Kevin Dominik Korte

I'm an angel investor and tech evangelist exploring Open Source, enterprise IT, AI, cybersecurity, and digital sovereignty. I share practical perspectives on startups and building resilient, human-centered technology. Occasionally, you'll also see rants, sarcasm, or dad stuff here.

I'm available as a speaker, board advisor, and mentor.

Website: https://www.korte.co
OS: UCS + CatchyOS + Ubuntu Studio

Public Key

npub1lwqfcc0mc3frqk3rycnwk5fr0sj8wvmju2zgwrpwz07svyt665uswuatsn

NIP-05 Address

kdkorte_at_fosstodon.org@momostr.pink

Profile Code

nprofile1qqs0hqyuv8aug53stg3jvfht2y3hcfrhxdew9py8pshp8lgxz9ad2wgpz3mhxue69uhhyetvv9ujuerpd46hxtnfduqs6amnwvaz7tmwdaejumr0dsagzys4

Publishing to

relay.damus.io

nos.lol

Author Public Key

npub1lwqfcc0mc3frqk3rycnwk5fr0sj8wvmju2zgwrpwz07svyt665uswuatsn

Show more details

Last Notes

2026-06-25 16:16:47 UTC

- reply

I'd say Valve looked at the data and thought about what the minimum viable specs are for an adequate Steam gaming system. The idea to run general purpose Linux is at most secondary. Or said differently, if they could have gotten 1% more performance for gaming, but broken compatibility with LibreOffice, well who wants to edit documents on their gaming rig.

2026-06-25 15:52:20 UTC

- reply

I would say the push for Linux has more to do with the uncertainty around Microsoft.
After all, Microsoft flip-flops around mandating the Windows Store, the future of Xbox/Microsoft Games, supplying different graphics interfaces, requiring signed MSI packages for installs, and weird support for ARM processors, all of which could significantly impact Steam.
Without any certainty around these issues, Valve might want an alternative.

2026-06-25 15:41:51 UTC

- reply

It's a bit like the console market, where the games subsidize the hardware.
Yet, unlike a console, they can easily abandon the whole hardware business if it flops. After all, there are enough PC manufacturers to keep Steam viable.

2026-06-25 13:21:05 UTC

- reply

Someone built their own Steam Machine. Components came to $700 without the case.
Bulk discount normally runs at around 10%. Case, I guess at $30. So Valve likely pays $650 for parts.
That leaves them with about $400 for assembly (probably $50-100), support (around $100), development ($50), sales and marketing ($10), and inventory ($10-50)
So, somewhere between $50 and $100 in profit per machine. Not a good deal for Valve
https://www.xda-developers.com/someone-built-their-own-steam-machine-at-valves-original-price-point-and-its-a-solid-contender/

2026-06-11 17:39:54 UTC

- reply

If you look at the statement, he deliberately uses the word margin because Microsoft Games is profitable. It just isn't profitable enough to cover all of Microsoft's AI spending and deliver sufficient revenue for Exec bonuses and shareholders.
I'd assume they are going to squeeze both the revenue by increasing prices and the bottom line by firing more nonessential staff, like designers and developers.
After all, Copilot Gaming is the future. (for the CEO Bonus)

2026-05-13 13:00:30 UTC

It is amazing how everyone at the OpenAI trial is nonchalantly admitting they committed crimes. So far we have:
- Lying under oath
- Falsifying business records
- Breach of fiduciary duty
All without remorse or even understanding, as if it were just another Tuesday.
#AI #business #governance

2026-05-06 13:57:50 UTC

- reply

Everyone clearly learned the lessons from WeWork. It's almost comical how Softbank was the biggest player in both cases.

