It may be difficult to see now. Since the start of the hosts of Openai’s Changpt and many other AI-powered chatbots and virtual assistants in late 2022, the focus has been around how these tools take over the work of journalists and other content creators. The media industry is already struggling and feels like it’s being attacked correctly.
Even from the inside. Not long after, owners of Politico and Insider Mathias Döpfner told his staff that AI could replace them earlier this year. Then, BuzzFeed’s entire newsroom was released, and CEO Jonah Peretti said the company will focus on AI. Experiment through AI to keep the newsroom list growing. Meta and Openai specifically attract journalists to train LLMS.
With the adoption of AI, humans have laid off employees. The reporter must have reason to worry. That said, media executives have adopted technology and cut humanity too quickly, seemingly after many dangerous events.
CNET and its sister company Bankrate were asked to publish dozens of articles and write them by AI; they have since stopped AI publications. Similarly, G/O Media (owners of sites like Jezebel and Gizmodo) publishes AI-generated stories without editing inputs, thus containing multiple errors. Microsoft users were shocked by inappropriate AI-generated polls, and a story about a woman who died was posted next to it.
All in all, AI is unlikely to replace journalists. Instead, AI may help news publications and make them more dominant. Why? The answer lies in the most critical product of AI labs: high-quality training content.
DéjàVu: How social media reshapes news
Just like the internet reshapes media business – some companies are working too high for shiny new toys, while others are benefiting greatly from the measured new advertising pathways and open distribution approaches, and AI will benefit.
Initially, media publishers were excited about the prospect of a rise in social media. They are no longer bound by physical barriers to printing. It turns out that they suddenly competed with the world, which included not only all other publications, but also personal bloggers and influencers. The New York Times has become a digital media royal family, attracting more than 11 million paid subscribers and has become one of the world’s largest news publishers. Many other publications struggle or have to close.
But it is possible that AI can reshape the entire field by bringing power back to the news media. Large language models require a lot of training content, and the quality of this content varies. It turns out that AI companies have had a great impact on the information captured by news organizations. This is because, unlike your X/Twitter feed and social media, these publications provide high-quality censorship information curated not only by one content creator, but by the entire newsroom of journalists and editors. Therefore, this information will be marked as more reliable and surfaced more frequently. This marks the valuable media company and work of their human employees’ products.
So, how does the New York Times view dealing with artificial intelligence? Well, they are suing Openai. And a large list of media businesses including The Guardian, Kant Naster, Forbes, etc., which prevented AI Crawlers from scratching content on their websites. The News/Media Alliance recently slammed Google’s new AI model by saying “the content was borrowed and there was no return” for publishers such as Condy Nast and Vox Media.
But this could be a negotiation strategy. AI companies and media organizations have begun to cooperate. Meanwhile, Openai has worked with more than 20 news publishers, including more than 160 media outlets such as The Washington Post, The New Yorker and Wired. Signed with Adweek, Independence, The Los Angeles Times and the World History Encyclopedia. AI labs are approaching the point where they exhaust a lot of high-quality public data suitable for training large language models and are actively looking for new content.
Therefore, these licensing partnerships are very important – not only are AI companies able to develop useful products, but newsrooms can distribute their articles to a broader foundation, but consumers can access researched, well-educated information.
New front page: Entering AI dataset
Because consumers have begun to use AI to search. Google and other search engines are losing their stance as the results have been surpassed by content created by marketers and SEO wizards that push helpless websites to the top. People are increasingly asking Chatgpt and other AI assistants for better, more professional content to search.
Gergely Orosz, the author of a developer-centered pragmatic engineer newsletter, mentioned in May that Chatgpt has driven traffic with Duckduckgo or Bing over the past month, and those visitors have read the page for longer.
Going forward, datasets entering major LLMs will be as important as appearing on the first page of Google search results. Consumers seek product advice, research applications and services, summarize information on complex topics, conduct basic market research or learn about new things. All of these instances are a great opportunity for businesses to attract new audiences in a fresh environment. Companies will fight for such teeth and nails, and the more people flock to AI searches, the more critical the field will become.
This brings us back to the beginning, because the best way to get into the LLM training dataset is to appear in major news media publications that produce high-quality journalism and have direct partnerships with OpenAI, Anthropic, Cloplexity and other AI labs. This further cements the media’s stance and provides them with real avenues for the future.
At the same time, the content optimization that will be included in the training dataset will become the new SEO.
