OpenAI, the AI research group backed by Elon Musk, has decided to hold off on releasing their latest discovery.

The AI, dubbed GPT-2, is basically a language system that tries to generate relevant-sounding text from any prompt. When fed a sentence, it uses statistical methods to try to guess what next words are most likely to be.

The scary thing is that it produces weirdly coherent paragraphs.

Developers are concerned that, in the wrong hands, the AI could be used to generate fake news and online hate speech, reports The Financial Times.

Elon is supporting the decision to delay the release, probably because he can relate. Given the right circumstances, he himself transforms into a random text generator on Twitter.

Futurism, with more on GPT-2:

The algorithm, GPT-2, was trained on some 8 million web pages, according to the new research. Given a prompt, GPT-2 is tasked with predicting the next word based how those words have been used on the websites it read. In the end, the algorithm churns out passages of text that are far more coherent than past attempts to build AI with contextual knowledge of language.

In the blog, the OpenAI researchers concede that GPT-2 works only about half the time. But the examples that the team showcased on the blog post were so well-written that you’d be hard pressed to say whether it was written by a human.

The following extracts are examples of what GPT-2 has generated:

There are occasional glitches and incoherent sentences, but for the most part, the above is almost indistinguishable from something written by a human being.

It’s fascinating as a project, but terrifying as a prospect.

I’m glad they’ve decided to keep it under wraps for now.

[source:financialtimes&futurism]