One very unfortunate curse we are stuck with is LLM providers pulling the rug from under us and swapping models with quantized versions during peak demand or on monthly plans. You never know if that's the reason AI starts writing shitty code or it's just the context size that has grown. On top of that there is the queuing mechanism which makes tasks take longer. Sometimes AI is super fast and sometimes a simple task takes 10, 15 or even 20 minutes of "thinking". You may not realize this if you are just using your same subscription but if you switch between providers you can immediately tell. I advise everyone to invest $5 in openrouter, deepseek and qwencode. Just do an experiment and use same model (deepseek-4.1 flash) asking same prompt on same code, then look at the result. Spoiler: you will see trash coming out of openrouter at 1-5 tokens/second, you will see slightly faster and better quality output from qwencloud and when you get to deepseek API where you are paying per token you will not believe this is the same model. Because it isn't.
Dude, coat hanger joint is broken. The free chatgpt gives openscat file I can just paste and convert to stl. I didn't even measure the diameter of the pipe, apparently those are standard. In 10 minutes I have a new 3d printed part to fix my coat hanger. No CAD UI app subscriptions, not reading CAD books or whatever. That's real life. Rest is in your mind.
People don't have time (or even mental capability / even more time) to learn CAD but they have a task at hand which they want done. If AI can generate a part which is "good enough" for the user at a cost that the user considers reasonable (I.e. zero) then AI gets used. If a part fails thats rarely a concern because these parts are just 3d printed and likely cost single digit cents in filament. User can always print a new one.
We are doing AI screening. Candidate works with AI for 15 minutes and session is recorded. We review real output not AI generated resume. If candidate can get AI to apply and produce meaningful output then maybe we should hire that candidate. Currently outputs are manually reviewed but easy enought to have AI screen outputs.
Oh if only cut and paste was broken... In Windows and Linux copy and paste itself is broken. Sometimes you copy with mouse, sometimes you copy with keyboard, and if it is a recent version of MS Office likely neither will work. Sometimes you paste with Ctrl+v sometimes with Ctrl+shift+v, sometimes with mouse, sometimes it just won't work. Such a simple operation should have been completely owned by the OS.
Nonsense. Many if not all legit Claude users are using Claude Code inside their Cloud servers. How else would you use it anyway? For just local dev? That's so 2000 and late bro.
There is no bona fide desire to solve the problem by the social media companies which are profiting from our kids. As a parent, I can never find the option to enable parental controls on apps like YouTube (especially on TV). This option is easy enough to implement locally without requiring any accounts or OS-level age support. However, clearly the goal is not to protect the children but to enable mass scale surveillance and profit from it.
Even when I manage to enable parental controls by using a dedicated app or account (what nonsense is it that I need to create an account to restrict access!!!) the content being served is dubiously appropriate for the age group, aside from having no value whatsoever other than the intended goal of perpetuating addiction.
I have done this, alias codex --yolo -p . It's very helpful not having to remember every odd command and its parameters. It's a bit more typing but I type faster than invoke and scan through man pages.
We always see consumers blamed for choosing price over quality. How about retailers taking the blame for dumbing down or removing product specs? If two items look identical but one costs more than the other how can consumer be blamed for choosing the cheaper ones? Especially in the age of LLMs, it you are building a quality product you need to include a spec sheet of what makes your product better than the competitor. Not dumbed down marketing speak like "lasts longer" but specific details justifying the premium, like "zippers made in Japan" or the stitching density, fabric specifics, etc. Consumers who care can use LLM to understand what it all means and make informed choice. But when the information is hidden consumer will choose the cheapest option.
LinkedIn also violates SPAM regulations on a regular basis. Despite of me having disabled all emails from this service I consistently receive promotional emails. LinkedIn defines a new "type of promotional email" for which it assumes it has implicit consent to send unsolicited emails and proceeds to do so. It then has a fake compliance apparatus by allowing the victim to once again "unsubscribe" from the newly created email subscription which they never consented to on the first place. I really hope there is a class action and these scumbags get fined.
It's terrible. Fake money is fueling the exhaustion of real resources in search of questionable outcomes ("AGI"). Imagine if all of these money were invested in curing cancer.
This! Absolutely needed regulation. Why is it that such a clearly beneficial and necessary piece of legislation is not making its way through the legislative bodies of the world while age checks somehow magically appeared universally?
First we need to understand what the root cause of the problem really is then we can discuss solutions. All we've been told is that "Android users are getting scammed, we are going to make side loading impossible". There is no clear cause and effect established, no data shared with the public on what percent of scams were caused by sideloaded apps and how the scams actually operate for us to be able to accept the solution.
Why would you say "semi-legally"? Nothing "semi" here. What is "semi-legal" is making hardware e-waste by deciding it is "no longer supported". It is "semi" legal because it is legal under the corrupt political systems in most of the world but is criminal against humanity and the planet we all call home. In that sense if you can prevent e-waste trough any means you are a hero.
Can one be opposed to age verification in the OS and yet totally happy that Meta got this fine? There is a very big difference between e2e encryption /telephone and social media. Social media is more akin to a phone book. I do not recall there ever being any phone books listing minors. That's completely unacceptable and unnecessary. I am totally OK with phonebooks (or their modern digital equivalents which enable people discovery and user generated content discovery) to abide by the same KYC rules as banks. And be only for adults. Your kids using e2e encrypted messaging to communicate with their friends whom they have met in person? Nothing wrong with that, we all have the right to privacy. Kids listing their contact information publicly? Absolute no.
What techies are missing is that AI doesn't make it possible for mom and pop shops to create and manage a website but it levels the playing field for enterprenuers. We can't expect plumbers and restaurant owners to spend 12+ hours fighting with AI website builders just to get a cookie cutter-website that is nothing more than a brochure. Nor can they fork thousands of dollars for web design agencies and spend months in mindless meetings. Thanks to AI now there is a way: small mom and pop local website builders can offer a white-gloves solution that scales and drives revenue for the SMBs.
You should mention that it uses PeerJS. That's not without servers. PeerJS runs their own signaling server which you are relying on to connect your peers.