Sasa Sekulic
@sasasekulic.bsky.social
CTO & CPO @ Develhope (edtech startup in Italy), backend and mobile developer with 30 years of experience. Not that much of a nerd. Ex-Deutsche Bahn, ex-United Nations, ex-Telecom Italia, ex- a lot of things...
Reposted by Sasa Sekulic
@ivanmorgillo.bsky.social and @sasasekulic.bsky.social's talk was beautiful, a wonderful chat with the audience about the use of AI for developers, friend of for? 💚
November 19, 2025 at 5:09 PM
@ivanmorgillo.bsky.social and @sasasekulic.bsky.social's talk was beautiful, a wonderful chat with the audience about the use of AI for developers, friend of for? 💚
Kudos to @benkadel for making the most interesting presentation on admittedly not very interesting topic of regex in #kotlin and #Android #droidconIT2025
November 19, 2025 at 9:55 AM
Kudos to @benkadel for making the most interesting presentation on admittedly not very interesting topic of regex in #kotlin and #Android #droidconIT2025
I'm somewhat disappointed I couldn't find any of the newer @scalzi.com books at the Dussman bookstore in Berlin, but still I managed to get something new to me - Starter villain! Also, this shelf is very different to how it looked like ten years ago, and only a few authors were in there 20 years ago
September 25, 2025 at 9:32 PM
I'm somewhat disappointed I couldn't find any of the newer @scalzi.com books at the Dussman bookstore in Berlin, but still I managed to get something new to me - Starter villain! Also, this shelf is very different to how it looked like ten years ago, and only a few authors were in there 20 years ago
Reposted by Sasa Sekulic
I am (self-evidently) very interested in language. And also in technology. From that perspective, the rise of LLMs has created many new things for me to think about.
One is: why is it that certain very small clusters of words are *clearly written by an LLM*? What is the quality of that writing?
One is: why is it that certain very small clusters of words are *clearly written by an LLM*? What is the quality of that writing?
September 16, 2025 at 7:32 AM
I am (self-evidently) very interested in language. And also in technology. From that perspective, the rise of LLMs has created many new things for me to think about.
One is: why is it that certain very small clusters of words are *clearly written by an LLM*? What is the quality of that writing?
One is: why is it that certain very small clusters of words are *clearly written by an LLM*? What is the quality of that writing?
Reposted by Sasa Sekulic
you don’t understand bro, if we didn’t lie, no one would buy our product 😩
September 16, 2025 at 3:31 PM
you don’t understand bro, if we didn’t lie, no one would buy our product 😩
Reposted by Sasa Sekulic
I keep reading about things like detailed up-front Sprint plans, about "failed" Sprints, about people panicking when a Sprint doesn't go according to plan, etc. There is no agility in that thinking, whatsoever.
1/7
1/7
July 27, 2025 at 7:08 PM
I keep reading about things like detailed up-front Sprint plans, about "failed" Sprints, about people panicking when a Sprint doesn't go according to plan, etc. There is no agility in that thinking, whatsoever.
1/7
1/7
So, last Sunday my tuxedo cat, Tom, thought to check if he knows how to fly. He does not. He was in the window of my second floor apartment, got spooked by a curtain and jumped down. 🤦All in all it could've been worse - a broken elbow and a perforated palate. Now for two months of meowing. 😅 #cats
July 27, 2025 at 11:37 AM
So, last Sunday my tuxedo cat, Tom, thought to check if he knows how to fly. He does not. He was in the window of my second floor apartment, got spooked by a curtain and jumped down. 🤦All in all it could've been worse - a broken elbow and a perforated palate. Now for two months of meowing. 😅 #cats
Reposted by Sasa Sekulic
Begging someone to go into the Google offices and just spend every day yelling at random managers "remember when your products worked? Remember when they worked so well your company name became a verb? Have you considered doing that instead of shoveling this AI dogshit all day????"
July 23, 2025 at 4:46 PM
Begging someone to go into the Google offices and just spend every day yelling at random managers "remember when your products worked? Remember when they worked so well your company name became a verb? Have you considered doing that instead of shoveling this AI dogshit all day????"
Reposted by Sasa Sekulic
I think it's now possible to make a poli-sci course that equips one for modern political analysis better than most classic theory and has a syllabus sourced entirely from random internet posts.
Text 1. Wilhoit's Law, born as part of a 2018 blog comment
crookedtimber.org/2018/03/21/l...
Text 1. Wilhoit's Law, born as part of a 2018 blog comment
crookedtimber.org/2018/03/21/l...
July 13, 2025 at 1:07 AM
I think it's now possible to make a poli-sci course that equips one for modern political analysis better than most classic theory and has a syllabus sourced entirely from random internet posts.
