AI is not replacing work evenly. It performs best on structured tasks, but it still struggles with the nuance. The near-term shift is not simply jobs disappearing. It is jobs being redesigned.
If an AI workflow cannot survive a restart, a stale login, a bad source, or a partial failure, it is not production. It is a very impressive demo with a trap door.
Apply to write a guest post for Lenny's Newsletter
The vast majority of my most popular posts are guest posts, and I realized I haven't held an open call for guest posts in years.
What's a powerful lesson, tactic, framework, or way of working that has been extremely helpful to you in your work that others might find useful?
No prior writing experience is required. I will help you edit and refine your writing. The only requirement is that you've built successful product(s) and have something valuable to share with the world. This will become the best thing you've ever written online.
What I’ll look for in the proposals:
1. Is it highly actionable and concrete? Can a reader immediately do something with your advice?
2. Is it well-researched and/or rooted in something you are uniquely knowledgeable about?
3. Does this add something new to the conversation about building successful products, careers, or teams?
Turn-offs:
1. You're clearly trying to promote your product/company
2. High on pontification/theory, low on tactics/specifics
Note: The bar for guest posts is extremely high. Expect a month of iterating/editing to get it to where it needs to be. That being said, your initial pitch doesn't have to be fully formed. We'll build on it together.
Find the link to apply in the thread below.
I'll reach out to a subset of finalists within 2 weeks.
Excited to see what you come up with!
12 executives from TSMC, Quanta, MediaTek, Foxconn, ASUS, Delta, and more sat down ahead of Jensen Huang's GTC Taipei keynote to lay out the AI infrastructure from silicon and racks to agentic factories and physical robots. Conversations cover real roadmaps and real results.
Taiwan helped build the PC era. Discover how they’re building the AI era.
Watch NVIDIA GTC Live Taipei pregame show, presented in Mandarin with English subtitles.
https://t.co/96jy3Vz3yu
This is a great idea. What many folks do not understand is that you can use Open Source on your own servers, or private cloud, disable data retention, ensure outputs, logs, are not shared. This lowers the cost of AI + enables the biz to build proprietary data + expertise.
@Jason@Jason This is a great idea. What many folks do not understand is that you can use Open Source on your own servers, or private cloud, disable data retention, ensure outputs, logs, are not shared. This lowers the cost of AI + enables the biz to build proprietary data + expertise.
End-to-end automation is the hard part. Generating the thing is easy. Publishing it, checking it, recovering from failure, and documenting it is where the real work lives.
It will be difficult to always tell the difference, because somewhere between rarely and often it will require code review to really see the difference.
It's painfully obvious to me, after 12 years of shipping production code, that:
We are massively overestimating AI and massively underestimating it at the same time.
→ 96% of the code I write today is AI-generated
→ but I review every single line like my job depends on it
→ the developers who win won't be the ones who prompt the fastest but be the ones who know what "good" looks like
Here's what nobody wants to admit:
AI didn't make engineering easier.
It made judgment the entire job.
The bottleneck was never typing.
It was knowing what to build, what to throw away, and what will break at 3am six months from now.
Juniors are shipping 10x more code.
And introducing 10x more bugs they can't explain.
The skill isn't writing anymore.
It's reading. Reviewing. Saying no.
Taste is the new 10x.
The engineers who treated coding as typing are panicking.
The ones who treated it as thinking have never been more valuable.
Adapt accordingly.
I keep coming back to this: AI is less like hiring one genius and more like managing a very fast junior team that never sleeps and occasionally loses the plot.