The Amazon founder's Washington Post unveils an AI-focused offering, facing a drop in subscriptions by 250,000 users.
The Washington Post has had a tough stretch, shedding over 250,000 subscribers following the directive from owner Jeff Bezos to abstain from endorsing any candidate in the 2024 election. Yet, there's a silver lining: behold, the debut of "Ask The Post AI," a new AI chatbot from the Post!
This new AI chatbot, referred to as such, delivers AI-generated responses to inquiries, drawing upon The Post's extensive archive of stories published since 2016. For example, inquiring about the number of homes damaged in Colorado's 2021 Marshall Fire, the bot will reply, "The Marshall Fire that struck Louisville, Colorado, in December 2021 left over 1,000 homes in ruins, making it the most devastating blaze in the state's history." Links to relevant Post articles are included at the bottom of its response.
The Post communicated excitedly about this development in its announcement, expressing that the rising trend of generative AI and conversational formats is an opportunity "to bring our rich tradition of fact-based, meticulously reported journalism to life in fresh and engaging ways, delighting and informing readers."
"Ask The Post AI" is basically a conversational interface for the newspaper's existing reporting.
News organizations have been scrambling to safeguard themselves against the emergence of AI chatbots like ChatGPT. Some have aligned with AI companies for monetary compensation, while others have opted for lawsuits to impede unauthorized utilization of their journalism. In The Post's instance, it is launching its very own AI chatbot.
The Post acknowledges that chatbots are often prone to speculation or falsehood, as they are merely statistical programs that predict the probability of the next word in a sequence. They are incapable of distinguishing between reality and hearsay or distinguishing between trustworthy and dubious sources.
However, chatbots generally perform better when constrained by a precisely curated dataset rather than having open-ended access to information on the web. "By restricting the search results to our published content, we are ensuring that every datum synthesized by the AI is derived from work already published by The Washington Post newsroom." According to The Post, its AI will refrain from responding if it deems the information inadequate.
The article emphasizes that individuals should confirm the chatbot's results with the original articles as "the chatbot may not function exactly as planned." Nonetheless, the bot may serve as a useful starting point, addressing even complex queries like "why are so many Brazilians in Cape Cod?" with a straightforward response while also offering links to additional interesting, in-depth stories.
Proponents of AI in search contend that chatbots will not divert traffic, as individuals will continue to yearn for in-depth coverage and will employ chatbots as mere entry points or to answer straightforward questions like "why is the sky blue?" However, anyone who has used Google recently is aware of its AI-generated replies often placed atop search results, which are proficient at addressing many questions (the accuracy of such answers is another matter entirely). It is challenging to dismiss the possibility that some traffic will inevitably fall prey to AI.
The Post aspires that a conversational chatbot may serve as a sufficient platform to "delight and inform" readers. With recent revelations concerning Bezos's eleventh-hour decision to withhold presidential endorsements from the newspaper – a move criticized as an attempt to curry favor or thwart additional animosity – the Post hasn't been a reader-delighter recently. Ownership by a business tycoon with diverse interests comes with its own set of challenges.
The Post believes that the integration of AI technology into journalism, as demonstrated by "Ask The Post AI," holds significant potential for the future of news delivery. ThisChatbot, powered by generative AI, can bring The Post's fact-based journalism to life in engaging ways, appealing to modern readers.
In the face of competitors like ChatGPT, The Post has chosen to leverage AI technology instead of fighting against it, recognizing its potential to enhance and expand the reach of its reporting.
