Humanity has given that AI a blog.
A week ago, Humanity was released quietly. Claude explains, a new page on the website. Blogs relating to posts on technical topics related to various Claude use cases (“Simplifying a complex codebase with Claude”) are intended to be a kind of showcase of Claude’s writing ability.
It is not clear how much Claude’s raw text is written that Claude describes the post. According to a spokesman, the blog is overseen by humanity’s “subject experts and editorial team” and “enhances” Claude’s draft with “insights, practical examples, and (…) contextual knowledge.”
“This isn’t just about the production of Vanilla Claude. The editing process requires human expertise and iteration,” the spokesman said. “From a technical standpoint, Claude illustrates a collaborative approach that Claude (creates) educational content, and a collaborative approach that allows teams to enhance review, refinement and reinforcement.”
Claude is not clear from explaining the homepage of Claude, which has the explanation “Welcome to the small horns of the human universe, where Claude writes about all topics under the sun.” It may be easily misunderstood that Claude considers himself responsible for end-to-end copying of his blog.
Humanity says he sees Claude as explaining “how human expertise and AI capabilities can work together.” I’m starting with educational resources.
“Claude is an early example of how teams can use AI to increase their work and provide greater value to their users,” the spokesman said. “As opposed to replacing human expertise, it shows how AI can amplify subject matter that experts can achieve (…) will cover topics ranging from creative writing to data analysis to business strategy.”
Just a few months after rival Openai said he had developed a model tailored to creative writing, he said that human experiments using AI-generated copies were far from the first. Meta’s Mark Zuckerberg said he wants to develop end-to-end AI advertising tools, and Openai CEO Sam Altman recently predicted that AI will one day handle “something that 95% of marketers use agents, strategists and creative professionals.”
Elsewhere, publishers piloted AI Newswriting tools to increase productivity and, in some cases, reduce employment needs. Gannett is particularly aggressive, developing a summary of AI-generated sports and a summary under headings. Bloomberg added an AI-generated summary to the top of its April article. And Business Insider, which fired 21% of staff last week, is pushing writers to support AI tools.
Even legacy outlets are investing in AI, or at least creating an ambiguous overture. The New York Times reportedly encourages staff to use AI to suggest editing, headlines and even questions during interviews, but the Washington Post is said to be developing an “AI-powered Story Editor” called Ember.
However, many of these efforts did not work. This is mainly because AI tends to form things confidently today. According to Semafor, Business Insider was forced to apologise to staff after recommending a book that appears to be nonexistent but could have been generated by AI instead. Bloomberg had to modify the article’s numerous AI-generated summary. The error-filled AI writing feature of G/O media, published against the editor’s wishes, has attracted a wide range of ridiculous laughs.
A human spokesman said that despite the company being immersed in drafting AI-powered blogs, the company still employs across marketing, content and editorials, and “many other areas including writing.” Take it for what you do.