What is GPT-3? How Can It Be Used in Journalism Practices?
Can GPT-3 technology replace journalism?
GPT-3, known as Generative Pre-Training Transformer 3, is an artificial intelligence technology that OpenAI company, founded by SpaceX and Tesla CEOs Elon Musk and Sam Altman, has been developing for years. This algorithm predicts future texts after a start. After entering a certain initial text, this artificial intelligence uses all the text found on the internet to give a reasonable response based on the input text it receives. Their answers tend to be pretty accurate as well, since they have a lot of data.
This algorithm, which can even do interviews, can be used easily in journalism. The first example of journalism in which this technology, which can produce articles, news, poems, translations, in short, any text with imaginable language algorithm, is featured in The Guardian. On September 8, 2020, “A robot wrote this entire article. Are you scared yet, human?" The text, published with the title of GPT-3 on the author's part, can give an idea for the future of journalism. Since it writes all kinds of texts with a language algorithm, it can write ideas on any subject. At the current stage, GPT-3 can write news in categories such as finance, real estate, weather, police-judicial, and sports. The basic element of all of them is numerical data. The algorithm used analyzes thousands of news and matches what kind of evaluations are made against which data.
The biggest advantage in the field of journalism is to lighten the burden of journalists. GPT-3 technology, which can write news on topics such as weather, sports, and economy, can lighten the burden of journalists, allowing them to do more investigative journalism or spend more time on special news. Journalists will still continue to do basic work such as information control, analysis, information gathering because artificial intelligence cannot replace them. It will save journalists from making boring routine news.
But on the downside, readers will likely want to see a person's contextualization and commentary on even this boring news story. This automation in journalism may not be accepted by most people. These problems can also be seen in news gathering and distribution. At the same time, software costs are so high that most media companies cannot use this technology.