Prediction:
The next 12-24 months, "UX-pilled" builders will be in massive demand.
Who can create intuitive interfaces, web+mobile+desktop apps that "feel good," natural, fast, and far better than the competition.
THIS will be the difference vs those building "just" with AI.
There is massive irony in how AI coding tools are starting to become TOO expensive for many enterprises - after eg Anthropic removed subsidizing AI subscriptions.
We might go from "everyone use AI for everything!" to "you have $300/month AI budget; use your brain for the rest."
BUT the handler is disabled when the text has any formatting!! So I cannot hit it.
Whoever decided to disable the three dots did a terrible job.
What you showed only works with unformatted text... this is a very bad regression itself
I'm sure there's a team at Notion patting their back for seeing "increased AI usage"
Yeah because you sabotaged a perfectly good UI that worked very well, and now are forcing senseless usage of AI
It's like removing buttons from a car and moving everything to a voice assistant
I hate AI when it's like this:
Notion used to have a one-click way to turn text into a quote. They removed this and now I have... AI
So instead of sub-500ms, it now takes 10+ seconds to turn selected text into a quote (type out the prompt, then wait ~4 seconds)
So backwards...
They were elected in one of the districts in Budapest promising a bunch of stuff including visibility into corruption happening - and did none of it, and refused to investigate corruption.
That party is a joke, unable to do any work. A documentary on them (Hungarian) youtu.be/Cr52io5lH6s?...
It’s also not just Magyar: for his future cabinet, he recruited some of the top Hungarian business people with international experience (a global VP at Shell, another global VP at Vodafone etc.)
Orban’s cabinet has minsters with 0 industry experience, 0 international, only Hungarian politics
Very trustable and extremely high integrity so far.
Ahold his party win, one thing he promised to pass is no person can be prime minister for more than a total of 8 years.
Orban is in year 20 (16 of this consecutive.) The first 8 years Orban was still sensible, then it went downhill
Today, Hungary votes. The choice is between an anti-EU, pro-Russia, pro-corruption party reigning for 16 years (Orbán’s party: Fidesz) or a pro-EU, anti-Russia, anti-corruption party (Tisza).
My mail-in vote went for Tisza ❤️🤍💚
A rendszerváltásért!
I use AI a lot for deep research and summarization. One thing I'm noticing across all models (Claude, ChatGPT, Gemini) is how they are becoming... more generic? More "AI-templated" in writing? Lazier? (Using the same tired phrases more)
I subjectively feel they are getting worse, not better.
Deepdives about MCP in @pragmaticengineer.com:
newsletter.pragmaticengineer.com/p/mcp and newsletter.pragmaticengineer.com/p/mcp-deepdive
MCP is very much alive! With co-creator of MCP David Soria Parra at AIE London
Brought to you by:
• WorkOS – Everything you need to make your app enterprise ready. Check out WorkOS.com
• Sonar – The makers of SonarQube, the industry standard for automated code review. Learn more about how SonarQube Advanced Security: www.sonarsource.com/products/son...
• Spotify: open.spotify.com/episode/3N2F...
• Apple: podcasts.apple.com/us/podcast/d...
Brought to you by:
• Statsig – The unified platform for flags, analytics, experiments, and more. Stop switching between different tools, and have them all in one place. statsig.com/pragmatic
(cont'd)
Just six months ago, David Heinemeier Hansson (the creator of Ruby on Rails and Omarchy) ssaid how he doesn’t use AI tools to write code, because they are not good enough. Things have changed, a lot: now he works agent-first.
• YouTube: youtu.be/JiWgKRgdgpI
(cont'd)
So lots of devs shipping overcomplicated PRs written with AI that score high "throughput" scores and/or use lots of tokens!!
Goodhart's law
This is absolutely not about DX btw - their TrueThroughput is actually a sensible approach to try to get a sense of code throughput. And they do NOT recommend to optimize for this.
But once devs see a number related to throughput or productivity, they just figure out a way to make it better!
Today, I got messages from 3 different devs at 3 different large tech companies (1,000+ devs each). "tokenmaxxing" is happening inside all of them.
Why?
Because all of them measure either token usage, or surface metrics like DX's TrueThroughput that reward complex code!
The more I use AI tools, the more I have to admit that I'm not that much more productive... I simply FEEL that much more productive.
In reality, the context switching of kicking several things off wipes out my perceived productivity gains. At least in many/most cases!
I'll be doing a podcast with Anders Hejlsberg, the creator of Turbo Pascal, C#, and TypeScript.
What would you like to know?
I am not offering any advice ha. I'm observing this repo, based off of Claude Code's original source is up, and very popular, and becoming even more so.
The workaround claimed:
We don't even know if fully AI-generated code has copyright btw (like Claude Code's). All open questions, no?
An interesting learning from this convo: expect multiple rewrites during hypergrowth.
The right architecture depends on how fast a product and company are growing. At Uber, repeated rewrites were common because each one “bought” another window of survival for the company.
Brought to you by:
• Sonar – The makers of SonarQube, the industry standard for automated code review. Check out SonarQube Advanced Security: www.sonarsource.com/products/son...
(cont'd)
• Spotify open.spotify.com/episode/13v4...
• Apple podcasts.apple.com/us/podcast/s...
Brought to you by:
• Statsig – The unified platform for flags, analytics, experiments, and more. statsig.com/pragmatic
• WorkOS – Everything you need to make your app enterprise ready. workos.com
(cont'd)
Why did Uber build thousands of microservices, why did it have two levels for "senior software engineer" (L5A & L5B), and why were some key infra services build in the Denmark office? No better person than Uber's first CTO, Thuan Pham to answer these & more:
YouTube: youtu.be/3jjRNVfm3V4
(cont'd)
Tomorrow on The Pragmatic Engineer Podcast: Uber's first and longest-serving CTO, Thuan Pham (now CTO at Faire)
We will talk about why Uber went all guns blazing on microservices, the unique L5A/L5B levels, and many epic, previously publicly untold stories.
Also another angle to consider:
Fully AI-generated code has no copyright
Almost all of Claude Code is AI generated (as shared by Anthropic)
So... does that AI-generated code itself have copyright, to start with?
bsky.app/profile/amyh...
no but a few things:
1. can AI-written code (that has no copyright) have copyright? Almost all of Claude Code was AI-written
2. The implementation claims a "clean-room Python rewrite" trying to position as not an exact derivative. LLMs themselves have some level of derivative characteristics also
also, Anthropic doesn't feel like they would want to argue to expland derived works protection, if done with eg a coding AI agent (that is one of their biggest renveue generators!)
