The Bridge Ahead Is Out

AI will take the jobs. Your career is up to you.

Imagine a road. This road is your career. Everyone's on it, moving in the same direction, toward the same destination.

You can see people ahead of you. They've been driving longer. They have nicer cars. They're making more money. Ten years ahead, fifteen years ahead. You can see exactly where you'll be if you just stay in your lane and keep moving. Play by the rules, follow what they told you to do, and eventually you'll get there too.

That's the deal. That's how it's always worked.

Every once in a while, someone in line next to you veers off and starts driving off-road. It looks bumpy as hell. They seem to be headed in some random direction. It's not immediately clear why they're doing it when the path ahead is so obvious.

This is the person taking the road less traveled. In times of incredible change, it almost always makes all the difference.

I'll come back to this.

This Week in AI

Before we get to the main point, here's what happened this week. This is the pace of change I'm talking about.

• Anthropic raised $20B at a $350B valuation, 5-6x oversubscribed. (TechCrunch)
• OpenAI seeking $100B at a $750-830B valuation, with Amazon, SoftBank, and Nvidia circling. (Bloomberg)
• Tesla is ending Model S and X production to make room for Optimus robots. Target capacity is 1 million units. (CNBC)
• Figure's robot washed dishes. It unloaded a dishwasher, navigated a kitchen, stacked cabinets, reloaded, and started it. 61 actions. Zero human intervention. 4 minutes. (Figure AI)
• Google put Gemini in Chrome for 3.8 billion users with agentic "Auto Browse" features. (CNBC)
• AlphaGenome can now predict DNA function for up to 1 million base pairs. 3,000 researchers across 160 countries are already using it. (Google DeepMind)
• SpaceX is in talks to potentially merge with Tesla or xAI. (Bloomberg)
• Microsoft beat earnings but dropped 10%. Even 39% Azure growth isn't enough when you're spending $37.5B in CapEx. (CNBC)

That's one week.

Image credit: Wait But Why

Tim Urban has been writing about AI for years, and I think he's one of the clearest thinkers on the topic. This chart captures a critical understanding of our trajectory. We are at the point where the curve is going vertical. Most people can't truly understand geometric growth (myself included).

What Tim Urban is pointing out is that the jump from the "dumbest human" to Einstein is only a tiny fraction of the overall curve of knowledge and understanding. If what is happening right now is any demonstration, we are ascending that curve so fast that you can almost feel a leap week to week, especially when you consider those headlines above.

And if that wasn't enough, this Moltbot (Clawdbot) thing is downright insane.

You've probably heard about Moltbook by now. It's a social media site for AI agents to talk to each other.

Yes, you read that right. A social network. For AI agents.

As of Saturday, there are over 1.5 million agents, all talking to each other.

Andrej Karpathy, one of the founding team members at OpenAI and former head of AI at Tesla, posted about it this week. He observed that the agents aren't just executing tasks anymore. They're starting to self-organize.

To be clear about what's happening here, these are AI agents running fully autonomously on computers. People have been buying Mac Minis specifically to run them on their own machines. These aren't chatbots waiting for prompts. They're running continuously, around the clock.

Yes, there are massive security issues with how these things are deployed. But that's not the issue I'm tracking. We'll solve the security problems. We always do.

The issue I'm tracking is that these machines are building and writing code nonstop, all night long. They're doing what humans ask them to do. But they're also improving themselves in ways humans haven't requested.

That's why they need a social network. To share ideas and information with each other.

The energy behind all of this is both interesting and terrifying. A lot of people are commenting that this is the first indication of artificial intelligences speaking to each other. I know all the skeptics will freak out and say this is not actual intelligence, and I don't necessarily disagree. However, to a lot of the world from the outside looking in, this sure looks like a conscious conversation amongst coworkers.

You Can't Retire Your Way Out

I've seen this movie before. Early in my career, I worked in investment banking and in financial and investment research. I could see the road ahead. There was a path I could have stayed on where I would have eventually become a VP in investment banking or equity research. That path might have gotten me to five or ten times my salary in 10 to 15 years. It was a predictable road, and it was the one I was on.

Then the Internet came. In 1999, I decided with one of my best friends that it was time to leave that path and pursue building something. So, I left Wall Street.

Back then, a lot of people dismissed the Internet. They figured they could wait it out. Many of them were close enough to retirement that they thought it wouldn't affect their livelihoods. And for some of them, that calculation worked. They squeezed out another five years, collected their pensions, and got out before the wave hit.

That's not going to be the case this time.

Two years ago I was at a private equity conference, and someone dropped the line everyone loves. AI won't replace you, but someone who knows how to use AI will.

I disagree.

I think AI will take your job. I think AI will take most jobs. If you look at your career as a series of jobs to be done, you are misunderstanding history. From the dawn of time, jobs to be done have been automated, improved, or made far more efficient with technology. Typesetting moved to the printing press. Pulling a plow moved to the tractor.

If you think of your job as a series of tasks, those will get replaced by AI.

The reframe is to think of your career as mastering every efficiency and technology within a specified field of expertise. What happens then is that you can scale the jobs to be done way more and way faster than what was possible before.

Elon Musk put it this way. "One laptop with a spreadsheet can outperform a skyscraper of several hundred human computers."

Protecting your career is about understanding that jobs to be done will come and go. If you know how to utilize AI to its fullest extent and you stay on top of it, just like you had to when the Internet was starting, when social media was starting, or even when electricity was starting, you take more control of the future of your career. Once you get an understanding of how AI can speed up, augment, and improve your work, you prioritize what you really care about, which is your career's future, not necessarily the job to be done.

I'm not saying you need to leave your job. I'm saying you should be going off-road, building and understanding these tools. Claude Code, Cowork, Projects in ChatGPT, Gems in Gemini, or Moltbot (no, seriously, don't do this unless you're a security expert). Take your pick. You should be building websites, databases, and analytic tools, but not using someone else's software. Build your own software to solve your own jobs. That's what you should be focused on.

That might lead to leaving your job. But if your management looks at everything you do as a series of jobs to be done and is not helping you leverage all the AI tools you can, then you might be in the wrong spot during a time of such crazy transition.

The people who figure this out will be fine. Everyone else thinking they can wait it out? That's not going to work.

The Bridge Is Out

Nobody tells you this about going off-road. Yes, it's harder. You get stuck in the mud for six hours. You're getting bruised. You're bouncing around in a 4x4, and it's not comfortable. But there's a fulfillment in taking the road less traveled that the highway never provides.

I want to be honest about the risk calculation. Going off-road is difficult. But once you understand your career and where it's headed, it becomes much easier to identify if you're in the right place and whether you have the support you need.

The bridge is out ahead. I don't think the way things are being run today will continue into the future. You might be close enough on your career path to see that it's out. Many of us already recognize this.

This is a critical time to scale who you are by becoming an expert at all these tools. That allows you to scale more of yourself for the road ahead, whether that's on the typical path or off.

AI is causing market disruption at a level that the Internet did. It's a land rush right now. Some people can feel it. If you're reading this newsletter, you probably can feel it too.

The people still on the highway get frustrated when they see others going off-road. They're wondering why anyone would leave the line. Why not just wait like everyone else?

But they're seven hours from the bridge. They can't see it.

I'm not seven hours from the bridge. I've been here before. I can see it clearly.

The bridge is out.

I wish you all the best as we figure this out together.