Troed Sångberg
@troed.swecyb.com.ap.brid.gy
Demoscene coder. Cybersec consultant. A child of the 80s home computer era.
Depending on who you are I either help secure your stuff against evil hackers or I play one and […]
🌉 bridged from ⁂ https://swecyb.com/@troed, follow @ap.brid.gy to interact
Depending on who you are I either help secure your stuff against evil hackers or I play one and […]
🌉 bridged from ⁂ https://swecyb.com/@troed, follow @ap.brid.gy to interact
https://codeberg.org/troed/meshtastic_distances for those who want to point and laugh at the LLM - but works it does.
meshtastic_distances
Local service using a connected meshtastic device to show information about nodes seen in the mesh, including calculating distances, sorting on the number of hops as well as color coded map display of the mesh around you.
codeberg.org
February 12, 2026 at 8:22 PM
https://codeberg.org/troed/meshtastic_distances for those who want to point and laugh at the LLM - but works it does.
@nopatience Have not, but I know this is a relevant Matrix room: https://matrix.to/#/#freestream:matrix.org
February 12, 2026 at 10:56 AM
@nopatience Have not, but I know this is a relevant Matrix room: https://matrix.to/#/#freestream:matrix.org
Reposted by Troed Sångberg
If you too want to feedback on the idea of Open Source CRA attestations (basically projects officially saying that they are "good projects" in a CRA sense), here's the survey
https://dialog-cybersicherheit.limesurvey.net/113884
https://dialog-cybersicherheit.limesurvey.net/113884
Powered by LimeSurvey – The Freshest Online Survey Tool
Create surveys in seconds with LimeSurvey. Easy to use, secure, and trusted by professionals worldwide. Get started free and unlock fresh insights today!
dialog-cybersicherheit.limesurvey.net
February 10, 2026 at 9:45 AM
If you too want to feedback on the idea of Open Source CRA attestations (basically projects officially saying that they are "good projects" in a CRA sense), here's the survey
https://dialog-cybersicherheit.limesurvey.net/113884
https://dialog-cybersicherheit.limesurvey.net/113884
@nopatience I wouldn't have known about it in that detail either unless it so happened it's part of a presentation I'm currently doing for a customer :D
However, the argument could be made that cybersec products with "admin access" should be class II, not I, requiring third party assessment.
However, the argument could be made that cybersec products with "admin access" should be class II, not I, requiring third party assessment.
February 9, 2026 at 5:42 PM
@nopatience I wouldn't have known about it in that detail either unless it so happened it's part of a presentation I'm currently doing for a customer :D
However, the argument could be made that cybersec products with "admin access" should be class II, not I, requiring third party assessment.
However, the argument could be made that cybersec products with "admin access" should be class II, not I, requiring third party assessment.
@nopatience I think this is exactly what the CRA tries to do. It's all about the classification those security products end up having. If they're class II, they will need third party assessments to be able to sell in the EU.
https://www.eucyberresilience.com/product-category
https://www.eucyberresilience.com/product-category
EU Cyber Resilience Act | Products Covered - Critical, Non-cricical
Know which product categories are covered by the Cyber Resilience Act. Understand whether or not your product is listed as important or critical.
www.eucyberresilience.com
February 9, 2026 at 5:26 PM
@nopatience I think this is exactly what the CRA tries to do. It's all about the classification those security products end up having. If they're class II, they will need third party assessments to be able to sell in the EU.
https://www.eucyberresilience.com/product-category
https://www.eucyberresilience.com/product-category
Reposted by Troed Sångberg
🇪🇺#chatcontrol Update: Draft is out! Text scanning is gone, but mass hash scanning remains. That’s 99% of reports! Almost NOTHING changes. 🛑
⏰ Amendments deadline is Tuesday!
📨 Email MEPs now: https://fightchatcontrol.eu/#contact-tool
ℹ️ Info […]
⏰ Amendments deadline is Tuesday!
📨 Email MEPs now: https://fightchatcontrol.eu/#contact-tool
ℹ️ Info […]
Original post on digitalcourage.social
digitalcourage.social
February 7, 2026 at 12:40 AM
🇪🇺#chatcontrol Update: Draft is out! Text scanning is gone, but mass hash scanning remains. That’s 99% of reports! Almost NOTHING changes. 🛑
⏰ Amendments deadline is Tuesday!
📨 Email MEPs now: https://fightchatcontrol.eu/#contact-tool
ℹ️ Info […]
⏰ Amendments deadline is Tuesday!
📨 Email MEPs now: https://fightchatcontrol.eu/#contact-tool
ℹ️ Info […]
@nopatience I agree fully - static models don't work so that paradigm needs a breakthrough.
I did dive down into neuropsychology and theories of consciousness a bit over a decade ago - this blog post contains my views on today's LLMs and "AGI" […]
I did dive down into neuropsychology and theories of consciousness a bit over a decade ago - this blog post contains my views on today's LLMs and "AGI" […]
Original post on swecyb.com
swecyb.com
February 3, 2026 at 8:21 AM
@nopatience I agree fully - static models don't work so that paradigm needs a breakthrough.
I did dive down into neuropsychology and theories of consciousness a bit over a decade ago - this blog post contains my views on today's LLMs and "AGI" […]
I did dive down into neuropsychology and theories of consciousness a bit over a decade ago - this blog post contains my views on today's LLMs and "AGI" […]
@nopatience The problem, I think, is that those who don't see how LLMs could ever "become intelligent" are falling for the classic AI trap, which can be explained like this:
Throughout the 20th century, as soon as we understood how something worked we stopped calling it AI.
A good example […]
Throughout the 20th century, as soon as we understood how something worked we stopped calling it AI.
A good example […]
Original post on swecyb.com
swecyb.com
February 3, 2026 at 7:31 AM
@nopatience The problem, I think, is that those who don't see how LLMs could ever "become intelligent" are falling for the classic AI trap, which can be explained like this:
Throughout the 20th century, as soon as we understood how something worked we stopped calling it AI.
A good example […]
Throughout the 20th century, as soon as we understood how something worked we stopped calling it AI.
A good example […]