A while ago I had a discussion with my friends that it is possible that in the future if 5G is sufficiently cheap, smart tvs come with a 5G SIM so they can force ads and updates even if you refuse to connect it to WiFi. I wonder if this will ever be a real thing. Either 5G, 6G or whatever comes next.
I fear this won't even required SIM cards. I'm worried that Apple's Find My and Amazon's Sidewalk networks are the precursors of this: They're effectively company controlled p2p networks that lets the company use their customers' internet access points like a commodity. If one customer refuses to give a device access to the internet, they could use that network to route it through the access point of another customer.
Also, personal experience: My own ISP (in Germany) experimented with some similar stuff a few years ago: They mandated use of their own home routers where only they had root access. At some point, they pushed an OTA update that made the router announce a second Wifi network in addition to the customer's. This was meant as a public hotspot that people walking down the street could connect to after installing an app from the ISP and buying a ticket.
The customer that "owned" the router wasn't charged for that traffic and the hotspot was isolated from the LAN (or at least the ISP promised that), but it still felt intrusive to just repurpose a device sitting in my living room as "public" infrastructure.
(The ISP initially wanted to do this on an "opt-out" basis, which caused a public uproar thankfully. I think eventually they switched to opt-in and then scrapped the idea entirely.)
The ISP named Free in France also did this a while ago.
It was fairly well implemented I think: separated from your network, bandwidth was limited (to avoid impacting the host), you could opt-out (which meant opting out of using the guest network), joining the wifi was automatic if you had a cellphone with the same ISP and it was the same "guest" network for all routers so in big cities, you could rely only on this to access Internet.
It was stopped a few years ago when they deemed cellular network was reliable enough to not need the guest network.
> Also, personal experience: My own ISP (in Germany) experimented with some similar stuff a few years ago: They mandated use of their own home routers where only they had root access. At some point, they pushed an OTA update that made the router announce a second Wifi network in addition to the customer's. This was meant as a public hotspot that people walking down the street could connect to after installing an app from the ISP and buying a ticket.
Not sure if you're referring to Vodafone, but Vodafone Germany definitely does this. You can opt out of allowing public access via your personal router, but this opts you out of being able to use other people's routers in the same manner.
If it had Ethernet ports I'd be tempted to just use my own wifi router and put the ISP's Trojan horse in a Faraday cage. All ISP-controlled hardware should be treated as just another untrusted WAN hop.
When I signed up with them, they were actually trying to withold access to the config web UI from customers and then charge extra just to enable Wifi. My response was exactly that - "fuck that" and put my own router in front of theirs.
(That was years before the other incident - since then they had dropped that idea and "generously" given customers access to the config UI)
You might be interested to read about the findings by Ruter, the publicly owned transport company for Oslo. They discovered their Chinese Yutong electric buses contained SIM cards, likely to allow the buses to receive OTA updates, but consequentially means they could be modified at any moment remotely. Thankfully they use physical SIMs, so some security hardening is possible.
Of course, with eSIMs becoming more widespread, it’s not inconceivable you could have a SoC containing a 5G modem with no real way to disable or remove it without destroying the device itself.
I hope this happens, because with the security track record of these companies it would mean free Internet. These would quickly become web torrent video portals.
I keep being surprised if why that is not a thing yet. Amazon launched whispernet with ads on the discounted Kindle years ago and I was totally predicting more companies jump on that.
Whispernet was a whole different thing, and it dates to the very first Kindle.
This Kindle did not have things like idle-screen advertising. That wasn't yet a thing yet.
These first edition devices were available with unlimited data access (IIRC in the US via AT&T) on cellular networks without a separate subscription. It was slow (everything was slow back then), but it would let a person download a book or have a look at a web page (with the very limited browsing that was possible with e-ink and a CPU that was meant more to barely sip power than to render megabytes of CSS and JS).
The expense of the data access was built into the one-time purchase price, and the hope was that people having the ability to buy books from "anywhere" would snowball into a thing that was both very popular and profitable.
It was simple and, functionally at least, it worked very neatly: Take new Kindle out of the box, switch it on, and download a book with it. No wifi or PC connection or other tomfoolery needed.
That was back in 2007 -- a time when many people still had landlines at home if they wanted to make a phone call, or a dumb phone in their pocket if they wanted to do that on-the-go. Some folks had Blackberries or connected Palm devices, but those things were rare.
