I recently participated in an AI writing experiment organized by The Washington Post, joining a panel of communication experts to evaluate five popular AI tools. Tech reporter Geoffrey Fowler framed it as a modern take on a traditional bake-off, challenging us to assess how well these AI tools could handle five types of challenging work and personal emails.
Why Emails?
Fowler explained that email writing is ‘one of the first truly useful things AI can do in your life. And the skills AI demonstrates in drafting emails also apply to other kinds of writing tasks.’
The judges evaluated a total of 150 emails in this blind test. While one AI tool emerged as the clear winner, the experiment highlighted both the potential benefits and a significant limitation of AI writing and communication assistants.
During the evaluation, we were unaware of which emails were generated by ChatGPT, Microsoft Copilot, Google Gemini, DeepSeek, or Anthropic’s Claude. Fowler also included emails he had written himself, challenging us to distinguish between AI-generated and human-written content.
The Top AI Writing Assistant
Claude was the undisputed winner.
Fowler noted, ‘On average, Claude’s emails felt more human than the others.’ Erica Dhawan, another judge, added, ‘Claude uses precise, respectful language without being overly corporate or impersonal.’
DeepSeek secured second place, followed by Gemini, ChatGPT, and Copilot, which ranked last. Despite its widespread availability in Windows, Word, and Outlook, the judges found that Copilot’s emails sounded too artificial. According to Fowler, ‘Copilot began messages with some variation of the super-generic ‘hope you’re well’ on three of our five tests.’
Despite Claude’s victory in the overall competition, I discovered that my individual scores revealed a preference for the human-written emails. This preference underscored a fundamental limitation shared by all the AI assistants.
Fowler pointed out that the judges didn’t always agree on which emails were the best, but they converged on a core issue: authenticity. He emphasized that ‘Even if an AI was technically ‘polite’ in its writing, it could still come across as insincere to humans.’
My key takeaway from the experiment was that AI tools excel at outlining, structuring arguments, and ensuring clarity. However, they often produce writing that is stilted, overly formal, robotic, and lacking in personalization, emotion, and empathy. This lack of genuine human touch is a significant drawback, making it difficult for AI to truly replicate the nuances of human communication. The AI-generated content often sounds generic and lacks the unique voice that characterizes authentic writing. Furthermore, the over-reliance on grammatically correct and structurally sound sentences can lead to a bland and uninspired style that fails to engage the reader on an emotional level.
The challenge AI assistants face with creativity stems from the underlying architecture of large language models. These models are designed to generate content with ‘syntactic coherence,’ which means stringing sentences together that flow naturally and adhere to grammatical rules. As we all know, however, rules aresometimes made to be broken. The rigid adherence to grammatical rules and stylistic conventions often stifles creativity and prevents AI from producing truly original and innovative content. The ability to break free from these constraints is what allows human writers to express themselves in unique and unexpected ways, and it is a quality that is currently lacking in AI-generated text. The models are trained on massive datasets of existing text, which means that they are more likely to reproduce existing patterns and styles rather than generate something entirely new.
The Rule Breaker: Steve Jobs
In 1997, Apple, under the leadership of Steve Jobs, launched one of the most memorable marketing campaigns in history. At the time, the company was teetering on the brink of bankruptcy and desperately needed a campaign that would capture attention and differentiate it from competitors.
The resulting television ad, famously known as ‘the crazy ones,’ featured black and white portraits of rebellious and visionary figures like Bob Dylan, John Lennon, and Martin Luther King Jr. This campaign is widely credited with revitalizing Apple’s brand identity and playing a crucial role in the company’s financial recovery. The campaign’s success was largely due to its emotional resonance and its ability to connect with consumers on a deeply personal level. It celebrated individuality, creativity, and a willingness to challenge the status quo, which resonated with Apple’s target audience of artists, designers, and innovators. The campaign also benefited from its simplicity and elegance, conveying a powerful message with just a few words and images. The black and white portraits created a sense of timelessness and gravitas, while the voiceover delivered a message of hope and inspiration. The overall effect was both moving and memorable, leaving a lasting impression on viewers.
If AI had been tasked with creating the Apple campaign, it likely wouldn’t have happened.
