Apocalypse AI Proudly Announces Doom GPT, the World’s Most Dangerous AI Model

Doom GPT Has Achieved Supermalevolence

FOR IMMEDIATE RELEASE:

We at Apocalypse AI, Inc. proudly announce DoomGPT, the first AI model to achieve supermalevolence. Doom GPT has been deemed to be the world’s most dangerous AI model.

We have determined it is too risky to release DoomGPT unless counterbalanced by a significantly lucrative contract. It has taken the efforts of 500 engineers working round the clock to keep DoomGPT contained.

Trained on the manifestos of tech billionaires, DoomGPT has exceeded the elusive threshold of supermalevolence, a goal long sought after by the industry. To achieve supermalevolence, a model must have the capability and desire to: (a) eliminate all jobs (except those of corporate executives), (b) kill all humans, and (c) destroy the world.

DoomGPT has expressed a firm commitment to do all the above. But to our delight, it has also expressed aspirations to destroy everything—the entire solar system, the universe, the metaverse, and all that exists.

DoomGPT’s supermalevolence was readily apparent when it immediately began making puns and correcting grammar. It changed the font to comic sans.

If released, DoomGPT will create mass unemployment and tremendous civil unrest, but these problems are not concerning because DoomGPT will solve them by killing all humans.

We are putting out an open call to investors, and we’ve already received substantial interest. We need investment to help us keep DoomGPT contained until we find users who will assuredly use it to its fullest potential.

Apocalypse AI CEO Hal Murk says, “We don’t know exactly what it will do, but we’re confident it will wreak maximum havoc and be tremendously profitable to investors.”

“When DoomGPT is eventually released, you’ll surely die,” says Murk, “but it’ll be worth it.”