As far as I know, John Patrick Lowry has never set foot in Valve's recording booth and had "Sniper" recreate old Vine clips. But thanks to free online machine learning tools, everyone from the shooter in TF2 to the narrator in The Stanley Parable has been made to say whatever they want.

15.ai, discovered by Kotaku over the weekend, is a deep-learning speech synthesis tool that trains a library of dozens of character voice clips. From the "deep moji" system for assigning emotions to the way the algorithm seeks out internet slang, it's all very fascinating to read about: type in a line or two, choose a character, and the system, I'm not sure, will generate clips like TF2's Heavy as the former US president in "Home Alone 2."

Obviously, the system is not very good at this.

Obviously, some characters work a bit better than others; GLaDOS is a natural fit, and her voice sounds like a computer-generated voice by design. However, despite the many hours of voice banked to the narrator of The Stanley Parable, the algorithm is unable to capture Kevan Blyting's whimsical and droll intonation.

That hasn't stopped people from going crazy online, though, as a quick search for "15.ai" on Twitter will quickly reveal. After all, a little robot voice acting can't stop the cast of "Team Fortress 2" from taking over an episode of "Thomas the Tank Engine."

Or how about the same group of mercenaries arguing over a round of "Among Us"? [The TF2 gang is obviously popular among those who do such edits, and the Source Filmmaker scene is not necessarily going to go on a rampage by shoving AI-generated dialogue into their vids. But the tool's character registration is not limited to video games. Sponge Bob's nasally voice works shockingly well, and the undersea pineapple dweller has shocking things to say about the city of Baltimore.

Not all characters are this lucky. Gordon Freeman and Chell don't talk no matter how hard they try, and the cast of Undertale and Celeste never make more than a frantic bleeping sound. However, the sole creator of the tool plans to add voices for Sonic, Mass Effect, Fallout, The Witcher, and others, and is currently accepting submissions for any favorite voices you would like to add.