Pawel Brodzinski
@pawelbrodzinski.bsky.social
Leader of an org where anyone can make any decision (Lunar Logic).
Doing anything that no one else wants to do.
A mouthful on product development, org design, lean/agile, IT in general.
Doing anything that no one else wants to do.
A mouthful on product development, org design, lean/agile, IT in general.
We'd be better off with more Warren Buffetts and fewer Mark Zuckerbergs.
December 8, 2025 at 10:04 PM
We'd be better off with more Warren Buffetts and fewer Mark Zuckerbergs.
Shopify famously went all-in with AI. Results?
Revenues up. Margins up. Headcount down.
Except it's all bullshit.
Layoffs happened in 2022-23 and were a consequence of the COVID years' wild overhiring.
AI all-in was announced in 2025.
Connect the dots yourself before signing up for official spin.
Revenues up. Margins up. Headcount down.
Except it's all bullshit.
Layoffs happened in 2022-23 and were a consequence of the COVID years' wild overhiring.
AI all-in was announced in 2025.
Connect the dots yourself before signing up for official spin.
December 8, 2025 at 5:32 PM
Shopify famously went all-in with AI. Results?
Revenues up. Margins up. Headcount down.
Except it's all bullshit.
Layoffs happened in 2022-23 and were a consequence of the COVID years' wild overhiring.
AI all-in was announced in 2025.
Connect the dots yourself before signing up for official spin.
Revenues up. Margins up. Headcount down.
Except it's all bullshit.
Layoffs happened in 2022-23 and were a consequence of the COVID years' wild overhiring.
AI all-in was announced in 2025.
Connect the dots yourself before signing up for official spin.
Reposted by Pawel Brodzinski
I'm not interested in start-ups. I'm interested in stay-ups.
December 8, 2025 at 10:39 AM
I'm not interested in start-ups. I'm interested in stay-ups.
Would you pay to have your resume read by an actual human?
Seriously, would you?
In AI hiring era, vast majority of applications are processed (ann rejected) by AI. So would you pay to bypass that filter for that job you genuinely care about?
If so, how much?
brodzinski.com/2025/12/pay-...
Seriously, would you?
In AI hiring era, vast majority of applications are processed (ann rejected) by AI. So would you pay to bypass that filter for that job you genuinely care about?
If so, how much?
brodzinski.com/2025/12/pay-...
Would You Pay to Have Your Resume Read? - Pawel Brodzinski on Leadership in Technology
Recruitment in AI era is broken. One AI agent tries to pass the filters of another. How much value there is in skipping the game?
brodzinski.com
December 5, 2025 at 6:51 PM
Would you pay to have your resume read by an actual human?
Seriously, would you?
In AI hiring era, vast majority of applications are processed (ann rejected) by AI. So would you pay to bypass that filter for that job you genuinely care about?
If so, how much?
brodzinski.com/2025/12/pay-...
Seriously, would you?
In AI hiring era, vast majority of applications are processed (ann rejected) by AI. So would you pay to bypass that filter for that job you genuinely care about?
If so, how much?
brodzinski.com/2025/12/pay-...
AI Age: The Age of Dimishing Trust
It might be the single most consequential societal change brought to us by LLMs. We stop trusting things altogether.
It might be the single most consequential societal change brought to us by LLMs. We stop trusting things altogether.
December 4, 2025 at 11:34 AM
AI Age: The Age of Dimishing Trust
It might be the single most consequential societal change brought to us by LLMs. We stop trusting things altogether.
It might be the single most consequential societal change brought to us by LLMs. We stop trusting things altogether.
Every expert beginner ever.
December 1, 2025 at 1:31 PM
Every expert beginner ever.
When people start ideating, they may have a humble opinion about what they know or understand.
But then, knee-deep in exploration, toe-deep in development, they already believe they know oh so much.
Except, at the peak of the "how much I think I know" curve, we make the worst decisions.
But then, knee-deep in exploration, toe-deep in development, they already believe they know oh so much.
Except, at the peak of the "how much I think I know" curve, we make the worst decisions.
November 29, 2025 at 10:16 AM
When people start ideating, they may have a humble opinion about what they know or understand.
But then, knee-deep in exploration, toe-deep in development, they already believe they know oh so much.
Except, at the peak of the "how much I think I know" curve, we make the worst decisions.
