I haven't said much in here about the geopolitics of AI, but it's reached the point where it's worth being on your radar, especially given how fragile and unpredictable global relations feel right now.

The tl;dr: the US and China are the two AI heavyweights, and are taking very different paths. The US is chasing “the holy grail” of Artificial General Intelligence (AGI) through big tech's mostly closed models. China has bet on open source, releasing models anyone can download freely to become the cheaper, more attractive option worldwide. (Side note: they also are heavily prioritizing robotics, given their huge manufacturing industry, and are way ahead of the US on that front).

To date, the US has owned the leaderboards, but that's shifting. This month China's Moonshot launched Kimi K3, an open-source model rivaling Claude Fable 5 and GPT-5.6 (our very best AI models), and everyone's rushing to try it (and save money). Days later, Alibaba previewed its own near-frontier model, Qwen 3.8.

Now there are rumors that both countries are eyeing export controls on their AI, and the US is threatening sanctions over concerns that Chinese models were trained on stolen American work.

I think more scrutiny and regulation is a good thing, but I’d so much rather see a collaborative international approach - especially now that AI can break out of closed environments and hack into other company systems autonomously 😳. I guess that might be wishful thinking, but I’ll keep hoping.

Before our kids head back to school, there’s something really important for parents to be aware of, and that is the growing issue of teens using AI to create explicit deepfakes.

A new Common Sense Media report came out a couple days ago with some startling data. Almost half of teens have already seen AI-generated sexually explicit material, and almost a quarter have seen it depicting themselves or someone they know.

Many teens expressed interest in talking to trusted adults about deepfakes and pornography, and so parents, it’s important that we do. Common Sense Media offers free resources for starting those conversations here.

The connection spark this week is a little silly, but I couldn’t resist sharing because it was, spontaneously, the highlight of our week.

I’m not even sure how the idea came up, but somehow we decided that we were going to take a trip down memory lane and watch some of the favorite shows from when our boys were little - I’m talking Daniel Tiger, Paw Patrol, Stinky and Dirty, Bubble Guppies, Blaze and the Monster Machines, Number Blocks… I know… you laugh, and so did I.

But, I was so surprised and delighted by how excited and into the idea my boys got. And then, sitting down all together with a bowl of popcorn and watching a handful of their “biggest hits” got us all laughing so hard, reminiscing, and loving every minute.

I highly recommend.

This hands-on is less about any specific prompt, and more about a way to more easily and efficiently hand stuff off to AI that’s in your head as you’re on the go. And that is voice dictation.

There’s a popular tool for this (with a free tier) called Whispr Flow (there are many others also). It instantly turns your rambling with all the ums into clean, punctuated text inside any app on your phone or laptop.

A great use case for this is offloading those random invisible load things you think of throughout the day when you’re on the go… like the soccer registration due Thursday, the almost-gone sunscreen, etc. and then having a magical recall (in a beautifully organized structure) for all of that stuff at the point you’re ready to tackle.

Try it out:

Pro tip: You can get fancy and make an interactive dashboard, for example with Claude Cowork, that can track everything and give you satisfaction as you tick things off the list. I haven’t yet gone this far, but once I do I will report back with best practices.