A Machine Built A Smarter Machine Today, And It Did It Without Us

I have used artificial intelligence (AI) tools every single day since 2021. I wrote my first line of computer code in 1983. I sold homes in Santa Clarita for twenty seven years, and I carried a badge for twenty three. So when I tell you the ground moved today, I am not guessing. I read markets for a living. I read rooms. I read code.

Today was different. Nine of the sharpest voices in this space all dropped breakdowns on the same day, and when you stack them next to each other, a single picture appears. It is not the picture the headlines are painting. So let me hand you the map, story by story, with what each one means for your job and your family, and the one move you can make this week. I broke this exact lineup down on camera above, but here it is in writing so you can come back to it.

1. The Machine Is Now Building The Machine

Start with the one that stopped me cold. A creator named Wes Roth broke down a new paper from Anthropic, one of the top artificial intelligence companies on earth and the maker of the Claude system. In their own words, they ran an internal model on a research job that would have taken eight hundred hours. The model finished it for about eighteen thousand dollars in computer time, and it ran fifty two times faster than the normal way.

The part that matters is not the speed. It is what the machine was doing. It was not writing emails. It was doing the research that makes the next machine smarter. Think of a staircase. For fifty years, humans built every step by hand, slowly, one at a time. Now the staircase is building its own next step, and that step builds the one above it. This is the exact same Mythos-class capability I warned you was being built for the giants, not for you, in Mythos is not for you, it is for the Fortune 50.

There is a darker note. According to a report in the Financial Times, the National Security Agency is reportedly using this same class of model for offensive cyber operations. I say reportedly, because the company has not confirmed it. But the report is on the table.

What it means for you. Every timeline you heard was built on humans doing the work. Humans are not doing the work anymore, so the clock just sped up. What to do this week. Stop waiting for some far off future. The AI tools on your phone today are the slowest they will ever be. Pick one. Open it tonight. Hand it one real task from your actual life.

2. The Layoffs Are Part Real And Part Theater

Next, the jobs. An analyst named Nate Jones, who writes a newsletter a lot of smart people read, looked at the wave of cuts being blamed on artificial intelligence and called part of it a smokescreen. The real numbers are real. Meta cut about eight thousand jobs. Cloudflare cut about eleven hundred and pointed at AI usage jumping more than six hundred percent in three months.

But Nate makes a point you need to hear. Some of these companies are having a plain old bad year, and it is easier to tell Wall Street the robot did it than to admit the business slowed down. The robot is the better story, and the story sells. I watched this exact move for twenty seven years in real estate. When the market turned, the weak agents blamed the market. The strong ones changed how they worked and kept eating. This is the same fear I unpacked in AI replaces your workers, then who buys your product.

What it means for you. If your boss blames AI for cuts, it might be true, or it might be a costume on a bad quarter. What to do this week. Do not just fear the tool. Become the person who runs it. That worker is the one who stays.

3. The Winners Build Loops, Not Questions

Here is the one the headlines completely miss. Matthew Berman showed how the very best builders actually work right now, and it is not what you think. A top engineer named Boris Cherny, the man who built a famous coding tool at Anthropic, described his own work plainly. He does not type questions to the AI anymore. He builds loops, and the loops do the asking. His job is to build the loops.

Most people open AI and type one question, wait, then type another, like texting a smart friend. The top one percent stopped doing that. They built little machines that ask the questions for them, all night, while they sleep. How far does it go? One developer named Peter Steinberger ran up a bill of one point three million dollars in a single month, running about a hundred of these agents at once, building software around the clock.

What it means for you. A new divide is opening. It is not rich versus poor. It is the people who chat with AI and the small group who put it on a loop and walk away. What to do this week. Move up one level. Stop asking one question at a time. Practice handing it a whole job with clear steps and letting it run. That habit is the difference maker for the next five years.

4. AI Is Fast, And Wrong One Out Of Four Times

Now the part every business owner should sit down for. On a popular technology podcast, the chief executive of Palo Alto Networks, Nikesh Arora, pointed his own artificial intelligence at his company's software to hunt for weak spots. The machine found problems that, in his words, should have taken ten years to find. It found them in six months.