And the Internet, and indeed Amazon itself, was a very different place back then. Having an Internet connection that was very quietly always available on a Whispernet-equipped Kindle was pretty cool at that time.
---
Sidewalk is a different kind of network. It uses consumer devices (like Echo Dot speakers) to act as Sidewalk bridges. This generally works at a low frequency (900MHz-ish), to provide a bit of relatively slow, relatively long-range wireless network access for things that are otherwise lacking it.
The present-day operation works like this: Suppose I've got some Amazon Echo speakers scattered around my house. If a neighbor's Internet connection is on the fritz, then their Ring doorbell can use a tiny slice of my Internet bandwidth using Sidewalk via one of my Echo speakers to keep itself connected to the network and thereby still function as a doorbell.
Or, maybe their Ring doorbell is out on a post by the gate, where their wifi coverage sucks. If it can gather up a little slice of 900MHz Internet access from anyone's near-enough Sidewalk bridge, then they've still got a button for their gate that notifies them on their pocket supercomputer when some visitor is waiting out there. They don't even necessarily need to plan it this way in order for it to Just Work.
Or, what GP was referring to: Your hypothetical new smart TV might use the neighbors' Sidewalk-enabled device(s) to update or patch itself, produce new ads to show you, and/or send telemetry back home to Mother. It might do this even without you ever having deliberately connected it to any network at all.
---
Either thing (some modern equivalent to Whispernet, or the already-loose-in-the-wild Sidewalk system) could potentially be utilized by smart TVs and other devices to get access to the network and simply sidestep the oft-repeated, well-intended, and somewhat naive mantra of "It can't have Internet access if you never connect it!"
And it will require an uncovered camera and microphone, or it won't display an image. Sony TVs already come with "optional image optimization" cameras.
This is why I self host my blog. My email. This is why i try to stay away from the convenience of big tech. It is not the first time this happens and it will not be the last.
100% agreed. The AI "bubble" has been "going to pop" for the past year or more. Sounds a lot like Bitcoin, that was definitely popping any second now but actually did not really pop and you can still buy Bitcoin at prices higher than 6 or 12 months ago.
I need people to understand that if everyone thinks something is going to pop then it won't pop because people don't put money in risky assets. Assets only pop when everyone thinks they are absolutely safe and they can never go down in price, like what happened with the housing market.
Facebook knew for years its social media was hurting the mental health of teenagers, and not only they doubled down on it because it makes money, they will also face zero consequences.
This will never ever happen in the US because free speech is obviously more important than children's mental health. Allowing 14 YOs onto the Internet is but a mere side effect of the Constitution.
My current favourite is Koofr. European and rclone native. If Koofr ever goes out of business, your Koofr Vault data is easily decryptable with open source tools.
What I have observed is that a project's popularity has almost zero correlation to the amount of effort behind it, or how well done it is. You just need to get lucky. Maybe one person noticed your utility, invited you to some podcast, and then your repo explodes in popularity.
One thing I can tell you is that you should stop seeking to get notoriety, because the effort you are spending on that is effort you don't spend in improving your open source tools. I know it is probably quite frustrating, but you just have to continue building stuff in public and hoping to get noticed at some point.
Thanks for the advice, and i think you are 100% right.
i am already working as a software developer, so the only reason that i want it to get noticed, its because i just think its cool to have made something other people find useful. And i love programming, and this project is definitely making me a better software developer, with or without tons of downloads or stars :)
Just wanted to make sure that i didn't do anything miserable wrong
Your employer needs you because writing code was never the hardest part of programming and software engineering in general. The hardest part is managing expectations, responsibilities, cross-team communication, multi-domain expertise, corporate bureaucracy and pushing back against unnecessary requirements and constraints. None of which LLMs can solve, and are especially terrible at pushing back.
I agree. I remember once the full especification I got was
> Enough
After talking for 4 hours and 3 coffe cups, I got enough corner cases and main case to understand what they wanted. 1 week later I got a list of criteria that can be programmed. 5 years later most of the unusual but anoying rought corners were fixed. We still had a button to aprove manually the weird cases.
I wonder if it is safer than a Pixel with Graphene. It is a completely different OS, of course, and many of the Graphene OS hardening features like isolating Google Apps, are not necessary in a phone that literally has no Google Apps, but Graphene also has other more generic privacy protection mechanisms including a duress pin or inactivity reboots allowing the phone to eject the decryption keys from memory. I might actually buy one and try it.
reply