I think one of the major reasons why it’s not here is because most AI tools are great for getting a prototype built up but undertaking a program like Photoshop which we can assume has a couple millions of lines of code is actually not easily replicable by vibe coding.
Can it be done? Probably. But the token cost to do so would be astronomically expensive. You still need a human (read prompt engineer) to steer the AI. Even small ideas I’ve thrown at it to test viability is often plagued by code that just simply doesn’t work. I am constantly having to send it back and say “this doesn’t work” and it will eventually figure it out but that is more wasted tokens just to get something that doesn’t throw an error during the build process.
One-off specialty programs are absolutely going to feel the heat. Take a multi-tone generation application (think Motorola pagers). You can now ask an AI to create that program for you, complete with tone generation and .wav recordings to use later.
Large programs are relatively safe for the moment just because of scale. But any small application that is usually behind some sort of paywall or license is absolutely going to be threatened since it’s no longer difficult to throw together a program in an evening and have 100% of the features you want and none of the features you didn’t need or want.
Starlink by virtue of being your ISP would have access to any DNS queries you send over the Internet over UDP port 53 in plain text. Starlink is also able to redirect those queries to their own servers. Even if you manually specify 8.8.8.8 or 1.1.1.1 Starlink can redirect traffic to their own DNS servers and return responses as if they came from those servers.
By itself DNS can tell a pretty detailed picture about you and what you do on the Internet without the need for SSL inspection or other deep packet inspection techniques.
Even as a licensed ham it's getting increasingly difficult to even get hardware that allows utilization of frequencies I'm duly licensed to transmit on in the 2.4 GHz band. Short of building and designing your own transmitters it's become impossible to repurpose hardware like it was before. Our club has aging M2 Rockets from Unifi that were modified for this use that are now decaying and dying. It's unfortunate too because once these stop working that's it. A few club members have been championing GLiNET but same problems. They are relying on older models which weren't as locked down and already show signs of suffering the same fate as the Rockets.
The other more compelling reason why people would have a rooted phone is to run ROMs that may still be providing OS support where the stock OS has been abandoned or EOL'd by the developer.
Having an unlocked bootloader at the minimum would be required in those scenarios. It actually saves hardware that still works from ending up in landfills.
This is the way. Widevine is a cancer that only serves to lock down the browser market to a small handful of web engines that have been approved by Google. If your browser isn't based on Chrome, Firefox, or Safari you're out of luck.
Most people will not use a browser that can't open youtube videos and they know and exploit this with extreme precision.
Yes. I tried using Chrome on Linux just to watch movies that I purchased on Youtube at HD/4K and watched as the stream was limited to 240P. IMHO regardless of what Google says in their ToS they have already broken the trust agreement by not providing what I paid for. Regardless of what the studios want, all this does is push me back towards piracy because once again the industry fails to understand that piracy is a accessibility problem, not a financial problem. If I pay for 4K then regardless of where I want to watch that movie it better be in 4K, that's what I paid for. Google hides behind their ToS to get around the fact that they sold me a product then failed to deliver.
> ChromeOS gets 1080p/4K not because it has massive market share but cause the hardware and boot chain are locked down by the almighty Google.
ChromeOS is based on Gentoo Linux underneath just very stripped down and Googlefied. It's the same BS that Bungee pulled with Destiny 2 and Linux. If you so much as dared to run Destiny 2 on Linux you would be banned. Stadia used Linux but because Google controlled the platform they allowed it to be played there.
These are the games they play to make other platforms that aren't MacOS/Windows appear like they are incapable but in reality it's just corporate greed and grift.
I think the difference is really more noticeable if you're on a limited connection. For example, on Starlink I only have 50 GB to play with. It's entirely ineffective if the browser downloads the ads and only scrubs them out of the view after the fact. Same with anybody using a mobile hotspot over LTE. In those situations bandwidth is super limited (I have 5 GB of hotspot data a month) unless you can convince the carriers to zero-rate data pulled for advertisements (they won't) I'll continue blocking ads before they can be loaded.
Edit: and I'm not on some cheap MVNO, I'm paying over $80 a month with AT&T on their post-paid plan. The phone gets unlimited data but any other device I may need to share that connection needs to be as efficient with bandwidth as possible. Only Firefox and derivatives provide proper ad blocking at this time.
Oh I agree 100%. I also play for my search engine so it's definitely not a lack of interest in doing so. I agree with your point as well. Get rid of the money vultures in the C-suite who are paying themselves exorbitant salaries and hand that money over to the Firefox devs. Give them the runway necessary to bring on more developers that would give Firefox the attention it needs to keep up with Chrome/Chromium and maybe start playing with the idea that if you want the latest updates when they release you pay for the browser. If you don't need immediate updates you'll get the deferred releases under a 1-2 month delay or whatever they deem fit with security fixes obviously being backported to keep those who refuse to pay happy enough to not abandon the browser entirely.
Unfortunately this is not unexpected because Mozilla needs to continue receiving money to survive and unfortunately nobody wants to have the tough conversation about paying for a browser so when whoever is funneling money into Mozilla (Google) says you need AI in your product you have no choice but to jump.
I think their logic is a bit wrong here. Microsoft is a "trusted" entity. Trust doing a lot of heavy lifting here, and even they had to roll back their AI ambitions after seeing the lackluster adoption rates of people using their AI features. The trust part just doesn't matter. It's the principal that we've had browsers for over 20+ years and we never needed AI in our browsers. I would quickly abandon Firefox for an alternative in a heartbeat that doesn't include AI in it.
The uncomfortable truth for all these companies though is that most people simply do not need AI in the places they are shoving it into. Like why does notepad need AI?
Fuck that noise. The places that shadow ban and encourage self-censorship do not deserve your traffic nor your content.
Start voting with your voice and your (digital) feet. Don't be sheeple. Keep the Internet weird. It is not on us to censor ourselves to protect the feelings of snowflakes who get all bent out of shape because of something someone said.
I think that's the model everyone (eg: your average consumer) likes and is most familiar with.
Advertisements on TV are tolerated because it's not targeted and just a random broadcast to everyone who might be interested.
Anxiety that certain things may be used against you in the form of advertising is a very real problem people face. A teenager who is researching transgender topics would be potentially frightened if the TV station the family usually watches that never shows any ads about transgender health suddenly started pushing those kinds of ads and the parents disapproved.
The problem lies in the tracking technology. In order to make informed decisions about what people might potentially be looking to buy includes learning that individual's preferences and the rather unfortunate part of everything being online is that includes everything you could possibly imagine. Everything from the type of job you have right down to your porn preferences and other proclivities are potentially up for grabs and I'm sure the vast majority of people would feel greatly unnerved if Google, Facebook or whoever showed them just what kind of person the algorithms have determined them to be just by the data they've already collected.
Can it be done? Probably. But the token cost to do so would be astronomically expensive. You still need a human (read prompt engineer) to steer the AI. Even small ideas I’ve thrown at it to test viability is often plagued by code that just simply doesn’t work. I am constantly having to send it back and say “this doesn’t work” and it will eventually figure it out but that is more wasted tokens just to get something that doesn’t throw an error during the build process.
One-off specialty programs are absolutely going to feel the heat. Take a multi-tone generation application (think Motorola pagers). You can now ask an AI to create that program for you, complete with tone generation and .wav recordings to use later.
Large programs are relatively safe for the moment just because of scale. But any small application that is usually behind some sort of paywall or license is absolutely going to be threatened since it’s no longer difficult to throw together a program in an evening and have 100% of the features you want and none of the features you didn’t need or want.