Artificial Intelligence has impacted every industry in the world. If we look at media, companies have deployed different AI and machine learning techniques to automatically produce news stories at scale. Here, AI/ML can be used to grow an audience, aggregate build loyalty, have better data insights, readership engagement.
Let’s look at a few examples. There is Bloomberg’s Cyborg which automatically extracts key data points from earning reports for thousands of companies. There is Yle News Lab at the Finnish Public Broadcasting Company with their smart news assistant Voitto for its personalised news. Wall Street Journal uses an ML-based dynamic paywall for personalised subscription prices based on reading habits. Reuters developed News Tracer and Lynx Insight. Both tools use machine learning and artificial intelligence technologies to support Reuters journalists in the newsgathering process.
Sign up for your weekly dose of what's up in emerging technology.
While AI is serving great value for news media houses, we know that AI is very complex technology. It encompasses various techniques that can be leveraged to build models and this is where the challenge presents to media professionals. So, should journalists learn the techniques? According to Google, the answer is yes. Google recently introduced a course on machine learning as part of Google News Initiative in collaboration with JournalismAI and VRT News.
Google has been working alongside media organisations, and journalism think tanks to strategise what AI tools can benefit reporters in doing their job. The tech giant has launched free training courses in machine learning and artificial intelligence to help them in their workflow.
From time to time, Google has provided journalists with better data tools. For example, working with Stanford University’s Big Local News and Pitch Interactive, Google News Initiative built the COVID-19 Case Mapper to make it possible for local journalists to easily embed up-to-date coronavirus map visualisations on their sites for readers.
Here Is Why Journalists Should Learn AI/ML
A journalist can definitely leverage AI in delegating redundant and repetitive tasks to machines using a variety of techniques. They can find in-depth analysis and investigations, which can then be taken forward for better engagement.
Also, we know that in coming years automation will have an impact on how journalists work. Learning basic AI/ML tools becomes an important consideration given redundant reporting processes are being automated. Not just to prepare themselves against future job automation, AI/ML holds tremendous value in things like fact-checking and deriving insights from various data resources.
The democratisation of AI/ML and algorithms will deeply impact journalism. Whether it is text, video, audio or others, information can be numerically represented for getting deep insights. So, a journalist needs to be AI literate to leverage all these Innovative tools to provide better content and information to readers.
Google says it is therefore imperative for journalists to embrace technology, gain some expertise to use different AI products. The latest machine learning course became a part of Google’s free training tools for media organisations. This according to Google will promote the efforts of journalists who are disseminating critical information to their readers.
“It’s a hard time for journalists and news organisations globally, as they try to evaluate the influence that COVID-19 will have on the market and editorial side of the industry. This course complements our freshly launched collaborative experiment, and also our effort to highlight profiles and experiments that reveal the transformative capability of AI and machine learning in shaping the journalist, and the journalism, of the future,” wrote Mattia Peretti Manager at JournalismAI, a leading organisation working as part of Google News initiative.
What’s In The Course?
The machine learning course for journalists focuses on helping journalists understanding the fundamentals behind machine learning and how journalists can actually train a machine learning model.
The course covers machine learning topics such as Supervised learning, Unsupervised learning, and Reinforcement learning. It teaches how journalists can assess the use case for machine learning, acquire data, getting data in shape, choosing an algorithm (such as Google Cloud AutoML Natural Language) for building a model.
It also specifies how journalists can use different tools to improve their model, validate and test the model, evaluate results. Finally, the course moves forward to understanding bias in machine learning. It covers the types of bias, sources of bias and preventing bias in models so journalists don’t report incorrectly.