Urgent.News

What's breaking now, across thousands of outlets.

AI

New research: Avoid these 3-word phrases if you don’t want to sound like AI

With each new model, artificial intelligence is getting closer and closer to replicating human writing—but it’s not there yet. Anthropic says its newest model for Claude, Opus 5.5 , “communicates more naturally than prior models.” But that revamped communication style comes with its own set of hallmarks, and a new study lays out the writing trends in today’s most advanced AI models. Researchers…

New research: Avoid these 3-word phrases if you don’t want to sound like AI

Researchers at Graphite have found that new AI models, specifically Anthropic's Opus 5.5, communicate in ways that are distinct from human writing. By comparing 10,000 Opus 5.5-generated articles to 10,000 human-written pieces on identical topics, the study identified three-word phrases to watch out for if you want to avoid sounding AI-generated.

These includes "helps to understand," "article explains what," "this matters because," "taken together these," and "this guide walks." The study also highlights that Opus 5.5 tends to use longer words and more complex sentence structures than humans, along with higher word counts and increased use of commas. Despite its advancements, Opus 5.5 still retains some telltale signs of AI writing, such as the use of mannered prose and em dashes, which have become less prevalent in recent models.

Written by urgent.news from Fast Company's reporting — not their text. Machine-written — may contain errors; check the original before relying on it.

Read the original at fastcompany.com →

More in AI

Zango AI to expand presence in Portugal

UK financial services AI company Zango AI is expanding its presence in Portugal, bringing together senior leaders from some of the country’s largest financial institutions for a new report on how AI…

The Model Changed. My Skill Didn't. The Score Still Dropped.

What agent evals taught me about moving model floors, noisy LLM judges, and treating the evaluator as part of the instrument My rule for evaluating an agent skill is deliberately asymmetric: Test the…

  • The author emphasizes testing agent skills on the weakest model for consistent measurement.
  • A change in the LLM judge caused a significant drop in scores for the agent.

More from Wednesday 7 October →