iiinova — where your digital self keeps evolving

Ólafur Páll Geirsson @geirsson.com · 07/06/2025

The best mental model for buying strollers is to think of renting them by the month. The resale value holds pretty high so the final cost isn’t so bad even for the $2k premium twin strollers.

010

Ólafur Páll Geirsson @geirsson.com · 03/06/2025

The usual answer is “no” whenever people ask whether LSP can be used in a novel way. The protocol’s strength is also its weakness, it’s very much optimized around a single human user interacting with an IDE.

000

Ólafur Páll Geirsson @geirsson.com · 31/05/2025

Meanwhile we’ll get ads in ChatGPT.

000

Ólafur Páll Geirsson @geirsson.com · 31/05/2025

Anthropic is the king of function calling and deserves their incredible revenue growth. They’ve paved the way for AI agents, not OpenAI or Google. It’s only a matter of time before Google gets the memo and Gemini starts taking function calling more seriously.

110

Ólafur Páll Geirsson @geirsson.com · 30/05/2025

Three kids done with chickenpox this month.

020

Ólafur Páll Geirsson @geirsson.com · 29/05/2025

On a second iteration, it seems like it's the web tool that's causing troubles. Disabling the web tool makes Claude 4 reach the right syntactic solution although not with the optimal token edits. Goes to show that you need to be careful with what tools you're exposing. Less is more.

000

Ólafur Páll Geirsson @geirsson.com · 29/05/2025

Sonnet 3.7 is the only model I've seen that delivers the perfect solution, it replaces the tokens for `.` and `apply` and nothing else. All other models I've tested use the worse tree replacement APIs.

100

Ólafur Páll Geirsson @geirsson.com · 29/05/2025

Surprisingly, Sonnet and Opus 4 both fail on one of my go-to codegen tests for new models > Implement a Scalafix rule that converts foo.apply(...) to foo(...) and explain why it's semantic or syntactic They both think it needs to be semantic (aka. have access to types and symbol).

110

Ólafur Páll Geirsson @geirsson.com · 27/05/2025

Amp Tab is coming along nicely, it's not too far from being able to replace Cursor Tab as my daily driver.

000

Ólafur Páll Geirsson @geirsson.com · 27/05/2025

Last note, even with code that can be unit tested, I still think most of the tests that AI generates is crap. And the AI generated commit messages also miss the point. I'm seeing lots of PRs now where people add AI generated tests that aren't even testing anything meaningful.

020

Ólafur Páll Geirsson @geirsson.com · 27/05/2025

The Dwarkesh episode still gave a fresh perspective on how these models work, and I have probably underestimated how powerful they will become. If you're still judging AI capabilities by today's products and today's models then you are probably also underestimating how weird things are going to get.

100

Ólafur Páll Geirsson @geirsson.com · 27/05/2025

I am knee deep in the AI hype, and I don't think software engineering will ever be the same again. I love working on ampcode.com and I see daily anecdotes how AI coding is turning software development upside-down for our users.

ampcode.com

Amp

Everything will change.

110

Ólafur Páll Geirsson @geirsson.com · 27/05/2025

Even components that can be unit tested or e2e tested via behavioral assertions have lots of implicit constraints wrt. latency or how features interact with each other in long-running user sessions that are impractical to tests in an automated fashion.

100

Ólafur Páll Geirsson @geirsson.com · 27/05/2025

The fallacy is thinking that all software engineers do is deliver code that can be tested in isolation, and AI is very good at doing that now. The problem is that tests only cover maybe 0-50% of real-world constraints.

100

Ólafur Páll Geirsson @geirsson.com · 27/05/2025

I keep shaking my head hearing AI folks claiming software engineering will be automated this year. After listening to this conversation, I better understand what they at least mean by this. These AI researchers are super smart, but they're also sort of clueless over what "software engineering" is.

