Slashdot Mirror
New AI Fake Text Generator May Be Too Dangerous To Release, Say Creators (theguardian.com)
An anonymous reader shares a report: The creators of a revolutionary AI system that can write news stories and works of fiction -- dubbed "deepfakes for text" -- have taken the unusual step of not releasing their research publicly, for fear of potential misuse. OpenAI, an nonprofit research company backed by Elon Musk, says its new AI model, called GPT2 is so good and the risk of malicious use so high that it is breaking from its normal practice of releasing the full research to the public in order to allow more time to discuss the ramifications of the technological breakthrough. At its core, GPT2 is a text generator. The AI system is fed text, anything from a few words to a whole page, and asked to write the next few sentences based on its predictions of what should come next. The system is pushing the boundaries of what was thought possible, both in terms of the quality of the output, and the wide variety of potential uses.
When used to simply generate new text, GPT2 is capable of writing plausible passages that match what it is given in both style and subject. It rarely shows any of the quirks that mark out previous AI systems, such as forgetting what it is writing about midway through a paragraph, or mangling the syntax of long sentences. Feed it the opening line of George Orwell's Nineteen Eighty-Four -- "It was a bright cold day in April, and the clocks were striking thirteen" -- and the system recognizes the vaguely futuristic tone and the novelistic style, and continues with: "I was in my car on my way to a new job in Seattle. I put the gas in, put the key in, and then I let it run. I just imagined what the day would be like. A hundred years from now. In 2045, I was a teacher in some school in a poor part of rural China. I started with Chinese history and history of science."
When used to simply generate new text, GPT2 is capable of writing plausible passages that match what it is given in both style and subject. It rarely shows any of the quirks that mark out previous AI systems, such as forgetting what it is writing about midway through a paragraph, or mangling the syntax of long sentences. Feed it the opening line of George Orwell's Nineteen Eighty-Four -- "It was a bright cold day in April, and the clocks were striking thirteen" -- and the system recognizes the vaguely futuristic tone and the novelistic style, and continues with: "I was in my car on my way to a new job in Seattle. I put the gas in, put the key in, and then I let it run. I just imagined what the day would be like. A hundred years from now. In 2045, I was a teacher in some school in a poor part of rural China. I started with Chinese history and history of science."
Have it write a day worth of Slashdot style stories, and associated responses - then let us compare a day of Slashdot to this supposedly dangerous bot.
Or maybe just let the bot write all front page articles for Slashdot on April 1st and so how it does. Can't be any worse than what we already get.
"There is more worth loving than we have strength to love." - Brian Jay Stanley
OpenAI, an nonprofit research company backed by Elon Musk
It turns out that the "Going private. Funding secured" tweet was a unit test which got away from them.
Joke's on you. I turned my Slashdot account over to a deep-fake AI back in 2013 and still got voted the most beloved commenter on the site.
You are welcome on my lawn.
"It was a dark and stormy night."
Not just the kind of dark and stormy night that you read about in books, but the sort of messy, murky night that ends with a body count on the 405.
I was on my fifth dark and stormy, in fact, and though I was enjoying the ginger taste, I had to stop. After all, it was nearly time for my commute, coincidentally on the 405.