An email from NPR this week announced that NPR is “actively developing a framework and set of principles to guide decision-making on all aspects of AI investment and use.” bottom. The email said NPR will consult with a wide range of subject matter experts, including editorial, legal, security and data governance, to evaluate how AI is used across his NPR and the NPR network. increase.
Also recently, the Radio Television Digital News Association (RTDNA) published guidelines on the use of AI in journalism. According to RTDNA, these guidelines are among the first guidelines on AI from a national journalism organization. RTDNA guidelines are more of a warning label than a set of standards or policy statements. A section of the RTDNA guidelines labeled “Accuracy, Context and Clarity” states that newsrooms should consider the following questions:
- Can you fully understand the functionality and source material of the AI program prior to implementation?
- What are the precautions to protect against inadvertent plagiarism?
- Can I independently verify the accuracy of my AI tools?
- Will there be an opportunity to test the AI tools before going public?
- What are your newsroom systems and expectations for pre-publication human review?
AI has been used to produce journalistic work for some time. It has been using AI since 2014 to generate data-driven automated text stories covering topics like financial reports and sports results, according to an Associated Press report. The Associated Press says AI has boosted the output of corporate earnings stories for all publicly traded US companies by a factor of 10.
But today’s AI is not yesterday’s AI. Current AI, called generative AI, can explore large amounts of data at high speed and synthesize the information contained in that data into narrative sentences that closely mimic human natural language. Leveraging sophisticated language models, this new feature enables AI chatbots to generate relatively well-balanced articles on topics based on a set of parameters programmed into their code.
The impact of generative AI (and what comes after) on journalism is far-reaching. On the bright side, AI can now be used by journalists to mine previously unmanageable data to develop deeper, factual stories. On the dark side, tech companies (many of which are now media companies, I think Google and Apple are) compensate journalists who created original sources as a way to maximize profits. Instead, they will use AI to generate content cheaper and faster. Raw materials for AI stories.
Recent article by Maggie Harrison on the website futurism Discusses the dangers of AI to the fragile journalism economy. The imaginative title of this article says it all. “Google announces plans to use AI to disrupt the journalism industry”. Harrison discusses Google’s recently announced demo of its new AI-powered search interface. In this interface, every Google search is preceded by an AI-generated overview called an “AI snapshot.” Harrison believes this AI snapshot of his could be the real journalism, because research shows that information consumers rarely make it to his second page of search results, or to the end of his first page. They argue that it will be a big change that will keep journalism out and make journalism even more destabilizing. source of income to support it. Harrison writes: “…this demo raises a very important question for the future of an already dilapidated journalism industry: How can Google’s AI concoct an original work and then distill it at scale without ever connecting it to users? Are you going to offer it?” For original work, how will publishers continue to monetize their work?”
That’s a good question that deserves an answer. Our democracy may depend heavily on it.
