Creative Marketing & Event Strategist Gerald Valentine
based in North Texas

When AI Misfires: The Human Toll of Faulty AI Writing Detectors

It’s no secret that I’m a proponent of AI as a useful tool for creatives. But, as with any new tool, there are huge downsides that accompany the limitless upside potential. This is especially true in these early days of AI adoption as individuals and companies struggle to learn how best to implement these tools into their workflows. One of the most significant downsides we’re seeing right now involves AI detectors mistakenly flagging human-written articles as being AI-generated, which is causing real harm to writers’ careers.

Consider one of the major factors leading to some of these false flags: the use of Grammarly, a widely-used tool for fixing punctuation and grammar mistakes. Grammarly helps make writing clearer and free of errors, which is essential for professional work. However, this polished writing can sometimes look like it was created by AI, causing AI detectors to flag it. So, writers who try to improve their work may end up having their work flagged as being created using AI. This issue is especially frustrating because many companies actually require their employees to use Grammarly to clean up their work.

“AI trains on millions of human-written articles. AI gets really good at imitating human writing. AI thinks everything is AI.”

David Puddy the 2nd – Gizmodo Commenter

AI detectors are not very reliable, which is a big problem given what’s at stake. Bars Juhasz, speaking with Gizmodo, pointed out these concerns: “We have a lot of concerns around the reliability of the training process these AI detectors use. These guys are claiming they have 99% accuracy, and based on our work, I think that’s impossible. But even if it’s true, that still means for every 100 people there’s going to be one false flag. We’re talking about people’s livelihoods and their reputations.”

For writers, being wrongly accused of using AI can have serious consequences. Many writers depend on their reputation for original work to get jobs and keep steady work. If their work gets flagged, they can lose opportunities, damage their reputations, and even lose their jobs. Freelance writers and those working on strict contracts are especially at risk because clients may decide to end agreements rather than risk publishing something that might be AI-generated.

On top of the financial hit, these false flags also take a toll on writers’ mental health. Writers put a lot of time, creativity, and effort into their work. Being wrongly accused of using AI undermines their skills and dedication, sometimes tarnishing a career that took years to build. This can lead to lower motivation and increased stress, which hurts their productivity and well-being.

AI is here to stay, and it should be seen as a helpful tool, not something to be feared. Some of the biggest companies in the world also recognize the potential of AI and are making significant investments to integrate AI into their products. Microsoft and Google heavily invest in AI technologies to enhance their software and services. Similarly, Adobe has introduced AI tools into its Creative Suite to help artists be more effective, allowing for enhanced creativity and productivity. It seems hypocritical to provide and encourage AI tools for graphic artists, but make the use of AI tools a taboo for those whose creative interests tend to be more lexical. Instead of punishing employees based on dubious results from AI detectors, companies should help their employees use AI responsibly. By providing training and clear guidelines on ethical AI use, companies can empower writers to improve their work without fear of being made a pariah.

Boiled down to its core, the basic idea of AI being able to detect other AI-generated content seems questionable. As one Gizmodo commenter put it, “AI trains on millions of human-written articles. AI gets really good at imitating human writing. AI thinks everything is AI.” This shows the flaws in current AI detection methods.

There may never be an AI detection tool that’s infallible, and employers should question whether using these tools to make crucial employment decisions is really the best path forward. Instead it should be the quality of the work being produced that should be examined, regardless of the tools being used to create it.

Writers need to stand up for themselves and each other. The writing community needs to raise awareness about this issue and push for improvements in AI detectors. By working together with tech companies and industry groups, we can create are more effective environment that embraces AI as a tool that makes all of our lives easier.

After all, AI really can be a great tool for writers, helping them enhance their work. But the risk of false accusations by AI detectors is a serious threat. By recognizing this problem and working on better solutions, we can ensure that human writers are appreciated and valued for their contributions.

The Washington Post’s AI Expansion: A Risky Move in Troubled Times

In the wake of a significant financial loss, the Washington Post is making headlines for a bold new strategy: expanding the use of artificial intelligence (AI) in its newsroom. After reporting a $77 million loss, the media giant is turning to AI in hopes of streamlining operations and reducing costs. However, this move raises critical questions about the future of journalism and the potential consequences of replacing human writers with machines.

The Washington Post’s Chief Technology Officer, in a recent statement, encapsulated the organization’s new direction: “The way I see us operate is we have A.I. everywhere. We have A.I. in our newsroom; we have A.I. with our consumers; we have A.I. in the business.” This comprehensive integration of AI reflects a broader trend in the industry, where technology is increasingly being leveraged to enhance efficiency and productivity.

“The way I see us operate is we have A.I. everywhere. We have A.I. in our newsroom; we have A.I. with our consumers; we have A.I. in the business.”

Vineet Khosla, Washington Post CTO

The strategy seems innovative on the surface. AI can assist in various aspects of news production, from generating quick updates and automating routine tasks to analyzing vast amounts of data for investigative pieces. Yet, there is a fine line between using AI as a tool to aid human journalists and allowing it to replace them.

The primary concern with the Washington Post’s approach is the potential for AI to supplant human creativity, intuition, and critical thinking—qualities that are the bedrock of quality journalism. While AI can process information and generate text at an unprecedented speed, it lacks the nuanced understanding and ethical considerations that human writers bring to their work. The risk here is that news becomes more about quantity than quality, leading to a decline in the depth and integrity of reporting.

The move to replace human writers with AI can undermine the trust between the publication and its readership. Readers expect and deserve well-researched, thoughtfully crafted stories that reflect the complexities of the world around them. A shift towards AI-generated content could erode this trust, as the subtle biases and errors inherent in AI algorithms become more apparent.

