Fly On Wall Street

You’re Hired! This Site Generates Random Neural Network Résumés

A new website can write your résumé for you in just ten seconds — as long as you don’t mind sending employers a document of totally-made-up information and just a touch of gibberish.

The site, called “This résumé does not exist,” uses a neural network-based artificial intelligence system trained on information scraped from public job boards to create brand new résumés for made-up job applicants.

Training Data

There’s just one catch: the people on the résumés may not be so made-up after all. Neural networks trained on real-world examples churn out results that are as close as possible to those original examples. In this case, it seems like the algorithm found a shortcut.

Futurism reviewed at least five résumés generated by the website, each of which provided an email address for their fictional subject that turned out to be registered in real life.

Futurism also found that the résumés provided the links to real Facebook and Instagram accounts, suggesting that the résumés generator might be simply reproducing real people’s names and contact information. By press time, none of those people have responded to Futurism’s request for a comment, so the résumé with an active Facebook link has been excluded from this article.

The headshots provided on some résumés, however, are randomized and generated by a similar website that uses AI to generate fake photos, according to an FAQ section on the résumé site.

Shallow Pool

Aside from reproducing contact information, the AI-generated résumés don’t always paint applicants in the best light. The job descriptions highlight AI’s ongoing struggle to recognize context and the actual meaning of words.

For example, a résumé for a fictional person named Darren Charles boasts that “…I am activities and learn new things and I do noticity.” Another résumé’s summary reads “I am a passionate software developer who loves to preparo many talent” and says that the person spends over a quarter of their time “getting too little sleep.”

If this algorithm’s HR skills are any indication, it looks like the job search and hiring process will still need a human touch.

Exit mobile version