Ad
Short

OpenAI CEO Sam Altman believes that AI will eventually master poetry, reaching what he calls a "10 out of 10" human level. Still, he thinks this achievement won't resonate with most people. Altman argues that what gives art its value is its human origin. Even if an AI writes a technically flawless poem, it will likely be missing a genuine emotional core.

Altman draws a parallel to chess: although machines now consistently outperform people, players still prefer to compete against other humans. The fun comes from measuring yourself against another person, not a computer. Watching two AIs play is just not that interesting for most viewers.

Altman has said before that as AI-generated text and images become more common, people will start to value content from real humans even more. "My directional bet would be that human-created, human-endorsed, human-curated content all goes up in value dramatically," Altman said.

Ad
Ad
Short

Entrepreneur Niels Hoven has released an alphabet book featuring nearly 1,000 illustrations made with AI. Hoven says producing this many images by hand would have been too complicated and expensive for human artists. With each illustration taking about two hours, he estimates the project would have cost around $50,000. The use of AI instead drew criticism on social media and in Amazon reviews.

A screenshot highlights the controversy over AI image generators. Recent UK court decisions say training AI models on existing works does not violate copyright. | via X

Hoven addressed the criticism, explaining that without AI, a hardcover edition would have cost about $200. Thanks to generative AI, the book is now available as a free PDF and as a $30 hardcover, with all proceeds going to Amazon for printing and shipping. Hoven says he earns nothing from sales. He maintains the book couldn't have been made without AI, and that its main purpose is to help children learn to read. Still, the project doubles as advertising for Hoven's company, which offers a related learning app.

Ad
Ad
Ad
Ad
Short

Barclays estimates that data centers for AI giants like OpenAI, Meta, and Amazon could need as much as 46 gigawatts of electricity. That's about the same amount of power used by 44 million US households, or roughly a third of all homes in the country. Building out these projects could cost $2.5 trillion.

This massive expansion is already putting pressure on the power grid. Nvidia, Microsoft, and OpenAI are warning about possible grid instability caused by rapid swings in electricity demand. Some plans call for energy sources like solar plants and gas storage to be integrated directly into the data centers. OpenAI has even asked the US government to add 100 gigawatts of new power generation each year. It's still unclear how many of these projects will actually get built. According to Barclays, it's hard to draw the line between projects that are real and those that are still just speculation.

Google News