• 1 Post
  • 206 Comments
Joined 2 years ago
cake
Cake day: June 13th, 2023

help-circle
  • Funny thing is that ANPR is a solved issue that you can literally run on-device with minimal training. As in, you can literally run it on a $30 WiFi camera. With existing open models.

    Facial recognition is a bit more tricky but there are open AI models that translate facial data to what is essentially a hash that can be compared to other faces with high precision, and that too can run on the same hardware.

    Hell, my cheap $100 smart doorbell has built in facial recognition that doesn’t require any cloud connection or such. All on-device.


  • fonix232@fedia.iotoScience Memes@mander.xyzLatitudes
    link
    fedilink
    arrow-up
    1
    arrow-down
    1
    ·
    13 hours ago

    Even if the scale was aligned with absolute zero - like Kelvin - it would not be able to describe temperature changes in the multiples primarily because our FEEL of temperature is what matters here. And since humans live in the approx. temperature ranges of -40 to 80 (using an extended range to cover cases like the Arctic/Antarctic stations, or saunas), the best scale to use would be a Celsius scale shifted somewhat to make 0deg the most optimal neutral temperature - which is, in my opinion, 16 degrees Celsius.






  • Plagiarism should only ever be counted for explicitly unique sentences that provide actual value.

    It’s actually an ongoing debate in software engineering, due to licensing, as to what you can consider “stolen code” - i.e. plagiarism.

    In fact things went as far as to some companies employing AI-aided automatic cease-and-desist deliveries on GitHub, but the system was so badly configured, it detected even the most basic logic bits as license infringement. Things that are standardised in software development - like, for example, for loops, that happened to have generic parameter names (e.g if you were to create a graphic subsystem for displaying Views, whatever the primary implementation may be, you’d iterate through all views with a for loop, making it a generic call such as for(val view in views) { [do something here] }).

    Well this AI aided detector was so brilliant that it detected such minute coincidences of codebases as legitimate violations (as if any company could copyright generics), and sent these spurious C&Ds to dozens of git repos. What’s even worse is that the initial company’s codebase used some open source libraries that were directly attacked… for being 100% copies of their own codebase.

    IMO as long as the code/sentence isn’t a provably unique statement, plagiarism shouldn’t apply. A whole paragraph having 80%+ similarity to something unique? Now that’s worrying enough to investigate.









  • Of course it ain’t real users.

    Reddit has had a major bot problem a decade ago and little has been done to mitigate it - beyond banning legitimate users who dared to be too loud about it.

    I’ve moderated a relatively small sub, and pretty much for every legitimate post a day, you’d get 6 to 10 bot posts literally pulling an older post verbatim word for word, or maybe introducing a typo just to make detection harder…

    Reddit’s response to the issue? “Hey, why don’t you pay us ~$25 a month just so you can continue using that open source automatic bot detection system we refuse to build into the site itself?”.


  • If this was a unique instance of a potential Nazi dogwhistle I’d say let’s give them the benefit of the doubt, and consider it a mistake. After all, who of us has never ever fat-fingered some typing and accidentally pressed a letter twice?

    But the thing is, this is like, the hundredth Nazi dogwhistle around the orange turdsack to happen “accidentally” in the past few years. From the MyPillow 14.88 pricing, Kegbreath’s very obvious negative space 88 tattoo, to multiple people close to the admin just randomly throwing Nazi salutes, etc., and that’s not even counting the literal Nazi marches waving Nazi and MAGA flags together…