100

Ólafur Páll Geirsson @geirsson.com · 27/05/2025

The Dwarkesh episode on Claude 4 is the most in-depth, balanced, and (almost) non-hype conversation I have heard on why AI researchers believe AGI is around the corner open.spotify.com/episode/3H46...

open.spotify.com

How Does Claude 4 Think? — Sholto Douglas & Trenton Bricken

Dwarkesh Podcast · Episode

150

Ólafur Páll Geirsson @geirsson.com · 25/05/2025

Memory reminds me of the Facebook feed circa 2016. It was clearly beneficial for the company, it sure boosted engagement, but deleted my Facebook account and was better off for it.

000

Ólafur Páll Geirsson @geirsson.com · 25/05/2025

Memory in AI chatbots is overrated, it turns the LLM into a sycophant by tying every response with random pieces of information that got extracted from past conversations. I’m sure memory is great for engagements/likability, but it’s turned me off ChatGPT personally.

100

Ólafur Páll Geirsson @geirsson.com · 23/05/2025

After starting working on Amp (ampcode.com ): - No meetings - No code review, just push to main - Take responsibility for your changes - Rarely need to create a branch off main - Auto-release every few hours - Prioritize user bug reports whenever possible

ampcode.com

Amp

Everything will change.

030

Ólafur Páll Geirsson @geirsson.com · 22/05/2025

At the risk of being pedantic, when many people say “one shot” the actually mean zero shot pass@1. Technically, one shot means including one example output in the prompt, and most prompts don’t do that. Not blaming, I even catch myself saying one shot meaning pass@1.

000

Ólafur Páll Geirsson @geirsson.com · 22/05/2025

Contrary to popular belief, the models are surprisingly bad at writing CSS.

000

Ólafur Páll Geirsson @geirsson.com · 22/05/2025

There’s a different trajectory for the people who are excited about AI because it enables them to build more expertise, or skip building expertise. Concrete example, I love AI because it helps me learn CSS faster, not because AI writes all my CSS so I don’t have to learn it.

120

Ólafur Páll Geirsson @geirsson.com · 21/05/2025

git worktrees are overrated, they're a performance optimization that only makes sense when working in a repo that's super slow to clone. For normal repos, just clone twice and enjoy benefits like being able to check out the main branch in both clones at the same time.

030

Ólafur Páll Geirsson @geirsson.com · 19/05/2025

Sprinkling Copilot dependencies across the VS Code codebase is a great technique to make it more annoying to keep a fork up-to-date. Well played, Microsoft.

000

Ólafur Páll Geirsson @geirsson.com · 18/05/2025

Got o3 to think for a record-breaking amount of time (4m54s) with this prompt. “How many songs in the Eurovision finals are in English? I have a feeling there are unusually few this year.” It enumerated the songs and classified the lyrics, very cool. Answer was 11 out of 26 songs were in English.

000

Ólafur Páll Geirsson @geirsson.com · 17/05/2025

It’s naive to believe that tech companies are slowing down hiring due to AI productivity gains. Only a small fraction of people in tech work on projects with meaningful revenue generation, and it’s a better investment to buy GPUs for the winning projects than to employ people for losing projects.

030

Ólafur Páll Geirsson @geirsson.com · 17/05/2025

Working on Amp makes it clear that Observables is really all you need, no requests, no notifications, no promises. Wire it straight into Svelte readable and everything becomes magically reactive. Beautiful.

020

Ólafur Páll Geirsson @geirsson.com · 17/05/2025

Hand-rolled a JSON-RPC like protocol with JSONL on the wire and it’s so simple that it’s almost a crime that LSP popularized the complicated Content-Length based format on the wire.

100

Ólafur Páll Geirsson @geirsson.com · 15/05/2025

Move to a place where neighbors randomly invite you to enjoy freshly baked cinnamon buns on a Thursday morning

030

Ólafur Páll Geirsson @geirsson.com · 10/05/2025

Claude has the confusing name tbh, their Max is limited.

000

Ólafur Páll Geirsson @geirsson.com · 10/05/2025

So Claude Code has token-based pricing unless you subscribe to Max for a flat rate. Meanwhile, Cursor has a flat rate unless you use the MAX models, which have token-based pricing.