Text 1. Wilhoit's Law, born as part of a 2018 blog comment
crookedtimber.org/2018/03/21/l...
Text 1. Wilhoit's Law, born as part of a 2018 blog comment
crookedtimber.org/2018/03/21/l...
Reposted by Sasa Sekulic
There are lots of avenues to challenge any study like this, especially when the sample size is so low (n=16).
But it is 1000% more transparent in its methodology than ANY marketing claim about increased developer efficiency with AI tools. 1/2
But it is 1000% more transparent in its methodology than ANY marketing claim about increased developer efficiency with AI tools. 1/2
We ran a randomized controlled trial to see how much AI coding tools speed up experienced open-source developers.
The results surprised us: Developers thought they were 20% faster with AI tools, but they were actually 19% slower when they had access to AI than when they didn't.
The results surprised us: Developers thought they were 20% faster with AI tools, but they were actually 19% slower when they had access to AI than when they didn't.
July 11, 2025 at 12:42 PM
There are lots of avenues to challenge any study like this, especially when the sample size is so low (n=16).
But it is 1000% more transparent in its methodology than ANY marketing claim about increased developer efficiency with AI tools. 1/2
But it is 1000% more transparent in its methodology than ANY marketing claim about increased developer efficiency with AI tools. 1/2
Reposted by Sasa Sekulic
Reposted by Sasa Sekulic
AI agents get office tasks wrong around 70% of the time, and a lot of them aren't AI at all
AI agents get office tasks wrong around 70% of the time, and a lot of them aren't AI at all
More fiction than science
Feature IT consultancy Gartner predicts that more than 40 percent of agentic AI projects will be cancelled by the end of 2027 due to rising costs, unclear business value, or insufficient risk controls.…
dlvr.it
June 29, 2025 at 11:37 AM
AI agents get office tasks wrong around 70% of the time, and a lot of them aren't AI at all
Reposted by Sasa Sekulic
Often, when people are baffled by why change can't happen, they don't consider incentive systems. Incentive systems take many forms. Here are a couple:
* Individual bonuses. These create a competitive culture within the team, and that typically leads to hoarding skills and information.
1/10
* Individual bonuses. These create a competitive culture within the team, and that typically leads to hoarding skills and information.
1/10
June 21, 2025 at 5:58 PM
Often, when people are baffled by why change can't happen, they don't consider incentive systems. Incentive systems take many forms. Here are a couple:
* Individual bonuses. These create a competitive culture within the team, and that typically leads to hoarding skills and information.
1/10
* Individual bonuses. These create a competitive culture within the team, and that typically leads to hoarding skills and information.
1/10
Reposted by Sasa Sekulic
Reposted by Sasa Sekulic
“Why GUIs are built at least 2.5 times
Or Why I Think Lean Software Development is Wrong”
patricia.no/2025/05/30/w...
Or Why I Think Lean Software Development is Wrong”
patricia.no/2025/05/30/w...
Why GUIs are built at least 2.5 times | Patricia Aas
Or Why I Think Lean Software Development is Wrong
patricia.no
May 30, 2025 at 10:24 PM
“Why GUIs are built at least 2.5 times
Or Why I Think Lean Software Development is Wrong”
patricia.no/2025/05/30/w...
Or Why I Think Lean Software Development is Wrong”
patricia.no/2025/05/30/w...
Reposted by Sasa Sekulic
*Career artisanal programmer eloquently beefing about AI copilots. He probably got fired this week
deplet.ing/the-copilot-...
deplet.ing/the-copilot-...
The Copilot Delusion
Disclaimer: This post was written May 2025, and the arguments apply to AI code capabilities at this time. The arguments around lack of competence are certainly likely to become less prevalent-while th...
deplet.ing
May 25, 2025 at 8:49 AM
*Career artisanal programmer eloquently beefing about AI copilots. He probably got fired this week
deplet.ing/the-copilot-...
deplet.ing/the-copilot-...
Reposted by Sasa Sekulic
Just saw this interview where the question was “how far do you think we are from having AI operate as a junior engineer 24/7?”
I cannot stress enough that I would pay a reasonable fee to 𝘯𝘰𝘵 have an unsupervised junior engineer operating 24/7 in our codebase.
I cannot stress enough that I would pay a reasonable fee to 𝘯𝘰𝘵 have an unsupervised junior engineer operating 24/7 in our codebase.
May 24, 2025 at 1:25 AM
Just saw this interview where the question was “how far do you think we are from having AI operate as a junior engineer 24/7?”
