Kevin Goldsmith
@kevingoldsmith.com
This is my account for technology leadership posts and announcements. For personal stuff, https://bsky.app/profile/kmg.bsky.social.
Author of "It Depends: Writing on Technology Leadership 2012-2022"
CTO @ DistroKid
https://kevingoldsmith.com
Author of "It Depends: Writing on Technology Leadership 2012-2022"
CTO @ DistroKid
https://kevingoldsmith.com
I've just published a piece on building a technical strategy that people can use - not a wishlist, not a roadmap, and not a vague mission statement.
Strategy isn’t prediction. It’s clarity. And clarity scales.
open.substack.com/pub/kevingol...
Strategy isn’t prediction. It’s clarity. And clarity scales.
open.substack.com/pub/kevingol...
Crafting a Technical Strategy That Actually Works
Moving from disconnected decisions to a unified technical direction
open.substack.com
November 16, 2025 at 9:43 AM
I've just published a piece on building a technical strategy that people can use - not a wishlist, not a roadmap, and not a vague mission statement.
Strategy isn’t prediction. It’s clarity. And clarity scales.
open.substack.com/pub/kevingol...
Strategy isn’t prediction. It’s clarity. And clarity scales.
open.substack.com/pub/kevingol...
Anyone using this? Interested to hear how useful it is for you. Mozilla.ai releases universal interface to LLMs | InfoWorld share.google/pD8LMsPDGNxc...
November 9, 2025 at 9:58 PM
Anyone using this? Interested to hear how useful it is for you. Mozilla.ai releases universal interface to LLMs | InfoWorld share.google/pD8LMsPDGNxc...
A technical strategy isn’t a roadmap; it’s how you align your team to face the future together.
In this episode of It Depends, I discuss what makes a good strategy, how to document it effectively, and how to keep it alive as your company evolves.
pod.link/1724860326/e...
In this episode of It Depends, I discuss what makes a good strategy, how to document it effectively, and how to keep it alive as your company evolves.
pod.link/1724860326/e...
November 9, 2025 at 9:22 AM
A technical strategy isn’t a roadmap; it’s how you align your team to face the future together.
In this episode of It Depends, I discuss what makes a good strategy, how to document it effectively, and how to keep it alive as your company evolves.
pod.link/1724860326/e...
In this episode of It Depends, I discuss what makes a good strategy, how to document it effectively, and how to keep it alive as your company evolves.
pod.link/1724860326/e...
Every wave of tech promises transformation. Few deliver.
My latest newsletter, "Making Technology Choices That Last," is about leading with intent instead of impulse.
open.substack.com/pub/kevingol...
My latest newsletter, "Making Technology Choices That Last," is about leading with intent instead of impulse.
open.substack.com/pub/kevingol...
November 2, 2025 at 3:54 PM
Every wave of tech promises transformation. Few deliver.
My latest newsletter, "Making Technology Choices That Last," is about leading with intent instead of impulse.
open.substack.com/pub/kevingol...
My latest newsletter, "Making Technology Choices That Last," is about leading with intent instead of impulse.
open.substack.com/pub/kevingol...
The recording from my LeadDev NYC talk "From friction to function: Strengthening the product-engineering alliance" is now live! leaddev.com/communicatio...
From friction to function: Strengthening the product-engineering alliance
Learn practical strategies to turn product-engineering friction into alignment through better prioritization, communication, and shared decision-making.
leaddev.com
November 1, 2025 at 10:17 PM
The recording from my LeadDev NYC talk "From friction to function: Strengthening the product-engineering alliance" is now live! leaddev.com/communicatio...
Reposted by Kevin Goldsmith
Just having some innocent fun with Claude.
October 26, 2025 at 9:19 PM
Just having some innocent fun with Claude.
Every wave of tech promises transformation.
In the latest It Depends episode, I share how leaders can make thoughtful technology choices that last: balancing curiosity with caution, and learning before leaping.
Making Technology Choices That Last
itdependspod.com/episodes/mak...
In the latest It Depends episode, I share how leaders can make thoughtful technology choices that last: balancing curiosity with caution, and learning before leaping.
Making Technology Choices That Last
itdependspod.com/episodes/mak...
Making Technology Choices That Last - It Depends Podcast
Avoiding shiny object syndrome, encouraging curiosity, and building a system for smart adoption
Every few years, a new wave of technology arrives promising to change everything. Right now, it’s gene...
itdependspod.com
October 26, 2025 at 6:03 PM
Every wave of tech promises transformation.
In the latest It Depends episode, I share how leaders can make thoughtful technology choices that last: balancing curiosity with caution, and learning before leaping.
Making Technology Choices That Last
itdependspod.com/episodes/mak...
In the latest It Depends episode, I share how leaders can make thoughtful technology choices that last: balancing curiosity with caution, and learning before leaping.
Making Technology Choices That Last
itdependspod.com/episodes/mak...
