Kevin Dominik Korte

#AI Innovator / Growth Strategist / Board Member / #FOSS Enthusiast / #Angel Investor / Speaker

Website: https://www.korte.co
OS: UCS + KUbuntu
LinkedIn: https://www.linkedin.com/in/kevindkorte/

Public Key

npub1lwqfcc0mc3frqk3rycnwk5fr0sj8wvmju2zgwrpwz07svyt665uswuatsn

NIP-05 Address

[email protected]

Profile Code

nprofile1qqs0hqyuv8aug53stg3jvfht2y3hcfrhxdew9py8pshp8lgxz9ad2wgprpmhxue69uhhyetvv9ujumt0d4hhxarj9ecxjmnt0sjwgw

Publishing to

relay.momostr.pink

Author Public Key

npub1lwqfcc0mc3frqk3rycnwk5fr0sj8wvmju2zgwrpwz07svyt665uswuatsn

Show more details

Last Notes

2026-07-09 17:10:14 CEST

- reply

Even if they did, it then becomes the question of how much they were ready to spend. The ultimate security would be a camera looking at both the environment and the LED. Cost: 30 dollars per frame in components plus extra weight and power consumption. Defeatable likely by a 10-dollar one-way plastic sheet.
Or as wargames said: "The only way to win is not to play."

2026-07-09 16:51:54 CEST

- reply

as someone who builds model trains, I would bet a substantial amount of money, that I have the right paint and brushes to defeat the led and sensor.
Thus, ultimately Meta hundred thousand dollars software fix loose is in an arms race against $1 of paint and, as you pointed out, no one at Meta asks why they are even doing it. (Outside of it being Zuckerberg's latest vanity project)

2026-06-25 18:30:42 CEST

- reply

Valve is publishing statistics. Gaming on Linux has a nice tracker for general usage
https://www.gamingonlinux.com/steam-tracker/
SteamDB has the number of players per game. It doesn't look too different from Windows.
https://steamdb.info/charts/?category=999

2026-06-25 18:24:03 CEST

- reply

By now, yes, Steam on Linux is its own project at Valve, and they are doing a great job in expanding the market share. It certainly is great that it developed into this, and I'm rather happy to have gotten rid of Windows on my gaming system.
However, at the beginning, their dependence on Microsoft was the driver that led them to develop Proton and make it a widely usable compatibility layer, rather than porting their own games to Linux.

2026-06-25 18:16:47 CEST

- 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 17:52:20 CEST

- 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 17:41:51 CEST

- 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 15:21:05 CEST

- 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/

2026-06-24 17:24:52 CEST

- reply

I think starting with making things obvious is a good first step. No more hiding in terms and conditions. Big agree buttons, instead of complex opt-outs, and maybe force Google and Apple to personalize permission requests, e.g., "Do you want this app to read all contacts, including 'Aunt Marty Don't Pick UP'"

2026-06-24 16:46:13 CEST

- reply

It's not just convenience.
Tech companies, lawyers, and regulators have trained us to never look at Terms and Conditions, click "grant" on any permission request, and accept the little class-action data-breach card as part of our modern lives.
Of course, it could work differently. For example, cookie acceptance in the EU plummeted from over 90% to around 50% after regulators required clear explanations and explicit consent to tracking.

2026-06-24 16:38:01 CEST

- reply

I think that's the biggest problem. Everything gets judged by either this quarter's stock price or the biennial election results. Everything outside that time frame either "was someone else in the past" or "is tomorrow's problem"

2026-06-24 15:00:00 CEST

It's strange how little people value their digital identity and privacy. Upload your ID and selfie to access some AI toy: sure. Click on allowing any data to be used for marketing and AI training: Absolutely. Being annoyed about yet another data breach is part of modern life.
#identity #privacy

2026-06-23 17:44:41 CEST

- reply

the sad part, if you look at component prices, the price for the steam machine is unfortunately rather reasonable.

2026-06-20 16:56:22 CEST

- reply

I don't think that choosing DeepSeek means that your product failed. Microsoft doesn't produce a base model.
However, it does show that their strategy has failed, including their strategic investments. After all, why spend billions on OpenAI if they don't produce something that you can integrate at an acceptable price?

