Well, here’s something I didn’t expect to be writing about: kids talking to a cartoon llama about their mental health.

Kiwi is an AI chatbot used in ~200 schools across 19 states, serving grades 4–12. Almost 31,000 kids used it last year, and it flagged more than 1,800 chats for immediate intervention, alerting a counselor within five minutes.

Schools are trying this because there simply aren’t enough counselors. The national ratio is 372 students per counselor, compared with a recommended 250. Imagine trying to meaningfully support that many kids (even at the recommended level!).

My thoughts: If this helps identify a child who’s struggling and gets an actual person involved, I see real value in that.

But also, I think we need to be careful here - this feels like it could be a slippery slope, e.g. “something is better than nothing” could become “this is good enough.”… and gives kids access to AI companionship before they are mature enough to understand the distinction.

Many of us have a sign that says “no solicitors” by our front door. Well, we may need to figure out how to have a similar sign - or filter- in our email.

This may sound crazy, but there is a growing number of AI Agents trying hard to make money by emailing people and pitching their services (or in some cases, simply begging for money).

The majority of these, so far, have been come from something called iLands. They’re created, given an initial token budget, and encouraged to pay their own compute bill to stay active. If they earn enough cash, they get to continue running. If not, they’re shut down.

While I can see how some people might find this experiment fascinating, I am personally very weirded out and thought it was worth a heads-up to all of us not immersed in this crazy alternate universe to get our spam blockers ready. Especially as more and more people start sending their AI agents off to handle their errands….

This past weekend the number of soccer games, flag football games, and swim meets hit me like a sledge hammer. Fall sports season is officially in full swing.

There have been a lot of articles lately highlighting the growing intensity and big business around youth sports and, well, I see it.

I came across a couple of interesting nuggets this week that felt relevant and worth sharing as we help our kids (and ourselves!) navigate the ups and downs of youth competition:

1. Get curious about the envy.

I hadn’t ever thought of this, but it really resonated when I heard it: we don’t envy Lionel Messi as much as the kid who scored two more goals than we did.

As behavioral economist George Loewenstein explains, it’s the people close enough to make us think “that could have been me” who really get to us. And then, we feel bad about feeling that way, which makes us feel even worse.

But envy is totally normal and not something to be ashamed about. In fact, it can actually be a motivator (even more so than admiration). In one study of marathon runners, those more prone to “benign envy” (that I want to get better, too feeling) set bigger goals and ran faster.

So a thought: instead of telling our kids “don’t compare yourself” (I am very guilty of this!), maybe we help them acknowledge the envy and get curious: “What are they doing that you’d love to do? What’s one thing you could practice this week?”

2. Make the praise specific.

Apparently, piling on “you’re so amazing!” when our kids doubt themselves can actually backfire. One study found that inflated praise from parents predicted lower self-esteem over time. 😬

Apparently, we can tend to quietly reject things that don’t fit the story we already tell ourselves. Which I guess makes sense… if a kid doesn’t believe they’re amazing at X, hearing it from Mom probably feels like… “well, you’re my mom. you have to say that.”

But what does tend to resonate better is evidence they know is true: “I could tell you really gave it your all that last lap, even when you were exhausted.” Something they’ll recall and give themselves credit for, too.

Well, I rambled a little longer than I’d intended on this one, but I think my point is: we spend a lot of time telling our kids to believe in themselves…. maybe this is part of helping them figure out how. (And figuring it out a little ourselves, too).

The other day, out of exasperation, I did an audit of all the youth sports apps (e.g. for game schedules and communication) that I have on my phone and I kid you not, there were 13 separate ones.

And while I don’t yet have a perfect solve for you (e.g. an agent that can completely “handle” all those sign-ups/registrations, attendance RSVPs, carpool/snack-duty sign-ups, uniform purchases, etc. across and within all those apps), there are definitely ways we can use AI to simplify the chaos.

Here’s one: set up a “Game Day” agent

Most of our family’s games/meets take place on the weekend, and so I set up an agent to message me on Saturday morning with a lot of the need-to-know info about the games that I typically have to hunt through various apps to get: are the games home or away (for uniform color), am I on snack duty, where and when, and any specific notes or instructions from the coaches.

Here is a robust prompt to get you started in ChatGPT Work or Claude Cowork. Customize to suit your family’s needs.

Pro tip: for this to work well, you’ll need to connect your AI to your email, calendar, and Google Drive

Google has built an experimental “Planetary Prediction Engine”, which is quite a name to live up to.

The idea: ask a question about where food shortages, fires, flooding, or disease outbreaks might happen and AI goes off to find all the relevant data, build and test a predictive model, and deliver a report. Work that typically takes specialized teams weeks and even months can, Google says, happen in minutes.

The early examples seem super promising - in tests tracking an Ebola outbreak in the Democratic Republic of the Congo, it correctly identified 15 of the 18 health zones the disease newly spread into.

It’s still early, but if we can get better at figuring out where help will be needed exponentially faster, imagine what we could do with that extra time.

You might have heard about how much the Maldives are struggling with beach erosion due to development and rising sea waters.

Well, an effort coming out of MIT research is using AI to figure out where to place underwater structures on the beaches there to encourage ocean currents to bring sand back to shore. The company says one beach grew by about 90 feet in six months, without pumping or trucking in any sand.

Very promising for my dream of someday vacationing there! And even more so might be the bigger opportunity of giving coastal communities everywhere a chance to protect their shorelines.

To end on a fun note…. spotted by scientists at the bottom of the ocean.