New newsletter: The Hidden Architecture of Engineering kevingoldsmith.substack.com/p/the-hidden...
October 20, 2025 at 4:46 PM
New newsletter: The Hidden Architecture of Engineering kevingoldsmith.substack.com/p/the-hidden...
I've heard "Code Survivability Rate" come up as a useful metric for measuring AI-assisted developer productivity. I'm not sure it is a great metric as it highly depends on your workflow. I tend to use AI as a pair programming partner and follow XP processes of progressive improvement. ->
October 19, 2025 at 5:48 PM
I've heard "Code Survivability Rate" come up as a useful metric for measuring AI-assisted developer productivity. I'm not sure it is a great metric as it highly depends on your workflow. I tend to use AI as a pair programming partner and follow XP processes of progressive improvement. ->
Reposted by Kevin Goldsmith
Interesting impact of Claude Code:
I dabble with frontend, but not a frontend expert. I always reached for etiher some templates or something like Webflow to put together some landing page.
Now I started to... ask Claude Code to build me as per my spec.
A big reason for it:
I dabble with frontend, but not a frontend expert. I always reached for etiher some templates or something like Webflow to put together some landing page.
Now I started to... ask Claude Code to build me as per my spec.
A big reason for it:
October 18, 2025 at 4:28 PM
Interesting impact of Claude Code:
I dabble with frontend, but not a frontend expert. I always reached for etiher some templates or something like Webflow to put together some landing page.
Now I started to... ask Claude Code to build me as per my spec.
A big reason for it:
I dabble with frontend, but not a frontend expert. I always reached for etiher some templates or something like Webflow to put together some landing page.
Now I started to... ask Claude Code to build me as per my spec.
A big reason for it:
So used to seeing Meri Williams as a conference host, I forgot how good a speaker Meri is! #LeadDevNYC
October 16, 2025 at 6:39 PM
So used to seeing Meri Williams as a conference host, I forgot how good a speaker Meri is! #LeadDevNYC
Kudos to Pooja Varshneya for the prettiest slides so far. #LeadDevNYC
October 15, 2025 at 7:28 PM
Kudos to Pooja Varshneya for the prettiest slides so far. #LeadDevNYC
Really liked the talk from @addyosmani.bsky.social on leadership supporting AI coding. #LeadDevNYC
October 15, 2025 at 4:05 PM
Really liked the talk from @addyosmani.bsky.social on leadership supporting AI coding. #LeadDevNYC
Hello #LeadDevNYC! @yee.camp starting things off!
October 15, 2025 at 1:48 PM
Hello #LeadDevNYC! @yee.camp starting things off!
new newsletter: Values->Culture->Everything
open.substack.com/pub/kevingol...
open.substack.com/pub/kevingol...
October 5, 2025 at 9:10 AM
new newsletter: Values->Culture->Everything
open.substack.com/pub/kevingol...
open.substack.com/pub/kevingol...
Most company "values" are BS. Real test: Can you tell your boss "no" when they ask you to violate a core value without getting fired?
Culture = what gets rewarded, not what's on the wall.
If you're constantly fighting your workplace culture find one that fits instead
itdependspod.com/episodes/val...
Culture = what gets rewarded, not what's on the wall.
If you're constantly fighting your workplace culture find one that fits instead
itdependspod.com/episodes/val...
Values → Culture → Everything: Why Company Culture Actually Matters - It Depends Podcast
Kevin Goldsmith revisits his influential 2013 keynote on engineering culture, expanding the conversation beyond just engineering teams to address anyone navigating workplace dynamics. This episode bre...
itdependspod.com
September 30, 2025 at 5:19 AM
Most company "values" are BS. Real test: Can you tell your boss "no" when they ask you to violate a core value without getting fired?
Culture = what gets rewarded, not what's on the wall.
If you're constantly fighting your workplace culture find one that fits instead
itdependspod.com/episodes/val...
Culture = what gets rewarded, not what's on the wall.
If you're constantly fighting your workplace culture find one that fits instead
itdependspod.com/episodes/val...
Reposted by Kevin Goldsmith
We need very strong norms for people taking responsibility for what they generate + share, with or without AI tools
It should be considered rude to ask for someone's attention or effort if you yourself outsourced your thinking to an AI tool and did not critically evaluate the output
It should be considered rude to ask for someone's attention or effort if you yourself outsourced your thinking to an AI tool and did not critically evaluate the output
September 13, 2025 at 6:10 PM
We need very strong norms for people taking responsibility for what they generate + share, with or without AI tools
It should be considered rude to ask for someone's attention or effort if you yourself outsourced your thinking to an AI tool and did not critically evaluate the output
It should be considered rude to ask for someone's attention or effort if you yourself outsourced your thinking to an AI tool and did not critically evaluate the output
Reposted by Kevin Goldsmith
A periodic reminder that leadership and management are two entirely different things. Calling a manager a "leader" is an appalling bit of doublespeak. You cannot appoint or assign a leader.
