|
We are offering a limited cyber sale now that offers you savings of 75% at $1 a week. Don't miss out. Sign up now.Greetings!While the bidding war between Netflix and Paramount over Warner Bros. Discovery dominated the front half of the week, a myriad of headlines related to artificial intelligence takes us into the weekend. Taken separately, the news ranges from the boneheaded (like the Washington Post's error-prone AI podcasts) to the troubling (President Donald Trump's executive order limiting state regulations related to AI) and
in between (Disney's OpenAI deal, which I break down further in this newsletter). As a whole, they serve as a reminder of just how prevalent AI is in our lives — for better or worse. Here's a helpful recap: - On Thursday, Trump signed an executive order limiting the role of state regulations on AI, arguing that he didn't want AI companies following a patchwork of guidance. “To win, United States AI companies must be free to innovate without cumbersome regulation,” the executive order read. The only problem is there aren't federal laws yet, and it's unclear when one will actually pass.
- The International Alliance of Theatrical Stage Employees (IATSE) on Friday lambasted Trump's order. "Preventing states from passing and enforcing AI protection before Congress acts is an invitation for lawlessness that directly threatens IATSE members and their economic livelihoods," the union said.
- After the Washington Post launched its AI-generated podcasts on Thursday, its staffers quickly found fabricated quotes and editorializing. That's one of the core problems of AI, which is essentially a black box when it comes to how it generates its content. When we wrote about the company churning out 3,000 AI-generated podcast episodes a week, the owner said there was some human oversight on the most sensitive topics. I wonder how much oversight the Washington Post had on its episodes?
- Also on the AI error front but with less stakes, Amazon's experiment with AI-generated recap videos hit a snag with its look back at the first season of hit video game adaptation "Fallout." Several fans noticed the AI system misunderstood or misrepresented key details from the show, forcing Amazon to pull it down.
- Then there's the Disney-OpenAI deal, which I break down even further in my analysis this morning.
- OpenAI also released ChatGPT5.2, a model that it claims hallucinates less and is less prone to mistakes. This comes hot the heels of last month's release of Google's Gemini 3, which topped ChatGPT on many major AI model benchmarks.
If you can't get enough
of AI, stay tuned next week for my piece recapping 2025 and looking at what's in store for the technology next year. Have a great weekend! Roger Cheng
Managing Editor of Business and PRO
roger.cheng@thewrap.com
AI companies, whether its OpenAI, Google, Meta or countless others, are in an arms race with each other and, more broadly, with China and its host of AI ventures...
To continue reading, subscribe now with a 2-week free trial.
Free for 14 Days – Then Just $4/Week
Free for 14 days, then $4/week (billed annual at $199). Renews yearly. Cancel anytime to avoid future charges.
|