2026-05-06 13:00:27 UTC

Lessons from the OpenAI vs. Musk trial so far:
a) If you're planning dubious financial decisions, don't write them down in your diary.
b) If your story is good, people will believe it.
c) No one at OpenAI understood non-profit governance law.
d) Self-dealing is rampant.
#business #AI

2026-05-06 03:59:46 UTC

If your AI systems are solely focused on speed, human oversight might become nothing more than theater. And we all know from cybersecurity how effective theater can be at convincing people they are doing the right thing while changing nothing.
#AI #oversight #goverance
https://www.techpolicy.press/ai-efficiency-can-undermine-accountability-even-with-humans-in-the-loop/

2026-05-04 14:39:14 UTC

- reply

the childish side of me went to a different part of their body for the comparison

2026-05-04 13:00:05 UTC

The Musk vs. OpenAI trial shows the absurdity of the AI bubble. Either lose if their upcoming IPO is smaller than the others.
#AI #lawfare

2026-04-28 13:00:02 UTC

You cannot offload ethics to a machine.
https://www.korte.co/4t96
#AI #ethics
https://cdn.fosstodon.org/media_attachments/files/116/482/454/379/761/243/original/da3e25f090c10a75.jpeg

2026-04-25 17:58:20 UTC

It used to be a virtue to write spelling and grammar error-free text. Now, it creates enough suspicion of your text being AI-generated that we are building tools to introduce spelling errors. What a wonderful world ... (or rather wrold, I guess)
#AI #spelling
https://gizmodo.com/the-anti-grammarly-helps-people-hide-their-participation-in-the-ai-slop-era-2000750553

2026-04-23 13:00:15 UTC

Stop asking whether AI is ethical.
The real question is whether the developers building it and the users applying it are acting ethically. My latest article argues that accountability starts with people, not machines.
#AI #Ethics #TechLeadership
https://www.korte.co/2026/04/23/stop-blaming-the-machine-for-ai-failings/

2026-04-16 23:57:21 UTC

It isn't just the UK, who's purchasing decisions have put it into a big tech dependency. Unfortunately, it is most countries, organizations, and individuals.
#digitalsovereignty #bigtech #dependency
https://www.theregister.com/2026/04/15/uk_big_tech_dependence/

2026-04-10 17:11:14 UTC

There is a childish side in me that wants to take Mozilla's new tool and flood OpenAI, Anthropic, and co with meaningless bug reports.
#cybersecurity #AI
https://gigazine.net/gsc_news/en/20260410-mozilla-0din-ai-scanner/

2026-04-10 03:36:20 UTC

I think it is time to abandon anything related to Microsoft, from OpenAI to GitHub. While both sides likely contribute to the underlying issue, Microsoft's lack of support and communication channels is inexcusable for a billion-dollar company.
#microsoft #opensource
https://www.bleepingcomputer.com/news/microsoft/microsoft-suspends-dev-accounts-for-high-profile-open-source-projects/

2026-04-09 17:10:58 UTC

- reply

Given that it is Florida we are talking about, I would be afraid that a success makes them pivot from instructions on mass shootings to instructions on abortion or transgender issues.

2026-03-25 13:00:18 UTC

It's genuinely worrying how many Tech companies don't have a plan B for what they will do if their AI gambit fails. Given the amount of money that poured into the sector, it should be likewise concerning how many investors don't have a plan B. Maybe it is just a sign that money has become truly meaningless for some.
#AI #investment

2026-03-06 03:37:38 UTC

So, the Dutch government tried to whitewash Amazon's sovereign cloud offering, only to be called out so hard that they had to withdraw the paper.
#digitalsovereignty
https://nltimes.nl/2026/03/05/dutch-govt-pulls-report-dangers-american-cloud-service-criticism

2026-02-18 16:28:25 UTC

One of the biggest problems with AI slop in open-source is how demoralizing it is for the actual maintainers to deal with that sh*t. Godot shows how parts of the ecosystem are on the verge of breakdown.
#AI #opensource
https://www.pcgamer.com/software/platforms/open-source-game-engine-godot-is-drowning-in-ai-slop-code-contributions-i-dont-know-how-long-we-can-keep-it-up/

2026-01-30 18:17:14 UTC

While U.S. Big Tech may be entrenched in Canada at present, Europe shows that it only requires a bit of political will to break these limits. Now, it's just up to politicians to find the will to pivot to open-source.
#opensource #digitalsovereignty
https://thetyee.ca/Analysis/2026/01/29/Big-Tech-Spearheads-US-Threat-Canada/

2026-01-12 14:00:05 UTC

CES shows that too many companies are still prioritizing expensive, flashy AI solutions. It's unbelievable that none of these companies predict how badly this will end.
#AI

2025-12-11 19:47:47 UTC

- reply

I know. Unfortunately, they don't syndicate with anyone that has less tracking.
Of course, you could pay for pure, but the English version doesn't have enough content to warrant the price.