1/6
1/6
September 23, 2025 at 4:47 PM
A periodic reminder that leadership and management are two entirely different things. Calling a manager a "leader" is an appalling bit of doublespeak. You cannot appoint or assign a leader.
1/6
1/6
New newsletter! kevingoldsmith.substack.com/p/making-the...
September 22, 2025 at 12:15 AM
New newsletter! kevingoldsmith.substack.com/p/making-the...
Reposted by Kevin Goldsmith
Talked with the Claude Code team on how they build Claude Code. It feels I get a peek into the future, and I get why Dario said 6 months ago that 90% of code will be written by AI.
This team works SO differently than any eng team I saw. Will share in-depth soon. One example:
This team works SO differently than any eng team I saw. Will share in-depth soon. One example:
September 19, 2025 at 4:54 PM
Talked with the Claude Code team on how they build Claude Code. It feels I get a peek into the future, and I get why Dario said 6 months ago that 90% of code will be written by AI.
This team works SO differently than any eng team I saw. Will share in-depth soon. One example:
This team works SO differently than any eng team I saw. Will share in-depth soon. One example:
Reposted by Kevin Goldsmith
OpenAI researchers discovered that their AI models don’t just hallucinate (confident guesses) but also scheme (deliberate lies to the user).
Anyone who’s vibecoded for any decent amount of time has caught these sorts of lies. Luckily we aren’t putting these models in mission critical systems…right?
Anyone who’s vibecoded for any decent amount of time has caught these sorts of lies. Luckily we aren’t putting these models in mission critical systems…right?
OpenAI’s research on AI models deliberately lying is wild | TechCrunch
AI models don't just hallucinate. They also "scheme," meaning deliberately lie or hide their true intensions.
techcrunch.com
September 19, 2025 at 12:44 AM
OpenAI researchers discovered that their AI models don’t just hallucinate (confident guesses) but also scheme (deliberate lies to the user).
Anyone who’s vibecoded for any decent amount of time has caught these sorts of lies. Luckily we aren’t putting these models in mission critical systems…right?
Anyone who’s vibecoded for any decent amount of time has caught these sorts of lies. Luckily we aren’t putting these models in mission critical systems…right?
Back with a new podcast after my summer break! Let me know what topics you'd like me to cover in future episodes. This week's episode covers how to handle a gap in your resume (past or current) and answering the question "What does a CTO do with their day?"
pod.link/1724860326
pod.link/1724860326
pod.link
September 14, 2025 at 10:56 PM
Back with a new podcast after my summer break! Let me know what topics you'd like me to cover in future episodes. This week's episode covers how to handle a gap in your resume (past or current) and answering the question "What does a CTO do with their day?"
pod.link/1724860326
pod.link/1724860326
Reposted by Kevin Goldsmith
Young people going to Silicon Valley to work long hours in the hopes of getting rich isn’t news. Microserfs was documented this back in 1995.
That Gen Z are consuming less alcohol and having less sex also isn’t news.
It is news that “no fun” is also a part of this culture. What a grim existence.
That Gen Z are consuming less alcohol and having less sex also isn’t news.
It is news that “no fun” is also a part of this culture. What a grim existence.
AI Startup Founders Tout a Winning Formula—No Booze, No Sleep, No Fun
To reach the dream of creating a trillion-dollar company, the 20-somethings flocking to San Francisco give up everything but their laptops.
www.wsj.com
September 12, 2025 at 4:03 PM
Young people going to Silicon Valley to work long hours in the hopes of getting rich isn’t news. Microserfs was documented this back in 1995.
That Gen Z are consuming less alcohol and having less sex also isn’t news.
It is news that “no fun” is also a part of this culture. What a grim existence.
That Gen Z are consuming less alcohol and having less sex also isn’t news.
It is news that “no fun” is also a part of this culture. What a grim existence.
Reposted by Kevin Goldsmith
Full article: techblog.cloudkitchens.com/p/study-and-...
This feels like a lot more grounded take on AI tools, and one that skips the usual hype
Also good to see real feeback on all these tools. Interesting eg how Devin considered expensive and not improving as fast as others
This feels like a lot more grounded take on AI tools, and one that skips the usual hype
Also good to see real feeback on all these tools. Interesting eg how Devin considered expensive and not improving as fast as others
Study and Update on GenAI DevEx
Internal GenAI developer tools, and learnings about vendors
techblog.cloudkitchens.com
September 10, 2025 at 12:14 PM
Full article: techblog.cloudkitchens.com/p/study-and-...
This feels like a lot more grounded take on AI tools, and one that skips the usual hype
Also good to see real feeback on all these tools. Interesting eg how Devin considered expensive and not improving as fast as others
This feels like a lot more grounded take on AI tools, and one that skips the usual hype
Also good to see real feeback on all these tools. Interesting eg how Devin considered expensive and not improving as fast as others