will get through the filter.
Honestly I'm deeply against "standard CV", the only who benefit form them are giants who trade humans as livestock. If I choose to propose myself for a position I want another human on the other side reading who I declare to be and my motivations, digging as needed because choosing someone for a job MUST BE a serious human business. For similar reasons I'm deeply against those who mass-send CVs to anyone: when one need desperately a job, any kind of, and try fishing the best option MUST not be mass sending but simple State made public matching platform where any one <i>on both sides</i> for free can consult the other side mark some offers/people as interested to book a contact etc.<p>The actual "job market" is abomination that must be annihilated and companies who can't have enough HR resources for their scale simply must be resized by nature because being unable to humanly process human resources means a bad/too quick/too unbalanced dangerous growth. Similarly for companies who fail to have call centers etc (yes, Alphabet is one of them).<p>We are not robots, we are human who work to live, not the contrary, and we are a society not a factory.
This looks like this project will benefit machines more by screening candidates out automatically before they've reached the human stage.<p>So effectively candidates will be rejected before they have even applied.
This is an interesting concept but as we have seen when the semantic web started out it only really served to enhance search engines instead or large tech companies would internalize the semantics of the data. Making large formats for interoperability never really caught on.<p>Naming the standard Mac is also an issue (due to a product by Apple Inc.) but since its an acronym you should always format it as MAC
In my mind, "Use a standard, machine-readable CV to represent your skills and life's work" is practically saying, "Commodify yourself!"
The page says the structure is based on an internal format that was used for storing employee data. I can see value in a standard record format from that perspective, eg if employee bio and work history are being kept on file.<p>I agree with the other comments that this should not exist for job applications, and that asking for a job application in machine readable format is basically saying "we're not going to read this". I would never use a channel that asked me to apply this way.
This will only benefit recruiting firms in order to automate processes such as the one behind this project, I don't see how this would be beneficial for anyone else.
If you can’t easily and intuitively understand why you shouldn’t call it “Mac” here in 2022, I confess to zero confidence that you’ll have successfully achieved any other objectives. Just being frank. The “Go” team ignored sense and it seems to have worked out somehow for them, so it may be moot to lose my random individual interest so quickly.