My January note to my kids on AI was easily my most shared column ever. In the past few weeks, several of you have asked what I'd say now, given all the scary warnings about AI ending humanity. Here's what I sent them...

🕰️ You find yourself — not by choice, but circumstance — living in a life-will-never-be-the-same-again historic moment with the advent and explosion of artificial intelligence.

• Yes, people aren't killing each other like in our Revolution or the Civil War. And no, people aren't starving or suffering like during the Great Depression. In fact, the country is exceptionally peaceful and prosperous by historical measures.
• But we've unleashed a technology so powerful, so unknowable and, at this moment, so uncontrollable that warnings of scary scenarios cannot be ignored. The risk rises by the hour.

⏳ Why it matters: Just writing these words feels hyperbolic, even absurd. But having spent the past two years studying, using, and reading and writing about AI, I believe deep in my bones and soul that you're living — and will be part of shaping — history.

💡 The good news: We're early to seeing this new history unfold. And this technology could benefit everyone in unimaginably good ways if managed properly. So the potential to bend it toward good outcomes and society-wide benefit is real and very doable.

• This is no longer something you can ignore or wish away. In the nine months since my last letter, AI has gotten way better and at times scarier. You have a generational responsibility to understand the gravity and urgency of shaping AI to avoid some very bad outcomes for this nation and world.
• And a generational responsibility to rise up, first to understand it, then to demand and fight for ways to steer this technology to benefit and protect us all.

This might feel heavy, frightening, maddeningly unnecessary. You didn't ask for this technology, and most people don't even understand it or find it terribly useful or powerful.

• History doesn't ask permission. It just happens.

⚡ The most important four words in this note: It's not too late.

• AI is very powerful, and getting more powerful by the day. But it's also early enough that government, business, all of us can do what great Americans did during and after the Civil War or the Great Depression: Band together to do smart, hard things to emerge better, stronger.

🧬 This is the genetic genius of America. We're wired, deep in our DNA, to start and solve problems — be it revolutions, wars, economic collapses or new technologies. Like AI.

• Reverse-engineer America's finest moments, and you'll find ordinary people doing extraordinary things during extraordinary times. Let that be you.

Before offering a few steps you can take individually, let me explain why AI is suddenly more powerful and worrisome, so much so that dozens of people building it publicly warn that it could literally end civilization, as crazy as that sounds.

• Most people use AI as a souped-up Google search, or for writing emails, or cheating on exams. But the underlying technology is so mind-bendingly complex that AI creators still don't understand why it does what it does — and what it's truly capable of doing.

What we know: We've poured unfathomable money, data and compute into willing, and rushing, AI with superhuman possibilities into existence.

• Every AI company, many the size of nation-states, is explicitly working to create thinking machines with knowledge surpassing all humans on all topics, with the aim of letting them train themselves.

🚀 The chief aim isn't to methodically tame poverty or cancer. It's to create superhuman intelligence and set it loose on everything. And everything can be curing cancer, or creating and spreading a plague.

• This was aspiration and fantasy until the past six months. These AI companies are blowing past their wildest projections of how quickly they could achieve true superhuman intelligence. It's gone from years or a decade to possibly as soon as next year.
• What does that even mean? They want to achieve something called recursive self-improvement (RSI). That's when AI, which was trained on human knowledge under human guidance, starts to train itself without human restraints.
• Basically, AI builds AI. Remember, the labs that conjure these models don't even know why basic human-guided AI works. Imagine advanced AI building advanced AI. That's when we veer into truly creepy and possibly dangerous territory.

In some ways, we're already there. So what can you do?

1. 📚 Study this more than any other topic. I don't want you obsessing about the dark scenarios. Instead, take time to read, watch and listen to people who understand what AI can do. This is 1,000x more important than any other political, business or cultural topic of your time. You have a civic obligation to understand something that will define your generation. Here is a list of trustworthy sources on the topic.
2. 🧭 Understand motives. This is a hard topic to get your head around because of all the hyperbole and hysteria. On one side are AI optimists promising cancer cures and unfathomable riches for all. On the other are critics warning of mass job loss or humanity's end. In between are politicians woefully uninformed on this complex topic. The truth about AI, like most topics, sits in between hysteria and hallelujah. It's on you to understand the motives before reacting.
3. 📣 Take action. You don't get to sit this one out. You should demand that the people you elect, the companies you support, the products you buy align with the values and views you come to on AI. Don't assume the government or the companies will do the right thing, absent considerable and sustained public pressure. Your voice matters, particularly when you encourage family and friends to do the same. Learn about AI — and then share that knowledge.
4. 🤖 Use it. Resist the temptation to protest AI by refusing to use it. This might feel good or righteous, but you'll never understand it without using it yourself. AI can do extraordinary good, for you and for society. So take time to explore how it works, where you can experiment with it, and what it does well or poorly. I feel confident writing this note and talking about AI because I use it so aggressively, which allows me to better understand and explain it to others. Make yourself the most AI-savvy person doing your specific job. That will make you indispensable — and knowledgeable.
5. 🌱 Stay hopeful. The reason America always bounces back stronger even when times feel bleaker is our innate optimism. Never lose that. Our history shows we've overcome moments much tougher than this, usually in unimaginable ways. We just figure it out. Pessimism is cancer — it spreads silently and perniciously. You hear the word "doomers" a lot. They live in perpetual fear of awful outcomes. Be aware of pitfalls, but don't succumb to doomerism. Choose hope and action.
6. 🌳 Grab hands, touch grass. Prediction: The better AI gets, the more we'll appreciate humanity and nature. Some of our wonder, awe and collective soul got lost to infinite scroll, COVID and political tribalism. Use this moment to restore it. No machine will produce love, sacrifice or joy. No machine will displace a sunrise or mountaintop or river. The best remedy for fears about big things is to relish the small. Go for a walk. Call a friend.

This is a lot. I hope the dark warnings of today are truly overwrought. I hope AI never exceeds our control and understanding. I hope it cures cancer, creates wealth, makes life better for all. But it would be foolish not to pay attention — and get involved.

• It's your time.

Share this letter.

📱 Let Jim know what you think: [email protected].

📈 If you're a CEO or on a CEO's team: Ask to join Jim's new weekly Axios C-Suite newsletter.

Finish Liner Jackie Rick captured this colorful calm over the Atlantic from Harvey Cedars on Long Beach Island last week, ahead of the weekend storms.