The US midterm elections are Tuesday, and AI is all over the campaign trail.

What’s more, there's something called the "liar's dividend:” once anything could be fake, it's easier to dismiss things that are actually real.

So we're left believing things that aren't true and doubting things that are. Yikes.

A few things we can do:

And most importantly, go vote!

Two weeks ago, I shared a Harvard report on AI in schools. Fewer than 1 in 5 teachers work at a school with an actual AI policy.

Meanwhile, the gray areas keep piling up. Is it cheating to have AI quiz you? Fix your grammar? Write a text to a friend?

Good luck writing (and enforcing) a rule for every scenario.

Last year at the French Open, Carlos Alcaraz hit what looked like an incredible winning shot against Ben Shelton. The umpire gave him the point. But Alcaraz knew his racket had left his hand before hitting the ball (which isn’t allowed), so he spoke up and gave the point to Shelton.

That's the kind of judgment I want my boys to have with AI - knowing when to call a foul on yourself, even when nobody's watching.

AI can be an awesome trainer. It can explain tricky concepts, quiz you, help you practice. But then you need to be able to play the game and compete.

And so, beyond the rulebook, let's teach our kids AI sportsmanship by asking:

There’s one night a year when knocking on practically every door in your neighborhood is not only acceptable, but encouraged. Let’s not waste it!

Tomorrow night, resist the urge to treat trick-or-treating like a competitive sport. Linger on the porch, introduce yourself, compliment the neighbor who went all out with decorations. Give your kids a mission: Learn the names of three neighbors you didn't know before.

And for a little Halloween nostalgia, It’s the Great Pumpkin, Charlie Brown turned 60 years old this week! Fun fact: When it first aired, viewers felt so bad that Charlie Brown got rocks instead of candy that they mailed candy to the studio for years.

Which is a pretty sweet reminder of what Halloween can bring out in us. So check in on a neighbor who might be spending the night alone, drop off a little extra candy, or invite a family over to watch the 25-minute classic together.

I don’t know about you but I feel like spam calls are out of control lately. (PSA: opting in to call screening is life-changing).

I saw a post this week about someone who used Grok Bot (one of the personal AI assistant products popular right now) to find everywhere they were listed on telemarketing/people search sites and get them removed. Um, yes please!

So, I tried using my Instinct AI assistant to do the same thing and actually, it was pretty great. Granted, it couldn’t do it all autonomously - a lot of these sites (and there are quite a few) have gotten smart to prevent AI agents from doing just this, but Instinct was great at finding all the instances of where my number/address were published, removing/confirming take-down of what it could, and then spoon-feeding me the steps I needed to take to finish off what was left.

I plan on sitting down with my parents this weekend to do this same exercise with them, and here’s how you can try it too:

To do:

This will work best if you are using an agentic assistant like Instinct, Muse (from Meta), Grok Bot (from Xai), or Dots (from OpenAI). Simply ask it something like:

If you don’t have one of these products, you can also prompt Claude, Gemini, or ChatGPT and ask it to do as much as it can on it’s own and then give you an easy to use checklist with all the sites and exact links/steps to remove from each. A little more manual for you but still great to get the full audit and hand-holding.

P.S. This is also a useful step to get yourself “hard to hack” - the full checklist I shared last week is here, I highly recommend spending a little time going through it and helping your parents do the same - please share it with others as well.

P.P.S. If you’d like to try Instinct (the personal AI agent I’m currently using and recommend (for now!)), here is a link to my invites.

Over the past couple of months, AI (mostly from OpenAI and Anthropic) has made jaw-dropping progress on some of the world’s hardest and longest-standing problems in science, math, and physics, such as one of the $1 million Millennium Prize problems, a question about how water and air flow that had stumped mathematicians for about 90 years.

On the one hand, this shows how fast AI is moving from answering our questions to answering ones no human could - and shocking to see how quickly AI can upend the life’s work of some of the world’s smartest people.

On the other hand, it opens up a future where breakthroughs in medicine, energy, and technology could arrive exponentially sooner. Experts are still checking the work, but either way… buckle up.