Welcome to my first blog! In today’s era of digital transformation and the unprecedented cacophony of voices competing for attention and customer mindshare, the demand for quality content that cuts through this noise has never been higher.  And while content generated by Artificial Intelligence (AI) seems to be the holy grail to produce quality content in the easiest and most efficient way, AI models like ChatGPT-3 are not perfect. 

They do not have the ability to discern right or wrong and may not always produce text that aligns perfectly with the user’s intentions or meets specific requirements.  Instead, generative AI can turn into a Trojan horse of impersonalization, sameness and inaccuracy.  

In fact, according to the latest Global Risks Report from the World Economic Forum, misinformation and disinformation aided by AI is listed as the most immediate risk to the global economy over the next two years as it threatens to erode democracy and polarize society.

As we create content, it is crucial to understand the limitations of generative AI.  The recent federal lawsuit by the New York Times against OpenAI and Microsoft goes beyond its claim that its copyrighted material was used to train chatbots, adding that this in fact makes chatbots competitors to the Times (and others) as sources of reliable information.

Earlier, for example, the New York Times had reported that “Chatbots May ‘Hallucinate’ More Often Than Many Realize.”  These “hallucinations” vary widely among the leading AI companies: OpenAI’s technologies had the lowest rate, approximately 3%; systems from Meta, which owns Facebook and Instagram, were approximately 5%; the Claude 2 system by Anthropic, an OpenAI rival, topped 8%; and a Google system, Palm chat, had the highest rate at 27%.  

This can lead to some serious problems using generative AI. Most notably, ChatGPT cited a half dozen court cases that didn’t exist in a 10-page legal brief that an attorney submitted to a Manhattan federal judge. This lawyer then asked a judge for leniency after the discovery.

According to Morning Brew, there have been a number of other major AI “hallucinations” since ChatGPT was launched in November 2022:

• In April, George Washington University professor Jonathan Turley was falsely named by ChatGPT when a UCLA professor asked for five examples of college professors who were accused of sexual harassment. ChatGPT referenced a Washington Post article that did not exist, although the paper wrote a real follow-up article that vindicated Turley.
• Alphabet lost $100 billion in market value when Google’s chatbot Bard made false claims about the James Webb Space Telescope in its debut.
• Microsoft’s AI-powered Bing search engine offered blatantly wrong information on topics like Mexican nightlife, Billie Eilish and the Gap.
• ChatGPT reported that a mayor in Australia was imprisoned for bribery when he was actually a whistleblower and never charged with a crime.

Why Writing For Humans™ Is Important

With issues like these – and others – Writing For Humans puts a highly experienced human writer in charge of the editorial process of AI content editing.  This is the expertise and experience necessary to ensure that the finished content is factually correct, original, relevant to its key audiences and contains a personal point of view.  

Founded by an award-winning writer, journalist, and senior public relations and marketing executive, Writing For Humans is dedicated to delivering original and AI-generated content that forges a close connection to a company’s key audiences.  Writing For Humans believes in the power of telling a company’s story through clear, concise and compelling communications to its customers and other key audiences — content that moves them to act. 

Benefits of a Having Human in Charge of the Editorial Process

The combination of having a human writer in charge of the editorial process with generative AI leverages the strengths of both to produce high quality and ethical content.  Here are seven benefits of having a human in charge of the editorial process:

• Quality Control: AI can generate text that may contain errors, inaccuracies, biases or inappropriate content.  Human editors review and correct these issues to ensure that that the finished content meets the highest quality standards.
• Personalization of Style and Tone: Human editors can adjust the style and tone of AI-generated content to match the desired company “voice,” include key company messages and ensure a personal point of view that will resonate with a company’s key audiences.
• Context and Nuance: AI models may not fully understand the context or nuance of a given topic, especially when dealing with complex or sensitive subjects. This can lead to content that lacks depth or is inaccurate, misleading or irrelevant.
• Timeliness: Because of the large expense of keeping AI models up-to-date, they are not aware of the most current events in today’s rapidly changing world.  Human editors ensure that the content is current and relevant.
• Ethical Considerations: AI models can generate biased or offensive content because they are trained on massive amounts of data, and this data may not always be accurate or unbiased.  Human editors identify and eliminate such biases to produce more ethical and inclusive content.
• Legal and Compliance Issues: Human editors can help ensure that AI-generated content complies with such legal and regulatory requirements as copyright laws or industry-specific guidelines.  This is crucial for businesses and organizations to avoid potential legal and reputational risks.
• Accountability: We are now familiar with the issues of generative AI content that has not been reviewed by humans.  Company employees, stakeholders, customers and the media should feel more confident in content that has been reviewed and approved by humans.  This, in turn, leads to increased confidence and trust in the organization that produced the content.

Stay tuned to this blog for more updates and insights from Writing For Humans, as I explore the nexus between the necessity of human editorial oversight of AI-generated content and content strategy and creation.

###