Bring the Halftime Speech
Eight months. Zero customers. Zero income. My Story.
There was a stretch in there where I’d open my laptop, scroll LinkedIn for three minutes, and feel the floor drop out from under me. Every other post was about layoffs. Every third headline declared something I cared about dead. Some mornings I’d close the laptop and just sit there for a minute, trying to figure out who was supposed to give me a pep talk.
Nobody was coming. The halftime speech, if there was going to be one, had to come from me.
That memory came back hard this week. Bob just got home from Agile Coach Camp Cologne, and the room he described sounded a lot like the inside of my head during those eight months. Plenty of experienced coaches, plenty of folks one to five years into their career, and the dominant energy in the room was “agile is dead, my job is going away, AI is coming for me, and there is nothing I can do about it.”
Bob said people walked around the camp handing each other Kudo cards, and a number of his thanked him for being a positive inspiration. He came home a little shaken by that. Not because of the cards themselves, but because of what they implied. Inspiration was the rare thing in that room. Optimism was the rare thing.
That’s not just an agile community problem. That’s the air right now.
The doom narrative is loud
Open Twitter, open LinkedIn, open any tech newsletter, and you can’t go ten minutes without somebody declaring software engineering dead. Or product management dead. Or design dead. Or whatever role you happen to do, dead. The headlines have a rhythm to them. They’re written for engagement, not for accuracy, and they work because fear reliably outperforms nuance in an algorithmic feed.
I get why the headlines land. Capital One did clear out a chunk of their agile org. Layoffs are real. AI is real. The job search is brutal in a way I personally remember. None of that is made up. But there’s a difference between “the market is hard right now and real disruptions are happening” and “your career is over and there’s nothing you can do.” The first one is true. The second one is clickbait.
If you actually pull back and look at the data, the picture is almost the opposite of what the doomscroll wants you to believe.
What the data actually shows
Lenny Rachitsky publishes a biannual State of the Product Job Market report, and his early 2026 edition is the most optimistic of the four he’s done.1
Product manager openings are at their highest level in over three years. Globally, there are more than 7,300 open PM positions, up roughly 75 percent from the 2023 lows and up around 20 percent just since the start of 2026. Software engineering openings sit at over 67,000 globally, with 26,000 of those in the United States, and the growth rate is accelerating. AI-specific roles are not just growing, they’re hockey-sticking off the chart.
The a16z team published a piece earlier this month called “The ‘AI Job Apocalypse’ Is a Complete Fantasy” and the data they pulled together deserves a look.2 Software development job postings have been climbing as a count and as a percentage of the overall job market since the beginning of 2025. On corporate earnings calls, executives mention AI as augmentation versus AI as substitution at roughly an eight-to-one ratio.
The Yale Budget Lab put out a tracking study in April that concluded AI’s impact on the labor market right now is one of stability, not major disruption at an economy-wide level.
A National Bureau of Economic Research working paper found AI adoption hasn’t yet led to meaningful changes in total employment.
A Census Bureau working paper found that of the firms using AI, only about 5 percent reported any headcount change at all, and that change was split roughly evenly between hiring more people and hiring fewer.
The summary version is this. The roles are still there. The roles are growing. New roles are showing up that didn’t exist three years ago. And the people screaming loudest about the death of your career haven’t bothered to check the numbers.
None of that erases the layoffs. None of that means the job search is easy. If you’re out of work right now, the macro data doesn’t pay your mortgage and I’m not going to pretend it does. But there is a difference between a hard market and a dead one, and the loudest voices in the feed have been telling you the wrong story.
This pattern has a name
There’s a 19th-century economist named William Stanley Jevons who noticed something strange about the steam engine. Engineers kept making it more efficient at burning coal, and the obvious assumption was that total coal consumption would drop. Less coal is needed for the same work, so less coal is burned. That’s not what happened. Coal use exploded. The efficiency gains made coal cheap enough to use for things nobody had bothered using it for before, and demand outran the savings by a wide margin.
Economists call this Jevons Paradox, and it shows up every time a foundational input gets cheaper. When tractors made farming more efficient, U.S. farm employment fell from roughly a third of the workforce to about two percent, and total agricultural output nearly tripled. The displaced workers didn’t sit at home. They moved into factories, then offices, then jobs that didn’t exist when they were born. When VisiCalc and Excel made bookkeeping more efficient, the country lost about a million bookkeepers and gained roughly a million and a half financial analysts. The work didn’t disappear. It moved up the stack.
This is the part the doom narrative leaves out. The price of cognition is collapsing, and the assumption baked into “AI is coming for your job” is that humans will respond to that by doing less thinking. That has never been what humans do.
When something we want gets cheaper, we want more of it. When the cost of writing software falls, we don’t write less software. We write software for problems we couldn’t justify solving before. We start companies that didn’t make sense at the old price point. We build things that used to be impossible to build. The work expands to absorb the new capacity, and then some.
That’s why the data is showing what it’s showing. Jevons isn’t a theory anymore. He’s old news, repeating himself.
What this means for leaders
The reason this matters for the leaders reading this is twofold. The first is personal. If you’re wrestling with the doomscroll and quietly wondering whether your role, your company, or your function has a future, the data says you have far more runway and far more optionality than the headlines are letting you believe. That’s worth knowing. It changes how you wake up. It changes how you walk into your one-on-ones. It changes whether you’re showing up with energy or showing up with dread.
The second reason is bigger. Your team is reading the same headlines you are. Your engineers are seeing the same “software engineering is dead” posts on Tuesday and the same “AI is coming for your job” thinkpieces on Thursday. They’re absorbing the same fear you’re absorbing. And nobody is going to walk into your standup tomorrow and give them perspective unless that someone is you.
Bob’s observation in Cologne was that the room was drowning, and people were starving for somebody to say something hopeful. Your team is in that room. The question is whether you’re the one handing out the positive Kudo cards or the one sitting at the back agreeing that everything is broken.
The speech itself
I’m not going to write a halftime speech for you. The good ones are personal, and they’re delivered in your own voice. But I’ll tell you what mine sounded like during those eight months, because it’s still close to my chest.
It went something like this. The market is hard, but the market is not forever. The headlines are loud, but the headlines are not the data. This is the hard part everyone talks about. This is where many give up. That won’t be me. Get up. Make the call you’ve been avoiding. Send the message. Open the doc. Do the next thing.
That’s not poetry. It’s not motivational LinkedIn quote material. But it got me from morning one to morning two, and morning two to morning three, until eventually the phone rang and the work came back.
If you’re a leader right now, your team needs to hear some version of that, often, and from you. Not because you have to be a fake cheerleader. Not because you have to pretend the disruption isn’t happening. Because the room is full of negativity, and somebody has to be the optimist in it. That’s the job description nobody puts in the offer letter, and it’s the one that matters most when the air is heavy.
Stay courageous,
Josh Anderson
The Leadership Lighthouse
P.S. This week’s Meta-Cast is Bob’s debrief from Agile Coach Camp Cologne, including the Kudo card moment and a long thread on why coaches and leaders alike have to learn to translate value into business terms before the market does it for them. Worth a listen.
P.P.S. If you’ve got your own halftime speech, I’d love to hear it. Reply to this and tell me what you say to yourself when the doomscroll wins for a minute. The good ones are going to find their way into a future Lighthouse.