2026-06-20 00:59:17 CEST

Spending a billion on OpenAI. Check. Hiring a bunch of AI developers and cutting the bonus of your rank-and-file. Check. Failing to produce anything and falling back to a cheap Open Source AI. Check.
Microsoft demonstrates what a leadership failure looks like.
#opensource #AI
https://letsdatascience.com/news/microsoft-explores-deepseek-v4-for-copilot-cowork-costs-73bcc00b

2026-06-18 00:36:06 CEST

- reply

The problem with that approach is that they still won't get anywhere close to the margin that traditional software and SaaS get.
After all, the compute doesn't get cheaper with more usage.
On the flip side, it might show people how ridiculous some of the AI use cases are, and they will cut back on what to use Copilot for.

2026-06-18 00:08:17 CEST

So Microsoft is raising prices on its AI tools to encourage more people to use them. I can't put my finger on it, but there seems to be an issue.
#AI #business
https://www.axios.com/2026/06/16/microsoft-copilot-cowork-tokenmaxxing-cowork

2026-06-17 21:57:06 CEST

- reply

that's about as likely as AI putting us on Mars this year

2026-06-17 20:55:11 CEST

- reply

You just pointed out the difference. For data centers, people and communities are fighting back and doing so successfully, even if it is just through zoning.
The same, unfortunately, isn't possible with ETFs, 401 (k) s, and pensions. They are buying SpaceX by following the law, getting additional management fees, and washing their hands of any responsibility when Mecha Hitler fails to deliver any returns on investment.

2026-06-17 20:43:34 CEST

- reply

the problem is, by including SpaceX and Co in the index funds quickly, it will be all the people with retirement accounts who hold the giant bag of excrement, when the party ends.
Or said differently it won't be the AI bros but the lower middle class that suffers.

2026-06-17 17:47:34 CEST

- reply

There is also a generous helping of Enron's pre-catastrophe playbook with barely legal market manipulation, including overloading retirement accounts with highly speculative assets.
As for the use cases, you have to believe that an automated spy network that tracks every move and accuses you using a Shakespearean sonnet is beneficial for humanity's future. Otherwise, you might need some "AI-enhanced" education.

2026-06-17 17:43:38 CEST

- reply

Let's be honest, they didn't have to do much. Most people, including most elected leaders, still see anything IT-related as a novelty, not worth the time to think about.
Add in the idea that machines, including AI, are logical and neutral, and you have fertile ground to avoid regulations.

2026-06-17 15:00:10 CEST

Why do AI CEOs increasingly sound like they run tobacco companies or illegal drug operations?
#AI #governance

2026-06-16 23:35:41 CEST

Nice and clear. Chinese AI companies have a different goal from US ones. US Big Tech has bought into building the latest and greatest AI. Chinese companies want to make money. Strangely, US AI gets billion-dollar valuations ...
#AI #business
https://fortune.com/2026/06/16/china-ai-deepseek-open-source-efficiency-global-expansion-strategy/

2026-06-15 17:06:03 CEST

- reply

You can buy knowledge, but understanding and wisdom require humility and listening skills. Both are in short supply in Silicon Valley, despite most HR Policies saying otherwise.

2026-06-15 04:12:28 CEST

Maybe it's time for me to get my eyesight fixed surgically, just so no one can accuse me of wearing glasses for something other than vision correction. Or we could not build this surveillance shit.
#smartglasses #surveillance
https://m.economictimes.com/ai/ai-insights/google-gemini-smart-glasses-set-for-fall-launch-as-warby-parker-and-gentle-monster-challenge-meta/amp_articleshow/131704679.cms

2026-06-11 20:29:31 CEST

- reply

I'd say the mobile train has left for Microsoft. They neither have the studios nor the IP for ultra-microtransaction games. Better get A, B, and indie games to keep pushing Game Pass.
The thing is, if Microsoft Games keeps the quality of the Game Pass games up, it wouldn't matter that Valve works on Linux. Game Pass works much better on Windows, and Steam can only reach parity.
Alas, they have to think of the "short-term margin."

2026-06-11 20:06:53 CEST

- reply

Easiest would be for Microsoft to spin off Microsoft Games. It would give them the cash for AI and protect the gaming system from more slaughter.

2026-06-11 20:05:31 CEST

- reply

Unfortunately, people are a high cost and one that can be quickly changed. Damed be anything that happens in the future.

