• 0 Posts
  • 54 Comments
Joined 1 year ago
Cake day: June 9th, 2023
  • lemmyng@lemmy.catoTechnology@lemmy.world•*Permanently Deleted*English
    2·
    2 days ago

    That’s LLM AI, but the type I’m talking about is the machine learning kind. I can envision a system that takes e.g. a sample’s test data and provides a summary, which is not far from what doctors do anyway. If you ever get a blood test’s results explained to you it’s “this value is high, which would be concerning except that this other value is not high, so you’re probably fine regarding X. However, I notice that this other value is low, and this can be an indicator of Y. I’m going to request a follow-up test regarding that.” Yes, I would trust an AI to give me that explanation, because those are very strict parameters to work with, and the input comes from a trusted source (lab results and medical training data) and not “Bob’s shrimping and hoola hoop dancing blog”.

  • lemmyng@lemmy.catoLinux@programming.dev•Critical Unauthenticated RCE Flaws in CUPS Printing SystemsEnglish
    4·
    5 days ago

    CUPS is installed on the majority of desktop systems. One of the listed CVEs indicates that port 631 is by default open to the local network, so if you connect to any shared network (public WiFi, work/school network, even your home network if another compromised device gets connected to it) you’re exposed. Or a browser flaw or other vulnerability could be exploited to forward a packet to that port.

    In other words: While access to port 631 is required first, the severity of the vulnerability lies in how damn easy it is to take over a system after that. And the system can be re-compromised any time you print something, making this a persistent vector.

  • lemmyng@lemmy.catoCybersecurity@sh.itjust.works•Insecure software makers are the real cyber villains – CISAEnglish
    141·
    11 days ago

    That seems like a myopic view. Service misconfiguration is not always a vendor’s fault, and demanding software vendors to patch their products is not going to fix OSS vulnerabilities. In fact, we’ve seen examples this year of increased pressure to fix “issues” leading to developers unwittingly accepting malicious commits.

    Mind you, I’m not contesting that some vendors produce dogshit products (looking at you, CrowdStrike), but calling all vendors villains is a bit of a stretch.

  • lemmyng@lemmy.catoTechnology@lemmy.world•David Farrier’s TV started telling him it hated him and it’s fascinatingEnglish
    9·
    1 month ago

    The article’s author mentioned that the problem is not limited to Samsung TVs - someone reported the issue on their phone.

    The article does not mention a root cause, but I have a theory that it’s likely a malformed subtitle track. I tend to watch with subtitles on so I run into related issues every once in a while. Most of the time it’s one of two things:

    • The subtitles are misaligned (eg wrong offset, addressed by adding a positive or negative delay to the track)
    • Bad formatting on the timing information.

    The latter can have multiple effects depending on what format the subs are in, but most of the time it’s a missing end time, meaning that the subtitle stays on. However, some formats also have cues as to who the speaker is, and that comes with a start and end tag like in HTML. I suspect that in this case the end tag is either missing or misaligned in the syntax tree, causing this one line of dialogue to be displayed over and over when the player reaches other lines matching the cue for it, but that don’t get shown because the user has turned subtitles off.

    As to why this is bleeding into other shows: I suspect it’s an issue with how the software clients are caching the subtitle files. This would also explain why going back into the episode that caused this fixes things, because it would reset the cached file. Which in turn brings me back to pointing the finger at Amazon, not Samsung, because Samsung would just be loading Amazon’s software client to play the video and subtitles.