100

Ólafur Páll Geirsson @geirsson.com · 10/05/2025

I once sent an email with some copy-pasted text from ChatGPT and realized afterwards you can tell from the HTML formatting. I’m still embarrassed by this.

110

Ólafur Páll Geirsson @geirsson.com · 07/05/2025

OpenAI and Microsoft must have a sour relationship when OpenAI acquires a VS Code fork to compete against Copilot.

020

Ólafur Páll Geirsson @geirsson.com · 06/05/2025

Date-based versioning is underused. I love how JetBrains IDEs have easy to remember versions like 2019.3. Meanwhile, we have Gemini Pro 2.5 Experimental (New).

060

Ólafur Páll Geirsson @geirsson.com · 06/05/2025

Feature request: testing library that automatically re-runs new tests N times to catch flakes.

010

Ólafur Páll Geirsson @geirsson.com · 04/05/2025

Highly recommend using o3 to geoguess screenshots of random images from your personal photo library. It's been off by ~30km on the two examples I've tried so far.

010

Ólafur Páll Geirsson @geirsson.com · 03/05/2025

Apple buying Anthropic would make a lot of sense. Strong win-win vibes.

000

Ólafur Páll Geirsson @geirsson.com · 03/05/2025

It’s definitely a different story on the server!

020

Ólafur Páll Geirsson @geirsson.com · 02/05/2025

It feels very good being able to develop on the main branch and pushing straight to origin. No branching, no PR. Every dev should try to at least experience this.

030

Ólafur Páll Geirsson @geirsson.com · 02/05/2025

Nix is so appealing on the surface but I always get turned off quickly after trying to use it. I get maybe 98% of the benefits via mise/direnv with none of the complexity. I must be missing something.

200

Ólafur Páll Geirsson @geirsson.com · 01/05/2025

The ArkTypr syntax is a lot more approachable, it looks like TypeScript types except inside string literals. As soon as you have unions of arrays and undefined, the Zod code becomes difficult to read and evolve.

010

Ólafur Páll Geirsson @geirsson.com · 29/04/2025

ArkType is absolute TypeScript sorcery. Very glad to see a strong zod competitor enter the market of runtime validation with a focus on performance and developer ergonomics. Runtime validation has always been a weak spot in the JS ecosystem, and ArkType might just fix that important gap for me.

150

Ólafur Páll Geirsson @geirsson.com · 29/04/2025

Maybe file-based issue tracking will have a comeback with agents checking off their TODOs. Wiring OAuth to something like JIRA/Linear feels heavy-weight when the files are just lying there and LLMs are good at editing files.

140

Ólafur Páll Geirsson @geirsson.com · 28/04/2025

First day today working with git-branchless. It takes some time getting used to, but I can already see how this workflow is greatly superior to the manual branch+rebase workflow that I've been using for years.

000

Ólafur Páll Geirsson @geirsson.com · 28/04/2025

Not that kind of rewrite. I'm thinking well reasoned rewrites. I believe people are generally too blindly opposed to rewrites.

010

Ólafur Páll Geirsson @geirsson.com · 28/04/2025

AI will definitely play a role. Many vibe coded apps will reach PMF and then get rewritten. I also think mainstream adoption of sync engines is going to justify clean rewrites in many cases. Canva paused development for two years to rewrite their codebase and it was wildly successful.

010

Ólafur Páll Geirsson @geirsson.com · 28/04/2025

Conventional wisdom says don’t do a rewrite, but I think we’re going to see many success stories of huge rewrites over the next years.

140

Ólafur Páll Geirsson @geirsson.com · 28/04/2025

I can’t believe I’m saying this but the world needs a new multiplayer AI chat application. Slack and Discord don’t cut it for me.

010

Ólafur Páll Geirsson @geirsson.com · 27/04/2025

I don’t expect Mercurial will gain mainstream adoption, but I do suspect we’ll see more large US tech companies grow out of git, drooling over what Meta and Google use internally.

000