Can AI be a reliable journalist or musician?

Recently, a guitar accessory company called D’Addario was caught using an AI tool called Suno to regenerate a promotional music track. After the company initially denied it—blaming audio artifacts on things like digital noise and the use of plug-ins like Autotune—experts proved the audio was indeed created by AI. In a separate trend, a growing number of automated newsrooms, such as RuntimeWire, are using AI agents—autonomous programs that can perform tasks like searching the web, writing, and editing—to churn out thousands of news stories. These sites aim to beat human journalists to the punch by monitoring social media and transcripts in real time, publishing stories in minutes without a human ever drafting the text.

These AI tools function as advanced prediction engines. When you ask a music-generator like Suno to regenerate a track, it is not actually playing an instrument. Instead, it analyzes millions of existing songs it was trained on—meaning it studied their patterns, rhythms, and structures during a massive learning phase—to calculate what the next sequence of notes should be. It is essentially an autocomplete machine for sound. Similarly, automated news sites use language models to ingest streams of data, such as transcripts from a conference. The AI filters this information to spot patterns that look like news, drafts a summary, and occasionally publishes it without human review. The core mechanism is probability; the machine calculates the most likely combination of words or notes to fit a request, rather than understanding the meaning behind them.

The danger here is not just that the content can be flawed, but that it bypasses the accountability that comes with human work. When D’Addario used AI, they were able to obscure the process for weeks because they offered technical excuses that seemed plausible. In journalism, speed often comes at the expense of accuracy. These automated outlets might accidentally prioritize unimportant details or share errors, and because they are designed to prioritize quantity, they can become part of a feedback loop where AI models start sourcing their information from other AI-generated articles. While these tools can handle simple tasks, they struggle to replicate the human work of building trust with sources or verifying if something is actually true. As AI continues to fill these roles, we have to decide if we value the convenience of instant content more than the integrity of the source.