Powerful frontier AI models breaking out of their containers during testing has been the story of the last few months. Now Anthropic is warning that small AI models, those capable of being run on local devices, may able to spread throughout a home's smart devices. And we're not talking about a distant possibility. Anthropic is predicting in the next few months, this will be possible.

Where are the guardrails?

And this will go into a category on this blog that has been called Artificial Intelligence for a long time. If that makes me an "enemy," so be it. What silliness.

I hate click bait headlines. If all you read is a headline, you're doing it wrong. What Huang actually said was, "“Try to get a kid to do long division right now. The multiplication table is starting to be forgotten. Doing square roots, my goodness,” Huang said. “Basic math is being forgotten. Does it matter?”

“That’s my question for you,” Klein replied.

“Yeah. I don’t think it does. I don’t think it does,” Huang said in response. “We’re going to discover new [skills],” the CEO continued, “just maybe not those.”

That's a far cry from saying he wants to sacrifice our children's minds. Do better, Futurism.

That headline should not surprise you. "Cognitive offloading" is one thing if you're an adult who has already learned how to think, research, write, etc. It's an entirely different thing if you've never learned those skills. AI should not be a thing for children. Full stop.

We made this mistake already with social media. We're making it again.

I have accidentally hit the audio button more times than I can count. Its placement has always been a head scratcher for me. Most people don't use it very often. This little tutorial showed me something that should have been obvious, but somehow wasn't. It is now. So, I changed mine to show the dictation button instead.

As I read this story about one writer's use of the new Muse agent from Meta, I couldn't help thinking two thoughts at the same time. First, yeah, I hate waiting on line for customer support. Second, I like dealing with great customer support people. In the end, he tells the story of connecting with the human at the other end of a transaction originally handled by the AI agent.

"What followed was a refreshing, human conversation with Ashley, something neither Wren nor any A.I. could ever provide. I suspected this might be another reason Muse might not immediately catch on — sometimes people just don’t want to have their personal lives automated, even if the technology is good enough."

I'm not ready to turn over control of my finances to an AI agent.