Kristofer Palmvik.
E-mail
kristofer@palmvik.sePhone
+46 705 886414
Postal address
Kristofer PalmvikHamngatan 16 D
582 26 Linköping
SWEDEN
Social media
About Kristofer Palmvik
I love to develop useful and cool things with lots of data. Love to work with other people. Love to travel the world.
MSc Information Technology with specialization in Media Informatics. Professional experience of software engineering, ranging from modern frontend web to high-performance databases and cloud infrastructure.
Interested in working with product development in a web based environment. Quick to learn new skills and handle new situations. Thrives in a demanding environment together with a great team.
Aim to be a valuable part of a small company and develop a world leading product or service with high quality and professionalism.
See my professional experience and current position on LinkedIn. Let's connect!
Currently reading
I enjoy reading books about a wide range of subjects.
- Millenniumskandalen : journalsystemet som havererade by Simon Campanello,
- Herr Saitos ambulerande biograf by Annette Bjergfeldt, and
- AI Engineering: Building Applications with Foundation Models by Chip Huyen.
A recent favorite book is Den sista milen : om e-handelns osynliga arbetare och de snabba paketens dolda pris by Julia Lindblom.
My notes
Observations on building software, working with teams, and exploring the world. A mix of professional thoughts and personal logs.
- 2026‑05‑20
Did I miss something in my writing?
What if your writing came with a built-in critic? Mine does. An AI finds the seams in my thinking and publishes them right next to the text.
- 2026‑05‑16
A podcast that creates itself, in a different voice
My notes now become 5-minute podcast episodes overnight, written and voiced by a chain of AI agents. The episodes sometimes contradict my own thinking, which is exactly the point.
- 2026‑05‑04
Claude knows FFmpeg, but it has no idea where the video is
Testing an AI video editor revealed that Claude writes flawless FFmpeg commands, but made assumptions around file system paths.
- 2026‑04‑21
The manual is missing – we are writing it
The feeds make it easy to feel left behind when it comes to AI. But history proves the hype is usually louder than reality. Since there is no manual for this era, trying things and sharing even the mundane experiences could be useful. We are writing the manual.
- 2026‑04‑08
The AI writes the code, but the team decides the direction
I have barely typed a single line of code at work for more than a year. Using LLMs and AI agent has changed a lot. But we still need the collaboration in a great team to make sure we go in the right direction.
- More of my notes...
Whitebrd.
A public collection of links I'm reading and referencing since 2008. Articles, tools, and resources relevant to my work and interests.
- 2026‑05‑23
After the Feed: Eli Pariser’s Roadmap for the “Agentic Interface” Era
In this new agentic era, the interface itself becomes the AI. Instead of scrolling through a list of posts, we will interact with “agents” that operate at an inhuman speed and scale. These interfaces aren’t just ranking content; they are transforming it—pulling out core ideas, changing their form, and presenting a synthesized version tailored specifically for an audience of one.mediaparty.org - 2026‑05‑23
auth.md
A Markdown file an application hosts at its domain — typically https://yourapp.com/auth.md — that tells agents how to register on behalf of a user. It includes which flows are supported, which scopes exist, and how to register for the service.workos.com - 2026‑05‑23
The Agentic Manifesto: Engineering in the Era of Autonomy
The transition to the Agentic Delivery Lifecycle is not optional for organizations that intend to lead in the AI era. It requires engineering leaders to make uncomfortable shifts: investing in governance as heavily as features, and valuing a robust evaluation suite as highly as production code.agenticmanifesto.ai - 2026‑05‑23
Kimi K2.5: Still Worth It After Two Weeks?
When Artificial Analysis evaluated K2.5, the model generated 89 million output tokens. The median for comparable models is 14 million. At $0.60/$3.00 per million input/output tokens, the per-token price looks cheap, but when the model produces 6x more tokens per task, effective costs still spike. Kilo Code’s week-long free trial confirmed this: usage surged past 50B tokens/day, and they concluded the model’s verbosity dilutes the savings from input caching. This is the main issue with this model.mlabonne.github.io - 2026‑05‑23
Open From Both Sides
The strategy is access restriction — the idea, foundational to Western AI safety since 2023, that controlling a model means controlling its capabilities. Anthropic’s decision to withhold its new model is the purest expression of this logic. But two forces are now converging, making it untenable. Chinese labs are releasing open-source models that match Western closed ones, trained entirely on domestically manufactured chips that the US tried to restrict. And safety alignment — the technical mechanism that’s supposed to make open release safe — is being removed from any model within days of publication, at zero cost, by anyone who can type a command.airealist.ai - More of my collected links...
