‘…after careful monitoring, OpenAI had not yet found any attempts of malicious use but had seen multiple beneficial applications, including in code autocompletion, grammar help, and developing question-answering systems for medical assistance.’

GPT-2, a text-generating AI model, created headlines when it was first announced in February this year. Its creators released only 8% of its code at the time, for fear it would be put to malicious use and spawn reams of fake news.

They are now making 50% available.

In between, they released a 25% version. The AI model was developed by OpenAI, a California start-up dedicated to AGI, or artificial general intelligence, a future form of AI that would replicate or replace human performance with multiple abilities.


OpenAI has released the largest version yet of its fake-news-spewing AI MIT TECHNOLOGY REVIEW | August 29, 2019 | by Karen Hao


Release Strategies and the Social Impacts of Language Models paper by OpenAI researchers