Every generation has its trendy baby names.
If you were born in 1925, chances are good that your parents named you something like Betty, Joan, Billy, or Gene. If you were born in 2005, you’re a lot more likely to be an Addison or an Aiden. And in a few decades? You might be naming your kid Nitnis. (That is, if we aren’t all named Zenon.)
As Co.Design reports, developer and designer Nate Parrott created a neural network with the sole purpose of coming up with new names. He trained it on 7500 American names, translating each name into a mathematical representation that could be tweaked by an algorithm to create new names.
He used the algorithm to generate a random sample of new names from this list of 7500 names, coming up with some plausible new ideas for soon-to-be parents. Mannie? Rusert? Halden? Not all that crazy. However, the algorithm did come up with some ideas that might not catch on soon, like “P” or “Suttttuuyy.”
Do you see your future progeny’s name in here?
Dive deeper into the process and the code in Parrott’s blog post.