Still mulling over the Block layoffs, and one thing that caught my eye was how many folks who were let go were there for easily 6, 8, even 10 years or more. People who gave literally a decade of their life, jettisoned in an instant.

In addition to the cruelty of it, there’s is NO WAY that you lose that much institutional knowledge and keep rolling along anywhere close to full operational efficiency, maybe not even 50%. There is a 0.0% chance that your fancy AI tools have that kind of context. And they certainly don’t have the collective centuries of technical, product, or design judgement of those people.

At the risk of sounding corny, as a grizzled (but experienced) old dude, I can tell you with great confidence that your values – what you believe way down deep – is the best thing to help you navigate your career choices. Over the long run it’s what’ll make you successful.

Definitions of success may vary wildly and it’s easier said than done, but if you can consistently align your values with your work, you’re already mostly there.

There are people in 2026 talking about how many lines of code they shipped with a straight face. Well, I think they’re people at least. 😑

Just read an AI hype post saying their vibe coding got them 90% complete on a project. Woo hoo! The “only” things they had to engage an engineer to finish off: authentication, API keys lockdown, VPN protection, and a full security audit.

Funny how all these over the top “AI is changing everything!” posts are largely from people running AI startups. Totally not suspicious at all and very believable.

Emily Gorcenski on Bluesky:

“But guess what—they all require you to still be a good software developer. Engineering has always been about tradeoffs and judgment and I have seen nothing that indicates that this will change.”

“You hear all these stories of people vibe coding a SaaS alternative in a weekend. That’s great for N_users=1. That’s not the hard part.”

This, this, this, this. 💯

I’m well past just muting folks with AI slop takes (“3 engineers can outperform 8 with just AI, here’s how!”). Full blast ridiculing and blocking now. ✌️

💯 You Can’t Prompt Your Way to Judgment by Dr. Claire Knight is a good one! A few nuggets of wisdom from the post:

But most of the software world isn’t Stripe. Some companies still have six-month release cycles. Some are still in the middle of their cloud migration. Panic-adopting AI in those environments doesn’t produce 4x productivity. It produces faster slop.

AI amplifies your existing engineering culture.

Experience debt is what accumulates when you’ve never been through a catastrophic release, never owned a 3am system failure, never made an architectural decision that seemed brilliant and then cost you eighteen months of pain to undo. It’s the absence of pattern recognition that only comes from proximity to failure."

This conversation on AI with Anil Dash is so good. I don’t know how anyone could argue with what he’s saying — measured, realistic, and well reasoned throughout. Y’all need to give it a listen.

Kind of hilarious to see a company make a big deal about how rigorous they are about hiring, but also have the same role open for literally a year in a market flooded with candidates. That’s not rigor, that’s just being bad at hiring.