2025-12-11 19:24:02 UTC

Germany's Interior Ministry has analyzed how much access US intelligence agencies have to data stored in cloud services and data centers outside the US. You might just as well post your company's internals and most private data on a public website.
#digitalsovereignty
https://www.heise.de/en/news/Opinion-US-Authorities-Have-Far-Reaching-Access-to-European-Cloud-Data-11111060.html

2025-12-04 03:58:14 UTC

You don't need to trust big tech for a modern workplace. @npub1qts…un2v partner @npub1rk6…hv4q has the perfect explanation of why and how Nubus + Nextcloud can deliver an open alternative that builds trust and performs well in all IT environments.
#IT #opensource #digitalsovereignty
https://nextcloud.com/blog/the-sovereign-alternative-to-microsoft-entra-nubus-from-univention-for-access-and-identity-management/

2025-11-26 15:44:47 UTC

- reply

The cost to privacy and security will fall on everyone, as most companies implement the strictest regulations necessary to comply with everything.
As for the financial costs, I'd be more concerned that, in the short term, the watering down will provide noticeable benefits to the affected companies. That, in turn, will lead to repeated calls to abolish privacy regulations, just like tax cuts and trickle-down economics in the US.

2025-11-25 21:04:02 UTC

- reply

unfortunately, it seems to be that they lack the will to change that.

2025-11-25 18:33:07 UTC

For one glorious moment, we generated significant shareholder profit.
#AI #regulation
https://www.theguardian.com/technology/2025/nov/24/us-europe-artificial-intelligence-ai

2025-11-07 18:18:54 UTC

While many Web Hosters are becoming big tech resellers, Ionos and @npub1rk6…hv4q just released an alternative to GSuite and M365. It's built entirely on open-source software and is even cheaper than most of its competitors. Let's hope others are learning from that example.
#opensource #office
https://www.heise.de/en/news/Office-alternative-from-Germany-by-Ionos-and-Nextcloud-is-now-available-11067308.html

2025-10-25 13:00:10 UTC

What surprised me most about this week's AWS outage is how many managers viewed it as a natural disaster. It just happened, it will happen again, and there is nothing you can do about it. Yet, it is not. In contrast to a hurricane or earthquake, it is much easier to remove single vendor failure points. Plus, no company would have accepted the same incident from its own infrastructure. Thus, we shouldn't give Amazon a pass.
#IT #leadership

2025-10-13 13:00:10 UTC

The biggest issue with Meta's new smart glasses is how they normalize surveillance and recordings. If we aren't careful, we'll soon accept it, just as we have accepted CCTV cameras in most public squares.
#surveillance #smartglasses

2025-09-28 13:00:00 UTC

Digging more into the Austrian Military's switch from MS Office to Libre Office. The money side is fascinating. They acknowledge that training and getting accustomed to the new interface will consume any savings in the first few years. To me, that speaks to how realistic they view this project and that it is more than just a money-saving attempt. Let's make sure we see more of these projects in the future.
#opensource #digitalsouvereignity

2025-09-19 13:00:26 UTC

Free Software Friday! With all the talks about encryption backdoors, breaking into chats, and spying on emails, I would like to give a massive shout-out to GPG. To keep my emails safe and encrypted. It does so without sharing my data.
#opensource #encryption
https://gnupg.org/

2025-09-11 15:35:15 UTC

- reply

Thanks for letting me know. Apparently, my hoster did a PHP update, and that disabled the URL shortening plugin. It should work now.

2025-09-11 13:00:13 UTC

AGPL isn’t a dealbreaker for startups or VCs. It enables open innovation, a strong culture, and customer trust. Licensing debates miss what really matters for growth: relationships and resilience. Read my latest insights on building with AGPL.
#OpenSource #AGPL #TechLeadership #StartupCulture #VentureCapital
https://www.korte.co/ca3y

2025-09-09 18:56:21 UTC

- reply

Yes, there were a lot of oligarchs fled Putin and Russia in he onset of the Crimea war. I guess they realized how unstable the situation would become for their global business empires.
You could also say, that Malta, Singapore, and the UAE gained a lot of wealth back then.

