OpenAI, a research center supported by Microsoft, Elon Musk, and Amazon has created a ‘state-of-the-art’ automatic text generator. But, you are unlikely to see it in action, anytime soon. The researchers are afraid of its potential misuses. Hence, in a disappointing turn of events, the technology is likely to remain under wraps in the near future.
However, OpenAI researchers have slipped out a few details of the advanced text generator. According to them, the text generator can write product reviews, news stories, and various other content pieces. The writing is apparently a notch above others in terms of its realistic appearances.
According to the OpenAI center, the technology achieves ‘state-of-the-art’ performance on many language benchmarks. It is capable of summarization as well as translation according to OpenAI. Furthermore, the program known as GPT-2 completely outperforms AI programs well trained in Wikipedia entries, news writing and other content forms.
During testing, scientists asked the program to write a story on a herd of unicorns living in a remote valley in the Andes Mountains.. The program generated a 300-word news story about it. Open AI researchers say that people can misuse this creative piece in the near future.
Growing concerns about the misuse of technology
Through a blog post on Thursday, the researchers made it clear that the program would not be open for public at large. The researchers expressed fears that the program could be used to spread fake news on the internet. The advances in AI help it generate fake news articles, automate fake content on social media, and impersonate others.
Scientists at OpenAI believe that public at large needs to be more skeptical of information online. Otherwise the advancements could lead to ‘deepfakes’ phenomenon in the near future.