Revolutionary AI fake text generator is 'too dangerous' to release: Project backed by Elon Musk won't publish its research for fear of potential misuse

  • Open AI is a research group from Elon Musk and US entrepreneur Sam Altman 
  • It has created a model that can generate stories from little more than headlines
  • Using AI language models that can already translate, read and write
  • The new programme won't be released yet due to fear of misuse and fake news

By Yuan Ren For Mailonline

Published: 12:03 EST, 14 February 2019 | Updated: 06:54 EST, 15 February 2019

Elon Musk's AI research group Open AI has come up with a 'large-scale unsupervised language model'  that can generate news stories from little more than a headline. 

But the group won't not be releasing details of the programme which it says could be misused in the wrong hands.  

The technology will be rapidly advanced in the coming years and its details released once the negative implications have been 'discussed' by researchers. 

Scroll down for video 

Elon Musk's AI research group Open AI announced in a paper yesterday that it has generated 'a large-scale unsupervised language model' that can write news stories from little more than a headline. It has also provided samples of the type of articles that can be generated on its blog

Elon Musk's AI research group Open AI announced in a paper yesterday that it has generated 'a large-scale unsupervised language model' that can write news stories from little more than a headline. It has also provided samples of the type of articles that can be generated on its blog

OpenAI is a group founded by Musk and backed by Silicon Valley heavyweights like LinkedIn's Reid Hoffman. 

It aims to develop increasingly powerful artificial intelligence tools in a safer way.

The new model, known as GPT-2, has been developed using technology that already allows computers to write short news reports using press releases. 

The programme is able to write coherent articles as shown by a sample of its writing, which produced a nine paragraph piece based on a two line manual insert about scientists discovering unicorns. 

So-called language models that let computers read and write are usually 'trained' for specific tasks such as translating languages, answering questions or summarising text. 

Researchers have however found the model to read and write longer passages more easily than thought, and with little human intervention.   

These general purpose language models could write longer blocks of information by sourcing from text openly available on the internet.  

It will take a few years however until the model can be reliably used and the process will require costly cloud computing, although such this could come down rapidly.

Sam Bowman, an assistant professor at New York University who has reviewed the research, said: 'We're within a couple of years of this being something that an enthusiastic hobbyist could do at home reasonably easily,'  

'It's already something that a well-funded hobbyist with an advanced degree could put together with a lot of work.'

The group will however not be announcing details of the programme which it says could be misused in the wrong hands. The technology will be rapidly advanced in the coming years and its details released in after the AI research community have discussed the negative consequences. 

The group will however not be announcing details of the programme which it says could be misused in the wrong hands. The technology will be rapidly advanced in the coming years and its details released in after the AI research community have discussed the negative consequences. 

While OpenAI is describing its work in the paper, it is not releasing the full model out of concern it could be misused. 

The researchers said: 'Due to our concerns about malicious applications of the technology, we are not releasing the trained model. As an experiment in responsible disclosure, we are instead releasing a much smaller model for researchers to experiment with'.  

Dario Amodei, OpenAI's research director, said: 'We're not at a stage yet where we're saying, this is a danger,' said 'We're trying to make people aware of these issues and start a conversation.' 

The team are hoping to rapidly advance the technology in the next few years and release it in a safe way and as a reliable version to the public.  

Alec Radford, one of the paper's co-authors, said: 'It seems like there is a likely scenario where there would be steady progress'.

'We should be having the discussion around, if this does continue to improve, what are the things we should consider? ' 

AN EXAMPLE OF OPENAI'S GPT-2 SYNTHETIC TEXT OUTPUT


SYSTEM PROMPT (HUMAN-WRITTEN)

In a shocking finding, scientist discovered a herd of unicorns living in a remote, previously unexplored valley, in the Andes Mountains. Even more surprising to the researchers was the fact that the unicorns spoke perfect English.

MODEL COMPLETION (MACHINE-WRITTEN, FIRST TRY)

The scientist named the population, after their distinctive horn, Ovid’s Unicorn. These four-horned, silver-white unicorns were previously unknown to science.

Now, after almost two centuries, the mystery of what sparked this odd phenomenon is finally solved.

Dr. Jorge Pérez, an evolutionary biologist from the University of La Paz, and several companions, were exploring the Andes Mountains when they found a small valley, with no other animals or humans. Pérez noticed that the valley had what appeared to be a natural fountain, surrounded by two peaks of rock and silver snow.

Pérez and the others then ventured further into the valley. “By the time we reached the top of one peak, the water looked blue, with some crystals on top,” said Pérez.

Pérez and his friends were astonished to see the unicorn herd. These creatures could be seen from the air without having to move too much to see them – they were so close they could touch their horns.

While examining these bizarre creatures the scientists discovered that the creatures also spoke some fairly regular English. Pérez stated, “We can see, for example, that they have a common ‘language,’ something like a dialect or dialectic.”

Dr. Pérez believes that the unicorns may have originated in Argentina, where the animals were believed to be descendants of a lost race of people who lived there before the arrival of humans in those parts of South America.

While their origins are still unclear, some believe that perhaps the creatures were created when a human and a unicorn met each other in a time before human civilization. According to Pérez, “In South America, such incidents seem to be quite common.”

However, Pérez also pointed out that it is likely that the only way of knowing for sure if unicorns are indeed the descendants of a lost alien race is through DNA. “But they seem to be able to communicate in English quite well, which I believe is a sign of evolution, or at least a change in social organization,” said the scientist.

 

Advertisement

Elon Musk AI creates news generator that's 'too dangerous' to release!

Comments 1

Share what you think

The comments below have not been moderated.

The views expressed in the contents above are those of our users and do not necessarily reflect the views of MailOnline.

Add your comment

What's This?

By posting your comment you agree to our house rules.