2025-09-09 16:48:35 UTC

- reply

Arkady Volozh has been living in Israel since the start of the Crimean War. He also holds Maltese Citizenship, which limits how closely EU regulators can look into his involvement.
Microsoft is likely facing more hurdles for approval, especially as its market share is increasing compared to the other big players. Interestingly, what might save this deal is that many regional EU service providers saw their share grow faster than Azure.

2025-09-09 16:31:48 UTC

- reply

It will be interesting to see how the regulatory and liability landscape plays out. Especially, with the recent clashes between EU and US regulators, we might see Microsoft getting into hot water on both sides of the Atlantic. Plus, at some point, their AI investments might become an antitrust issue. (If the LLM market doesn't have a correction beforehand.)

2025-09-07 19:37:55 UTC

- reply

thanks! I'll check them out.

2025-09-07 13:00:00 UTC

Are there any search engines left on the internet, that do not force AI at the forefront of your web searches?
#AI #searchengine

2025-08-27 18:03:18 UTC

Who would have thought? Cutting cybersecurity budgets for short-term gains can lead to significant long-term problems for businesses. Unfortunately, a lot of CEOs will find this acceptable as long as long-term is past their bonus payment.
#cybersecurity
https://www.forbes.com/councils/forbestechcouncil/2025/08/27/short-term-cybersecurity-savings-can-lead-to-long-term-risks/

2025-08-13 19:24:50 UTC

My friends at @npub1rk6…hv4q have launched a Digital Sovereignty Index. While not perfect, it gives a fascinating insight into which countries self-host and thus control their digital destiny.
#opensource #digitalsovereignty
https://dsi.nextcloud.com/

2025-07-26 17:15:31 UTC

- reply

the problem with this race to the top is, that neither is the top well defined, nor do they mind how much collateral damage they are causing

2025-07-26 13:00:10 UTC

Big Tech Companies are investing 10 times more into AI than they generate from it. At the same time, their solutions are crumbling from cybersecurity issues, and their employees are devastated by layoffs. The question is, where do these company valuations come from?
#bigtech #wallstreet

2025-07-02 12:39:24 UTC

- reply

Or these tests will boil down to some Facebook-esque quizzes, like "which of these politicians makes you angrier?" And you must then infer the political leanings of the website creator.

2025-07-01 23:05:31 UTC

- reply

It will be interesting to see what effects it will have.
In particular, that AI can already solve some of the common robot tests, like identifying a traffic light

2025-07-01 17:25:30 UTC

Let's hope Cloudflare's block on AI scrapers works. It's frustrating to see the significant impact these AI scrapers have on small websites and blogs.
#AI #scrapper #fightback
https://www.nytimes.com/2025/07/01/technology/cloudflare-introduces-default-blocking-of-ai-data-scrapers.html

2025-06-01 17:32:55 UTC

If you are enrolling your child in a US school this year, please remember that providing your or your child's Social Security Number is OPTIONAL. They cannot bar enrollment without it.
Given the latest cuts in cybersecurity spending, plus the increased attacks against school districts, do your kid a favour and opt out!
#cybersecurity #education #socialsecuritynumber
https://nces.ed.gov/pubs2004/privacy/section_3d.asp

2025-04-24 16:26:36 UTC

- reply

I think it's worse than the calculator. Even for the most advanced calculators (including Octave and MATLAB), you still need to understand the concepts to use them.
With AI, you can copy the question and get a wrong answer without knowing what a single word or symbol means.