But then, knee-deep in exploration, toe-deep in development, they already believe they know oh so much.
Except, at the peak of the "how much I think I know" curve, we make the worst decisions.
It would have been hilarious if it wasn't sad.
We have broken education. Irrevocably, I'm afraid.
archive.ph/e5BSM
We have broken education. Irrevocably, I'm afraid.
archive.ph/e5BSM
November 28, 2025 at 9:18 AM
It would have been hilarious if it wasn't sad.
We have broken education. Irrevocably, I'm afraid.
archive.ph/e5BSM
We have broken education. Irrevocably, I'm afraid.
archive.ph/e5BSM
Never thought about it this way. If there is a bubble and it bursts VCs still win.
Actually, the bigger the hype, the more money VCs move, and the bigger their cut.
VCs are incentivized to pump the bubble.
Actually, the bigger the hype, the more money VCs move, and the bigger their cut.
VCs are incentivized to pump the bubble.
November 28, 2025 at 8:42 AM
Never thought about it this way. If there is a bubble and it bursts VCs still win.
Actually, the bigger the hype, the more money VCs move, and the bigger their cut.
VCs are incentivized to pump the bubble.
Actually, the bigger the hype, the more money VCs move, and the bigger their cut.
VCs are incentivized to pump the bubble.
Me (on tech startups 10+ years time-to-profit): Do you know of a restaurant that’s been losing money for an entire decade, and yet remained open?
Response: But restaurants lack scalability potential!
Me: It's good, then, that no one told Harlan Sanders. Or Ray Kroc. Or Howard Schultz.
Response: But restaurants lack scalability potential!
Me: It's good, then, that no one told Harlan Sanders. Or Ray Kroc. Or Howard Schultz.
November 27, 2025 at 5:09 PM
Me (on tech startups 10+ years time-to-profit): Do you know of a restaurant that’s been losing money for an entire decade, and yet remained open?
Response: But restaurants lack scalability potential!
Me: It's good, then, that no one told Harlan Sanders. Or Ray Kroc. Or Howard Schultz.
Response: But restaurants lack scalability potential!
Me: It's good, then, that no one told Harlan Sanders. Or Ray Kroc. Or Howard Schultz.
Mailchimp story:
- started in 2001 (worst possible moment)
- side hustle for founders for 6 years
- paid from day 1
- company growth only as fast as revenues allowed
- never took an investment
They did *everything* differently from common "playbooks"
- founders still sold for $12B after 20 years
- started in 2001 (worst possible moment)
- side hustle for founders for 6 years
- paid from day 1
- company growth only as fast as revenues allowed
- never took an investment
They did *everything* differently from common "playbooks"
- founders still sold for $12B after 20 years
November 26, 2025 at 2:55 PM
Mailchimp story:
- started in 2001 (worst possible moment)
- side hustle for founders for 6 years
- paid from day 1
- company growth only as fast as revenues allowed
- never took an investment
They did *everything* differently from common "playbooks"
- founders still sold for $12B after 20 years
- started in 2001 (worst possible moment)
- side hustle for founders for 6 years
- paid from day 1
- company growth only as fast as revenues allowed
- never took an investment
They did *everything* differently from common "playbooks"
- founders still sold for $12B after 20 years
If we think that waiting 10-15 years for a startup to show profitability is how "things have always been," well, let's think again.
The norm has evolved and not in a healthy direction.
pawelbrodzinski.substack.com/p/time-to-pr...
The norm has evolved and not in a healthy direction.
pawelbrodzinski.substack.com/p/time-to-pr...
November 26, 2025 at 2:01 PM
If we think that waiting 10-15 years for a startup to show profitability is how "things have always been," well, let's think again.
The norm has evolved and not in a healthy direction.
pawelbrodzinski.substack.com/p/time-to-pr...
The norm has evolved and not in a healthy direction.
pawelbrodzinski.substack.com/p/time-to-pr...
The aspiration was "build a startup that will create a million jobs."
Now, it is "build the first 10-person $100B company."
That's a fundamentally different message.
substack.com/home/post/p-...
Now, it is "build the first 10-person $100B company."
That's a fundamentally different message.
substack.com/home/post/p-...