In addition to these concerns, expanding AI might not help the Washington Post’s credibility. In 2022, 48 percent of respondents believed the publication to be very or somewhat credible. This level of trust is significant in a media landscape where credibility is constantly scrutinized. Introducing AI-generated content could jeopardize this credibility, as readers may question the authenticity and reliability of AI-driven stories. The perception that news is produced by machines rather than experienced journalists could lead to skepticism and diminished trust.

The impact on newsroom staff cannot be ignored. Journalists facing the threat of redundancy may find their morale and motivation plummeting, which can further degrade the quality of the content being produced. The human element of journalism—built on passion, curiosity, and a commitment to truth—is irreplaceable. AI should be seen as a supportive tool, enhancing the capabilities of journalists rather than rendering them obsolete.

The Washington Post’s financial struggles are indeed a pressing issue, but turning to AI as a panacea might create more problems than it solves. Instead of viewing AI as a replacement, it should be integrated in a way that supports and augments the work of human journalists. This balanced approach can ensure that the publication continues to produce high-quality journalism while also benefiting from the efficiencies that AI can offer.

The Washington Post’s decision to expand AI in its newsroom amidst significant financial losses is a gamble. While AI can undoubtedly bring benefits, it should be implemented as a complement to human effort, not a substitute. The future of journalism depends on maintaining the delicate balance between leveraging technology and preserving the irreplaceable qualities that human journalists contribute to their craft. As the industry evolves, this balance will be crucial in ensuring that journalism continues to serve the public good with integrity and depth.

The Challenges of Predicting the Future of AI: ChatGPT and Beyond

When OpenAI unleashed ChatGPT onto the world in November 2022, it felt like a seismic shift in the landscape of artificial intelligence. This conversational AI, built on the GPT-3.5 architecture, showcased an unprecedented ability to understand and generate human-like text, setting a new standard for what AI could achieve. But as awe-inspiring as these advancements are, predicting the future trajectory of AI remains a Herculean task, laden with uncertainties and unexpected turns.

The Rapid Evolution of AI

ChatGPT’s debut was a pivotal moment, not just for AI enthusiasts but for the broader tech industry. In the months and years following its release, AI technology has continued to accelerate at a blistering pace. From significant strides in natural language processing to breakthroughs in computer vision and autonomous systems, AI is advancing faster than many could have imagined.

Why AI Predictions Are So Tough

The primary challenge in forecasting AI’s future is the dizzying speed of technological progress. Innovations are emerging so rapidly that even the most forward-thinking experts struggle to keep up. Moreover, AI development is a melting pot of various disciplines—hardware, software, data science, and ethics—all of which evolve in tandem and influence each other in unpredictable ways.

A Look Back: Tom Scott’s 2012 Prediction

To appreciate the difficulty of predicting tech trends, it’s helpful to revisit past forecasts. In 2012, YouTuber and technologist Tom Scott took a stab at predicting what the world would look like in 2022. While his predictions were insightful, he, like many others, missed some of the most significant changes.

Among other misses, and despite some of the first signs of the change already being felt, Scott did not foresee the mobile-first revolution and the sweeping shift from dedicated web pages and applications to mobile apps.

In 2012, after I’d spent more than a decade writing websites on and for desktop computers, I’d just missed that entirely. That transition was starting—just starting—and I hadn’t spotted it.

Tom Scott reflecting on his prediction oversight.

The explosion of smartphone use and the dominance of mobile apps have revolutionized user behavior, business strategies, and social interactions. This gap in prediction underscores just how challenging it is to foresee which technologies will gain traction and transform our daily lives.

The Uncharted Territory of AI’s Future

AI’s future is a vast, uncharted territory. While we can speculate about potential advancements—like more sophisticated language models, improved machine learning algorithms, and deeper AI integration across industries—the precise path is anything but clear. Factors such as regulatory changes, ethical considerations, and unexpected technological breakthroughs will all play crucial roles in determining AI’s trajectory.

The societal impact of AI, including its influence on employment, privacy, and decision-making, adds yet another layer of complexity. The ethical and philosophical questions that AI raises are profound and constantly evolving, making it even more difficult to predict the full extent of its future impact.

My Prediction

Now that I’ve laid out why it’s not necessarily a great idea to predict what the future of AI could look like, I’m going to throw my prognosticator hat into the ring and make a prediction of my own.

Looking ahead, I predict that AI will become another indispensable tool in the creative workbelt, much like Photoshop did in the early ’90s. AI will assist creators across various disciplines, enhancing their productivity and expanding their creative possibilities. This transformation will make AI a ubiquitous presence in creative industries, driving innovation and efficiency.

However, I foresee that the most popular application of AI will be as a personal assistant capable of handling a wide array of tasks. From sifting through your email archive for a specific attachment to making restaurant reservations on your behalf, AI will become an integral part of our daily routines. Companies like Google have already demonstrated some of these capabilities, and I believe this will become the most commonplace use of AI within the next 5-10 years.

Embracing the Unknown

As we gaze into the future of AI, embracing uncertainty becomes essential. The rapid pace of technological change demands continuous learning and adaptability. Instead of trying to pin down exact predictions, we should prepare for a range of possibilities and cultivate a mindset that can navigate the unknown.

In the end, the challenge of predicting AI’s future lies in its rapid evolution, interdisciplinary nature, and the sweeping societal changes it brings. Reflecting on past predictions, like those of Tom Scott, highlights the unpredictability of technological advancement. The best approach is to stay open-minded and flexible, ready to adapt to whatever surprises the future holds. As AI continues to advance, so too must our understanding and anticipation of its potential, ensuring that we remain active participants in shaping the future of this transformative technology.