Skip Navigation

Posts
23
Comments
704
Joined
3 yr. ago

  • Deleted

    Permanently Deleted

    Jump
  • How is this going to work if you’re browsing the internet on your phone??

    Easy, they force you to install an app, just click the link and agree to all the permissions and cookies and stuff. It might be malware, it might be legit, who knows?

    Google copying UI workflows from literal malware developers...

  • Deleted

    Permanently Deleted

    Jump
  • There's already the app published and it has already blatantly fake reviews

  • Deleted

    Permanently Deleted

    Jump
  • Banks love to reinvent the wheel and definitely can't trust Google to be part of their workflow, as Google will 100% change UI or steps or rename the app or discontinue the app while replacing it with an identical one but with a different name but less features with no advance notice. Relying on Google for banking workflow means that one day the bank user support will be overwhelmed by requests like "the button disappeared, where is it now"

    For the rest of stuff, this system IMHO has too much friction, the bounce rate will be too high. Businesses won't like to pay for a bot detection system (it costs $1 per 1000 verifications) that will push humans away while bots pass it without problems (either by using the accessibility workaround or by using those smartphone farms in southeast Asia)

  • Wow good to know, that means Xiaomi is dead for me now.

    2000 devices globally is a ridiculously low number

  • What happened about the bootloader? Afaik there's "only" a 7 days waiting time but that is to be done only once, if the serial number is tied to the Xiaomi account that makes the request, the next bootloader unlock has no waiting time.

    And devices that run a ROM for the Chinese market only accept a bootloader unlock request from a Chinese IP address

  • Did they vibe coded the backend?

    An hacker can get:

    • GPS position
    • Email address
    • Video feed
    • WiFi password
    • Root access to a Linux client in the same WiFi network of the victim which means can change the DNS servers in the router for a mitm Attack if the default password hasn't been changed (and nobody changes that)

    And they demonstrated to the journalist..

    1. Get a list of every "smart" lawnmower nearby a nuclear plant
    2. Check the emails of the owner in LinkedIn or something like that to see who could work at that nuclear plant
    3. Have access to his home network and a video feed on a robot that can be remotely moved to other position to check the perimeter

    Scary

  • Deleted

    Permanently Deleted

    Jump
  • This is awesome news for scammers:

    1. Fake page will say "you need to scan this qr code to verify you're human"
    2. Users normally dismisses this shit, but it has become normal nowadays, take out the phone to scan it
    3. Qr code opens a page on totallynotascam.com that say "you need to install this totally safe APK on your device for verification 😉"
    4. APK passes the new useless developer "verification" as the scammer either used a hacked dev account or just paid $25 with a stolen id + stolen credit card
    5. User see the message "APK verified by Google play protect" and would totally believe the bullshit, giving all the possible permissions to the app
  • Deleted

    Permanently Deleted

    Jump
  • I'm too lazy to take out the phone from my pocket.

  • Deleted

    Permanently Deleted

    Jump
  • Well the good thing is that now Google enshittified recaptcha and now if the website owner wants to implement it, needs to pay $1 for 1000 verification requests which is crazy expensive for something that as of now it's easy to pass as a bot than as a human (bots ask the voice verification and ask a LLM to interpret that, pass recaptcha in less than a second. Humans need to click on 25 traffic lights and give up in the process)

  • Deleted

    Permanently Deleted

    Jump
  • I would never scan a qr code to verify that I'm not a bot. I'd simply close the page

  • Wow 87% disapproval is a much higher number than what I imagined.

    Reading the headline I was expecting it was like a 60% disapproval

  • Assistant was made 100x stupider over the course of the years. I clearly remember in 2020 asking basic questions like "where's Lima" and it replied "Peru", while since 2023 it replied "ok, starting navigation to Lima, Peru, estimated time 2 days and 10 hours".

    Around 2023 also they made the voice assistant completely useless by changing the responses from "for this answer open the link in your notification bar on the phone" to "sorry, I didn't understand", which was pissing me off too much. I completely stopped to use it as 9 questions out of 10 would be "sorry I didn't understand" instead of "I understood but I'm not programmed to give a voice answer, use your phone".

    Navigation by voice was a complete disaster.

    "Navigate to

    <CONTACT NAME>

    " - "Ok navigating to <a random business 1000 km away with a name slightly similar to what I said>"

    Or, it dropped the road name: you asked to go to "street name, city" and it placed the destination on the geographic center of that city

    Also the navigation instructions would have maximum priority and would play even when listening to a command.

    "Send a message: I'm going to be late

    <assistant overlaps my speech>

    " - "Sending: I'm going to be late on the next exit take the right lane then after 600 meters turn left. Ready to send it?"

    "Call Anna" - "OK, I call Daniel"

  • Yes but unfortunately it’s the only one that is coded to strip paywalls instead of archiving the page as-is

  • i adjusted hue and saturation to increase contrast

  • Until now I never saw something written in OpenDyslexic that was this unreadable

    I know that you did it with good heart for accessibility but with this colors is really unreadable

  • Background: many people vibe coded a python proxy for the official Claude code app "converting" its outputs as an openai compatible API to be used with openclaw

    edit: i did some web search and found snippets like:

    OpenClaw spawns Claude Code sessions via ACP

    and

    ACP gives OpenClaw a way to run external coding harnesses — Claude Code, Codex, Gemini CLI — as supervised child processes instead of doing everything inline in the main agent loop.

    so it looks like openclaw users were using this ACP method as a workaround to use a $20 subscription to get $1000 worth of token usage. I'm guessing that openclaw JSON example posted in the article is the configuration of this ACP server or something like that.

  • The incompetence of the day is the administration placing Australian flags instead of the UK ones

  • Google is going to push an update that will automatically install without consent or permission to all android devices and after the install, the disable toggle will appear buried in the settings, somewhere in developer settings