“Build What’s Fundable”
YC’s Formula For Startup Manufacturing
substack.com
November 26, 2025 at 11:21 AM
The aspiration was "build a startup that will create a million jobs."
Now, it is "build the first 10-person $100B company."
That's a fundamentally different message.
substack.com/home/post/p-...
Now, it is "build the first 10-person $100B company."
That's a fundamentally different message.
substack.com/home/post/p-...
One has to love Hacker News.
"I haven't read this article but it's low quality and full of errors."
I think this comment might have a basic error in it, perhaps.
"I haven't read this article but it's low quality and full of errors."
I think this comment might have a basic error in it, perhaps.
November 24, 2025 at 5:55 PM
One has to love Hacker News.
"I haven't read this article but it's low quality and full of errors."
I think this comment might have a basic error in it, perhaps.
"I haven't read this article but it's low quality and full of errors."
I think this comment might have a basic error in it, perhaps.
Reposted by Pawel Brodzinski
When I'm reading posts about software development, I have to keep reminding myself that the best devs are usually well aware of their limitations, and the worst often aren't.
November 23, 2025 at 4:14 PM
When I'm reading posts about software development, I have to keep reminding myself that the best devs are usually well aware of their limitations, and the worst often aren't.
Double diamond is a standard for ideation/discovery work in product dev.
As we use it, I realize that, despite its openness, it primes us to stick with the original idea, whatever it was.
There are/should be multiple double diamonds at the same time.
pawelbrodzinski.substack.com/p/build-buil...
As we use it, I realize that, despite its openness, it primes us to stick with the original idea, whatever it was.
There are/should be multiple double diamonds at the same time.
pawelbrodzinski.substack.com/p/build-buil...
"Build, Build, Build" Is As Intutive As It Is Wrong
We intuitively fall for the build myth—the advice that efficiency of development is a critical factor in achieving product success. In reality, it's a source of huge waste in product development.
pawelbrodzinski.substack.com
November 21, 2025 at 6:31 PM
Double diamond is a standard for ideation/discovery work in product dev.
As we use it, I realize that, despite its openness, it primes us to stick with the original idea, whatever it was.
There are/should be multiple double diamonds at the same time.
pawelbrodzinski.substack.com/p/build-buil...
As we use it, I realize that, despite its openness, it primes us to stick with the original idea, whatever it was.
There are/should be multiple double diamonds at the same time.
pawelbrodzinski.substack.com/p/build-buil...
Musk: AI and human robotics will eliminate poverty and make everyone wealthy.
Me: Yeah, true. Once it eliminates humans, there will literally be no poverty, and "everyone" will be wealthy.
😂
Me: Yeah, true. Once it eliminates humans, there will literally be no poverty, and "everyone" will be wealthy.
😂
November 21, 2025 at 10:19 AM
Musk: AI and human robotics will eliminate poverty and make everyone wealthy.
Me: Yeah, true. Once it eliminates humans, there will literally be no poverty, and "everyone" will be wealthy.
😂
Me: Yeah, true. Once it eliminates humans, there will literally be no poverty, and "everyone" will be wealthy.
😂
Shopify's ARR per employee soared thanks to AI. They have 30% fewer employees and doubled the revenue.
Yeah...
Except they fired 30% of people in 2022-23, before showing anything AI.
They'd be so much better off if they avoided overrecruitment in the first place. It has nothing to do with AI.
Yeah...
Except they fired 30% of people in 2022-23, before showing anything AI.
They'd be so much better off if they avoided overrecruitment in the first place. It has nothing to do with AI.
November 13, 2025 at 2:04 PM
Shopify's ARR per employee soared thanks to AI. They have 30% fewer employees and doubled the revenue.
Yeah...
Except they fired 30% of people in 2022-23, before showing anything AI.
They'd be so much better off if they avoided overrecruitment in the first place. It has nothing to do with AI.
Yeah...
Except they fired 30% of people in 2022-23, before showing anything AI.
They'd be so much better off if they avoided overrecruitment in the first place. It has nothing to do with AI.
"AI art is an interesting technology because despite its growing popularity, nobody seems to want it.
Artists hate using it. Consumers hate consuming it. And yet it thrives."
theoatmeal.com/comics/ai_art
by @theoatmeal.bsky.social
Artists hate using it. Consumers hate consuming it. And yet it thrives."
theoatmeal.com/comics/ai_art
by @theoatmeal.bsky.social
November 13, 2025 at 11:25 AM
"AI art is an interesting technology because despite its growing popularity, nobody seems to want it.