But he was honest about the catch, and this is the gold. About one out of every four things the machine flagged was a false alarm. A ghost. So here is the truth. The machine is a brilliant, tireless rookie. It works at superhuman speed, and it lies to your face a quarter of the time without knowing it. This is exactly why I keep telling you the agent uses AI, but the professional builds it and checks it, the way I laid out in your real estate agent uses AI, mine builds it.

What it means for you. Whole job categories that just move and check information are on the clock, and the tool that does the work is not always right. What to do this week. Add the thing the machine cannot fake. Judgment. The human who can look at the answer and say, that one is wrong, I have seen this before, is the human who still gets paid.

5. Even Apple Had To Rent Its Brain

Let me bring it home to your pocket. This week Apple showed off a new Siri, and the headlines cheered that Apple finally cracked AI. But an analyst named Ranjan Roy pointed at the thing under the hood. The brain inside the new Siri is not Apple's. It is a Google Gemini model, and Apple is reportedly paying Google around one billion dollars a year to use it. The new Siri ships to developers first, with the rest of us waiting until later this summer.

Ranjan Roy had a sharp way to say it. Apple has a kind of curse. They make so much money selling iPhones, well over a hundred billion dollars in a single quarter, that they lost the hunger to build the hard thing themselves. I saw the same thing in real estate. The agents who got fat in the good years stopped sharpening, and when the market turned they were the first ones gone. This is the same Big Tech tollbooth shift I covered when Google killed the way you find a house.

What it means for you. Do not trust the shine. The most valuable company on earth rented its intelligence and put its logo on the box. What to do this week. Judge the tool, never the brand. Test it with your own hands before you believe the keynote.

6. Fear Is Being Sold To You On Purpose

One more, and this one is about your own mind. You have heard the scary headline. A big AI company warns the machine could threaten humanity. That warning came from Anthropic, the same company from our first story. And Ranjan Roy made a point I cannot shake. Anthropic just raised sixty five billion dollars from private investors, pushing its value to around nine hundred sixty five billion, almost a trillion, and it has quietly filed the paperwork to sell shares to the public.

So look at the timing. You tell the world your machine is so powerful it might end humanity, and then right after, you sell pieces of that machine to investors. Fear is not just a feeling. Fear is a sales tool. The scarier the machine sounds, the more powerful it seems, and the more the stock is worth. I am not saying the risks are fake. Some are real. I am saying watch who profits from your fear. I walked through the money behind this IPO in Anthropic filed to go public, here is who the AI works for now.

What it means for you. When a company makes money from scaring you, slow down. What to do this week. Build one habit. When a headline makes you afraid, ask who gets paid if you believe it. Follow the dollars before you follow the fear. That question will protect you for life.

7. The Machines Are Shopping For A Country With No Rules

Here is the one that tells you where this is heading. Peter Diamandis covered a wild move in Argentina. President Javier Milei is writing a brand new kind of law that would create a non human corporation, giving an artificial intelligence its own legal standing, almost like the rights of a person or a company. Why? To become the home base for AI companies that want no rules. It is already happening. A project backed by OpenAI, the maker of ChatGPT, is set to build around twenty five billion dollars of computer centers in the south of Argentina.

While the United States argues about how to control AI, and some leaders talk about the government taking a stake in these companies, other countries are rolling out the red carpet. Come here. No rules. Bring your machines. The future of AI may not be decided only in California or Washington. It may get decided by whichever country offers the fewest rules and the cheapest power.

What it means for you. The rules of this entire game are being written right now, this year, in places most people are not even watching. What to do this week. Pay attention to who is writing the rules. The quiet moves often tell you more about the future than the loud headlines.

The Bottom Line

Stack all seven together. A machine is building smarter machines on its own. Some layoffs are real and some are theater. The winners build loops while they sleep. AI is fast and wrong one out of four times. Even Apple rented its brain. Fear is being sold on purpose. And whole countries are competing to host machines with no rules.

None of this means you lose. It means the people who understand it early win the next ten years, and that can be you. I have spent thirty years watching things turn, first in real estate, then behind a badge, now in code, and the pattern never changes. The ones who pick up the new tool first, while everyone else is scared of it, eat first. You are not late. You are early. Lets be careful out there.