2025-04-24 16:11:34 UTC

Looks like our kids will soon be AI-literate. It's too bad that actual grade-level reading rates in the US are still abysmal.
#education #AI
https://www.whitehouse.gov/presidential-actions/2025/04/advancing-artificial-intelligence-education-for-american-youth/

2025-04-12 01:27:54 UTC

Microsoft Privacy Nightmare is finally rolling out. The wet dream of every authoritarian government and hacking group just became a reality.
#privacy #AI #microsoft
https://www.bbc.com/news/articles/cj3xjrj7v78o

2025-04-08 03:33:19 UTC

Building AI only to pass the Turing Test means we stop building technology to benefit humanity. Instead, we start to develop AI only for itself. This is an excellent insight and very relevant in our current tech environment.
#AI #philosophy
https://www.vox.com/the-gray-area/407154/jaron-lanier-ai-religion-progress-criticism

2025-03-22 19:35:07 UTC

Be aware of the new AI features if you still use Gmail as your primary E-Mail account. All your data is about to become training data for AI, beyond what Google already does for advertisement. Not might be a good time to switch to a better provider
#AI #privacy
https://www.forbes.com/sites/zakdoffman/2025/03/22/google-confirms-gmail-upgrade-3-billion-users-must-now-decide/

2025-03-22 13:00:11 UTC

Today is World Water Day! We must know how water is used to cool data centers, including AI servers, and reduce our consumption. While AI can be helpful, we cannot destroy life on the planet to feed our egos.
#worldwaterday #AI

2025-03-19 16:52:42 UTC

- reply

And the path where no one has to admit that the best design choice 30 years ago doesn't hold up against modern principles.
Yet, even if it was the right choice back then, the team (all different people now) should have known better than to make an error of judgment -> Thus, performance-based layoff.

2025-03-19 16:45:48 UTC

- reply

An unfortunate truth in IT, where many of our daily drivers have gotten too big and monolithic to be easily updated and maintained.

2025-03-19 16:40:50 UTC

- reply

From my understanding from some of our partners who build modules and extensions:
a) Some of the elements in the control panel are owned by different teams. Thus, making one big push would require too much coordination within Microsoft and with third-party providers.
b) The new interfaces don't have the same powers as the control panel. Thus, migration was hampered by needing to build new secure interfaces

2025-03-19 13:01:00 UTC

Yesterday, someone pitched to me that I can use AI to find out which job applicants unethically use AI to apply for a job. Now, I'm sitting here trying to figure out why using AI to sort through job applicants is more ethical than the applicants using AI.
#AI #ethics.

2025-02-21 03:13:15 UTC

AI is a computer program. No one asks whether a hammer is ethical, but a digital tool should somehow understand the varying forms of human ethics. Ethics will remain something we decide on, and change over time.
#AI #ethics
https://www.wired.com/story/the-prompt-ethical-generative-ai-does-not-exist/

2025-02-16 23:12:14 UTC

Weirdly, people think it's okay to pressure people who give them free things. Few people in Big Tech experience the same kind of direct user pressure that people try to exert on voluntary open-source maintainers. We need to change that.
#opensource
https://www.theregister.com/2025/02/16/open_source_maintainers_state_of_open/

2025-01-14 18:17:26 UTC

ActivityPub is a protocol that already surpasses most social media in terms of capabilities and openness. Why do tech millionaires feel the need to push another one?
#socialmedia #opensource
https://www.theregister.com/2025/01/14/free_our_feeds_decentralized_social_media/

2024-12-04 22:34:22 UTC

Refactoring is often a waste of programming time. Yet, some venture back founders want AI to refactor the US legal system. A system that is built upon 1500 years of English legal traditions and court cases. What could go wrong?
#AI
https://fortune.com/2024/12/04/peter-thiel-backed-entrepreneur-doge-should-use-ai-to-refactor-us-legal-system/

2024-10-10 01:52:50 UTC

It's worrying if your home security provider cannot protect its own data. After all, they want me to install a dozen wireless IoT devices and cameras in and around my house.
#cybersecurity #adt homesecurity
https://cyberscoop.com/adt-security-hack-employee-data/

2024-09-14 21:35:46 UTC

Writing documentation is complicated and often underappreciated. As an open-source community, we should do a better job acknowledging those who create tutorials and technical documents.
#opensource #documentation
https://www.zdnet.com/article/linux-and-open-source-documentation-is-a-mess/

2024-09-10 20:17:07 UTC

"A blueprint for responsible actions of military AI." Why do I have the slight feeling that China opting out is the most honest action in this whole debate.
#AI #military #ethics
https://www.reuters.com/technology/artificial-intelligence/south-korea-summit-announces-blueprint-using-ai-military-2024-09-10/