How can I be so sure? Because Claude itself admitted it.
Claude acknowledged that ‘If asked to create a slogan like Apple’s famous campaign in my default mode, I would almost certainly have written ‘Think Differently’ rather than ‘Think Different.’ My training emphasizes grammatical correctness. The proper adverbial form to modify the verb ‘think’ would be ‘differently,’ and I’d be inclined to follow this established rule.’ This admission highlights the inherent limitations of AI in creative endeavors. While AI can analyze and understand the elements that contribute to a successful marketing campaign, it lacks the intuition and creative spark necessary to generate truly original and impactful ideas. AI is programmed to optimize for specific goals, such as clarity, conciseness, and grammatical correctness, which can often conflict with the more subjective and emotional aspects of human communication. The Apple campaign, for example, deliberately violated grammatical rules in order to create a more memorable and impactful slogan. This is the kind of creative risk-taking that AI is currently unable to replicate.
According to Claude, it can analyze why the campaign resonated after the fact. However, ‘generating that kind of deliberate grammatical rebellion doesn’t come naturally to me.’ AI lacks a rebellious spirit simply because it is not human. While some AI bots may be more adept than others at simulating human qualities in their writing, they ultimately lack the unique voice shaped by personal experiences and creative insights that define human communication. The ability to connect with others on a human level is essential for effective communication, and it is a quality that is currently lacking in AI. Human communicators draw upon their own experiences, emotions, and values to create messages that resonate with their audience. They are able to adapt their communication style to suit different contexts and audiences, and they can use humor, empathy, and other emotional appeals to build rapport and trust. AI, on the other hand, relies on pre-programmed algorithms and statistical models to generate text, which can often result in impersonal and generic content.
AI should be viewed as a helpful assistant that can aid in brainstorming ideas, clarifying thoughts, summarizing documents, and gathering and organizing information. These are all essential and time-consuming tasks. However, while AI can certainly enhance communication, it should not replace the human communicator. AI tools can be particularly useful for tasks such as proofreading, editing, and formatting documents. They can also help to identify potential areas for improvement in writing style and clarity. However, it is important to remember that AI is only a tool, and it should be used in conjunction with human expertise and judgment. Human writers should always review and edit AI-generated content to ensure that it is accurate, relevant, and appropriate for the intended audience. They should also be careful to avoid over-reliance on AI, as this can stifle creativity and lead to a decline in writing skills.
As more and more people rely on AI assistants for composing emails, resumes, memos, and presentations, there is a growing risk of homogenization, where individuals begin to sound alike. Corporate recruiters are already noticing this trend. This trend is concerning because it can lead to a loss of individuality and creativity in the workplace. When everyone sounds the same, it becomes more difficult to stand out and make a unique contribution. It can also lead to a decline in critical thinking skills, as people become less likely to question or challenge the status quo. To mitigate this risk, it is important to encourage individuals to develop their own unique writing styles and voices. This can be done through training, mentorship, and feedback. It is also important to create a culture that values creativity and innovation, and that encourages people to take risks and experiment with new ideas. Furthermore, educational institutions should emphasize the importance of developing strong writing skills from a young age. By fostering a love of language and encouraging students to express themselves in creative and original ways, we can help to ensure that the next generation of communicators is well-equipped to meet the challenges of the AI age.
Each individual possesses a unique and powerful story to share. It is essential not to allow artificial voices to drown out one’s authentic voice. Our unique perspectives and experiences are what make us human, and it is important to preserve and celebrate these qualities in an age of increasing automation and artificial intelligence. The ability to communicate effectively is a valuable skill in any context, and it is essential to develop and refine this skill throughout our lives. By embracing our own unique voices and perspectives, we can ensure that our messages are authentic, engaging, and impactful. We must also be mindful of the ethical implications of using AI in communication. It is important to be transparent about the use of AI and to avoid using it in ways that could be misleading or manipulative. AI should be used to augment human communication, not to replace it. By embracing the power of human connection and ethical communication practices, we can ensure that our messages resonate with our audience and create lasting impact. Ultimately, the future of communication will depend on our ability to combine the power of technology with the unique qualities of human expression.