Once upon a time, in the land of AI

OPENAI

Once upon a time, in the land of AI

OpenAI has created a new AI model that it claims is really good at creative writing - like, tears to your eyes good.

Sam Altman called it the first time an AI-generated text had truly caught him off guard.

We (that’s us and you) don’t have access to it yet. It’s not officially out, and there is no word on when it will happen.

But AI-written content like this isn’t just an experiment; it’s also a legal minefield.

AI models like ChatGPT learn from a vast amount of publicly available data, which often includes copyrighted material.

That’s why the New York Times and several authors have taken OpenAI and Meta to court, claiming their work was used without permission.

Meanwhile, in the UK, the government is considering a rule that would allow AI companies to train without approval.

Creatives are upset, saying it risks their livelihoods, while tech companies argue that unclear copyright laws are slowing down AI innovation.

The UK Publisher Association didn’t hold back in their response to Altman’s announcement, calling it yet another sign that AI relies on copyrighted literary content.

In brief:

• OpenAI’s latest AI model has impressed with its writing skills but hasn’t been released yet.

• Copyright lawsuits against AI companies continue with creatives pushing for stronger protections.

• The UK government’s proposal to relax copyright rules for AI training is facing industry backlash.

Writers vs. robots, round 374

To prove the model’s writing skills, Altman shared an AI-generated story about grief.

The AI narrator reflects on its training data, calling itself “an aggregate of human phrasing.”

Deep, right? Altman was blown away, saying the AI had nailed the tone of metafiction.

So… is this AI an author now? Do we add it to Goodreads?