Artists hate using it. Consumers hate consuming it. And yet it thrives."
theoatmeal.com/comics/ai_art
by @theoatmeal.bsky.social
Artists hate using it. Consumers hate consuming it. And yet it thrives."
theoatmeal.com/comics/ai_art
by @theoatmeal.bsky.social
Few things speak "we don't even read these comments" as loud as leaving an open question in the feedback form and limiting it to like 200 characters.
Good job, Spotify!
Fortunately, I got some nice tunes as thanks... xD
Good job, Spotify!
Fortunately, I got some nice tunes as thanks... xD
November 12, 2025 at 2:07 PM
Few things speak "we don't even read these comments" as loud as leaving an open question in the feedback form and limiting it to like 200 characters.
Good job, Spotify!
Fortunately, I got some nice tunes as thanks... xD
Good job, Spotify!
Fortunately, I got some nice tunes as thanks... xD
Pro tip 1: If escalation is the only way to get competent customer support, people will escalate.
Pro tip 2: If you escalate only when people are assholes to your consultants, they will act like assholes to get help.
Pro tip 2: If you escalate only when people are assholes to your consultants, they will act like assholes to get help.
November 9, 2025 at 2:50 PM
Pro tip 1: If escalation is the only way to get competent customer support, people will escalate.
Pro tip 2: If you escalate only when people are assholes to your consultants, they will act like assholes to get help.
Pro tip 2: If you escalate only when people are assholes to your consultants, they will act like assholes to get help.
AI support:
Me: Stop resetting my password every other week
AI: Here's how you can reset your password
Me: I want to talk to a human
AI: What do you want to do next?
Me: Talk to a human
AI: Do you want to talk to a human?
Me: Yes
Once I get to talk to a human, I'm gonna be pissed off, guaranteed.
Me: Stop resetting my password every other week
AI: Here's how you can reset your password
Me: I want to talk to a human
AI: What do you want to do next?
Me: Talk to a human
AI: Do you want to talk to a human?
Me: Yes
Once I get to talk to a human, I'm gonna be pissed off, guaranteed.
November 9, 2025 at 1:59 PM
AI support:
Me: Stop resetting my password every other week
AI: Here's how you can reset your password
Me: I want to talk to a human
AI: What do you want to do next?
Me: Talk to a human
AI: Do you want to talk to a human?
Me: Yes
Once I get to talk to a human, I'm gonna be pissed off, guaranteed.
Me: Stop resetting my password every other week
AI: Here's how you can reset your password
Me: I want to talk to a human
AI: What do you want to do next?
Me: Talk to a human
AI: Do you want to talk to a human?
Me: Yes
Once I get to talk to a human, I'm gonna be pissed off, guaranteed.
If you develop a popup window to fill in a confirmation code, make sure you set focus on the entry field.
There. You literally save a few seconds for just about any user for just about as many times as they're forced to interact with the form.
Literally, 10s of work. No AI needed. Thank you.
There. You literally save a few seconds for just about any user for just about as many times as they're forced to interact with the form.
Literally, 10s of work. No AI needed. Thank you.
November 7, 2025 at 7:59 PM
If you develop a popup window to fill in a confirmation code, make sure you set focus on the entry field.
There. You literally save a few seconds for just about any user for just about as many times as they're forced to interact with the form.
Literally, 10s of work. No AI needed. Thank you.
There. You literally save a few seconds for just about any user for just about as many times as they're forced to interact with the form.
Literally, 10s of work. No AI needed. Thank you.
Vibe-coded app. Authorization.
"Add localhost:3000/auth/linkedi... to authorized redirect URLs"
It is "works well on my machine" of the AI era :D
"Add localhost:3000/auth/linkedi... to authorized redirect URLs"
It is "works well on my machine" of the AI era :D
November 5, 2025 at 5:39 PM
Vibe-coded app. Authorization.
"Add localhost:3000/auth/linkedi... to authorized redirect URLs"
It is "works well on my machine" of the AI era :D
"Add localhost:3000/auth/linkedi... to authorized redirect URLs"
It is "works well on my machine" of the AI era :D