2026-06-11 20:04:22 CEST

- reply

XBox was the grant idea to protect their Windows business by bringing the same games into the living room that they develop for PC.
The PC at home is the best way to teach kids about Windows, and thus make them demand a familiar OS at work.
Thus, as long as one understands that relationship, one can see how critical Microsoft Games really is to protecting the whole business empire.

2026-06-11 19:39:54 CEST

- 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)

2026-06-03 15:00:17 CEST

As much as I want to believe in disarming AI, the cynic in me is sure it will never happen. Just like with nuclear weapons, we have opened Pandora's box and won't be able to stuff AI back into it. The only thing we can do now is to anchor our society in morals and hope to prevent abuse.
#AI #armsrace

2026-06-01 17:54:02 CEST

- reply

even if they remember, it wouldn't change their behavior. As humans the tendency to say "that wasn't us" is strong and easily translates into "that won't happen again" or "we wouldn't do that anyways."
That's why it's so important to build a society of shared power and accountability. The more people you need to actively agree the harder it gets to drift into the extremes.

2026-06-01 17:01:55 CEST

@npub1vu6…vcka IBM still has the access and the nostalgia factor to get the first few projects going. They would then have to show that they are doing the right thing and meeting the customer's needs, which is easy given their tech stack. That would quickly earn them the trust they need.
Unfortunately, with the dashboard, they more likely show customers that they neither understand the issue nor care. After all, neither the ICC nor the Dutch Civil Service would be secured by a dashboard.

2026-05-31 19:40:47 CEST

It looks like Microsoft's legal department is now outrunning its engineering unit. Maybe they should work more on fixing vulnerabilities, assuming they haven't laid off everyone yet.
#cybersecurity
https://www.tweaktown.com/news/111895/microsoft-threatened-a-security-researcher-with-criminal-charges-and-the-cybersecurity-community-isnt-having-it/index.html

2026-05-27 23:03:41 CEST

The Dutch government has stopped US companies from acquiring the supplier behind its digital national ID. The move is meant to prevent anyone from using digital tools for blackmail.
#digitalsovereignty
https://www.biometricupdate.com/202605/dutch-govt-blocks-us-companys-acquisition-to-maintain-digital-id-sovereignty

2026-05-18 05:21:35 CEST

Hey, a new medical AI use case. So much for AI making the world a better place.
#AI #medical #dental
https://futurism.com/health-medicine/dentists-ai-unnecessary-dental-work

2026-05-13 15:00:30 CEST

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 15:57:50 CEST

- reply

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

2026-05-06 15:00:27 CEST

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 05:59:46 CEST

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 16:39:14 CEST

- reply

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

2026-05-04 15:00:05 CEST

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 15:00:02 CEST

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 19:58:20 CEST

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 15:00:15 CEST

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/

2026-04-17 01:57:21 CEST

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 19:11:14 CEST

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 05:36:20 CEST

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 19:10:58 CEST

- 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 14:00:18 CET

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 04:37:38 CET

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 17:28:25 CET

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 19:17:14 CET

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 15:00:05 CET

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 20:47:47 CET

- 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 20:24:02 CET

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 04:58:14 CET

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 16:44:47 CET

- 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 22:04:02 CET

- reply

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

2025-11-25 19:33:07 CET

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 19:18:54 CET

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 15:00:10 CEST

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 15:00:10 CEST

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 15:00:00 CEST

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 15:00:26 CEST

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 17:35:15 CEST

- 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 15:00:13 CEST

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 20:56:21 CEST

- 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 18:48:35 CEST

- 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 18:31:48 CEST

- 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 21:37:55 CEST

- reply

thanks! I'll check them out.

2025-09-07 15:00:00 CEST

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 20:03:18 CEST

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 21:24:50 CEST

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 19:15:31 CEST

- 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 15:00:10 CEST

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 14:39:24 CEST

- 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.

2025-07-02 01:05:31 CEST

- 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 19:25:30 CEST

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 19:32:55 CEST

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 18:26:36 CEST

- 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 18:11:34 CEST

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 03:27:54 CEST

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 05:33:19 CEST

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 20:35:07 CET

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 14:00:11 CET

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 17:52:42 CET

- 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 17:45:48 CET

- 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 17:40:50 CET

- 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 14:01:00 CET

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 04:13:15 CET

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/

2025-02-17 00:12:14 CET

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 19:17:26 CET

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 23:34:22 CET

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 03:52:50 CEST

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 23:35:46 CEST

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 22:17:07 CEST

"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/