I cannot stress enough that I would pay a reasonable fee to 𝘯𝘰𝘵 have an unsupervised junior engineer operating 24/7 in our codebase.
I cannot stress enough that I would pay a reasonable fee to 𝘯𝘰𝘵 have an unsupervised junior engineer operating 24/7 in our codebase.
Reposted by Sasa Sekulic
Newsletter: We are in the era of the Business Idiot, where middle management has seized power, breeding out true meritocracy and value-creation in favor of symbolic growth and superficial intelligence. This piece ties together everything I've ever written.
www.wheresyoured.at/the-era-of-t...
www.wheresyoured.at/the-era-of-t...
The Era Of The Business Idiot
Fair warning: this is the longest thing I've written on this newsletter. I do apologize.
Soundtrack: EL-P - $4 Vic
Listen to my podcast Better Offline. We have merch.
Last week, Bloomberg profiled ...
www.wheresyoured.at
May 21, 2025 at 4:41 PM
Newsletter: We are in the era of the Business Idiot, where middle management has seized power, breeding out true meritocracy and value-creation in favor of symbolic growth and superficial intelligence. This piece ties together everything I've ever written.
www.wheresyoured.at/the-era-of-t...
www.wheresyoured.at/the-era-of-t...
Reposted by Sasa Sekulic
Physics is learned by solving physics problems.
A guy who "learns" physics by talking to chatgpt for hours a day, can spend an entire year never once solving a single problem, but thinking that he is learning physics.
A freshman on their very first day, has learned more physics than he has.
A guy who "learns" physics by talking to chatgpt for hours a day, can spend an entire year never once solving a single problem, but thinking that he is learning physics.
A freshman on their very first day, has learned more physics than he has.
A year with zero calibration, happily regurgitating nonsense back and forth with the answer-shaped object machine, and you now have a guy who is *ruined* for the next several years. It'll take years of hard effort to fix what he mutilated.
May 19, 2025 at 6:13 PM
Physics is learned by solving physics problems.
A guy who "learns" physics by talking to chatgpt for hours a day, can spend an entire year never once solving a single problem, but thinking that he is learning physics.
A freshman on their very first day, has learned more physics than he has.
A guy who "learns" physics by talking to chatgpt for hours a day, can spend an entire year never once solving a single problem, but thinking that he is learning physics.
A freshman on their very first day, has learned more physics than he has.
Reposted by Sasa Sekulic
Spent too much time on this
May 17, 2025 at 9:32 PM
Spent too much time on this
Reposted by Sasa Sekulic
"Individual developers report productivity gains using these tools.
And at the same time, data shows that the more teams use them, the bigger the negative impact on team outcomes like delivery throughput and release stability.
How can both be true?"
codemanship.wordpress.com/2025/05/11/i...
And at the same time, data shows that the more teams use them, the bigger the negative impact on team outcomes like delivery throughput and release stability.
How can both be true?"
codemanship.wordpress.com/2025/05/11/i...
It’s the System, Stupid!
Since this “Age of A.I.” arrived in late 2022, something’s been nagging at me. As more and more data rolls in, we see an apparent paradox emerging where “A.I.” coding …
codemanship.wordpress.com
May 11, 2025 at 10:41 AM
"Individual developers report productivity gains using these tools.
And at the same time, data shows that the more teams use them, the bigger the negative impact on team outcomes like delivery throughput and release stability.
How can both be true?"
codemanship.wordpress.com/2025/05/11/i...
And at the same time, data shows that the more teams use them, the bigger the negative impact on team outcomes like delivery throughput and release stability.
How can both be true?"
codemanship.wordpress.com/2025/05/11/i...
Reposted by Sasa Sekulic
Did AI kill your job? If so, I want to hear about it.
The response to my story on the AI jobs crisis has been overwhelming. I heard from so many people impacted by AI that it made me want to collect and organize these stories to better understand the phenomenon.
Share yours at: [email protected]
The response to my story on the AI jobs crisis has been overwhelming. I heard from so many people impacted by AI that it made me want to collect and organize these stories to better understand the phenomenon.
Share yours at: [email protected]
Did AI kill your job?
If so, I want to hear about it. Send your story to [email protected]
www.bloodinthemachine.com
May 8, 2025 at 10:18 PM
Did AI kill your job? If so, I want to hear about it.
The response to my story on the AI jobs crisis has been overwhelming. I heard from so many people impacted by AI that it made me want to collect and organize these stories to better understand the phenomenon.
Share yours at: [email protected]
The response to my story on the AI jobs crisis has been overwhelming. I heard from so many people impacted by AI that it made me want to collect and organize these stories to better understand the phenomenon.
Share yours at: [email protected]
