advertisement
advertisement

Elon Musk-funded AI text generator was too dangerous for the public

Elon Musk-funded AI text generator was too dangerous for the public
[Photos: Franck V./Unsplash; Pixabay/Pexels]

AI researchers fed their piece of software the first sentence to George Orwell’s 1984, and it produced the following text:

advertisement

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.

True, that’s not how the story goes, but that sure doesn’t seem like something a robot wrote either. The text was generated by an AI model called GPT2, built by an organization called OpenAI–which is funded by Elon Musk and Reid Hoffman. The software is good–scarily so. And the creators realized the world may not be ready for what they built.

According to the Guardian, the people at OpenAI are not releasing their research to the public for fear that the system could be used maliciously.

GPT2 is essentially a text generator. All it needs is a few lines fed, and it will figure out the tone and subject. It then writes its own version of what should come next. The software was given a few sentences from a news report about Brexit, and it successfully wrote the rest–including fabricated quotes from Jeremy Corbin.

Right now, OpenAI has decided to hold off on releasing the system for now, as the group figures out what bad actors might be able to do with the technology. The researchers are hoping to use this as an example of what tech advances will be available in a few years time. Instead of giving everyone the keys, they want to try and figure it out themselves to be one step ahead.

If we’re this close to an AI text generator that’s so scarily convincing, the future sure looks dark.

You can read the full Guardian report here.

advertisement
advertisement