ChromeOS is a distinct brand, people expect a particular experience.
Windows is a distinct brand, people expect a particular experience. Giving users a restricted experience while using the same branding leaves a bad taste in the mouth.
Thats why all the windows-ce laptops failed miserably too, as did windows mobile. People bought it expecting the normal windows desktop experience, got something inferior and incompatible, and word soon spread.

A professional is paid, an amateur is not.
Absolutely nothing to do with ethics.
A professional is less likely to have ethics because money is their primary motivator, whereas an amateur may have more altruistic motivations.

By drawing a mouse, he copied the appearance of a mouse. So he copied something that already existed and expanded upon it.
Without public domain, you have no source material to draw from, virtually everything is a derivative of something else.

Yes a mouse is a creation of nature, copyright does not exist in nature it's an artificial construct. Why should there be artificial constructs to protect some forms of work but not others?
Perhaps the copyright for all natural things should go to the government too, or perhaps an environmental fund for protection of nature?

And while the mouse is a product of nature, the steamboat is not. Should disney have paid licensing fees for his use of the image of a steamboat? Someone designed and built steam powered boats.

If you lock away the public domain in perpetuity you will absolutely stifle future work. Meanwhile people managed to create things just fine before the concept of copyright existed.

It's also not just reading a script, it can simultaneously have a feed of the original audio which allows it to better replicate it.

Replace "AI" with pretty much any technology.
Why should dubbing actors receive any special treatment? What about blacksmiths, factory workers, crop harvesters, and all manner of other professions that have been eliminated or severely reduced by advances in technology?

In some developing countries you get a guy in a booth who collects parking fees, in developed countries there is generally a machine. Should we rip out all those machines to protect the jobs of those who sit in booths to collect parking fees?

Or they clone the voice of the original actor, so the dubbed voice more closely matches the appearance.

That's exactly what happened in some places. There were various restrictions set up to impede the spread of cars, for example requiring that someone has to walk in front of any car waving a flag.

Also why should dubbing actors receive special treatment compared to all the other professions that either have been, or will be rendered obsolete or reduced to a much smaller niche by technology?

Non technical people are not going to buy and operate a high end server with 128GB ram, and then manually deploy hundreds of virtual machines on it...
But buying an off the shelf NAS appliance, many people can and do exactly that.

Many of these NAS devices will let you install additional packages as simple as choosing them from a list, you get things like owncloud, torrent clients, media servers etc as well as the built in capabilities of the NAS.
Adding an SSL certificate is usually also a built in capability, and the NAS vendors also typically provide a dynamic DNS service so you can access the device remotely by name with proper certificate validation.

Here's just a sample list of a few thousand synology NAS appliances which are publicly accessible, with valid SSL certs via IPv6-only:
https://www.ev6.net/v6sites2.p...

This setup is not default, these people have explicitly enabled external access and chosen a dyndns name for their appliances.

There's lots more. Self hosting is actually very common.

Use devices which support open standards such as Matter.
Connect them to an open source hub such as HomeAssistant.
Keep the devices connected to HA via an isolated network
Self host your HA instance using IPv6, and only allow remote access to the HA hub rather than the individual devices.
Use an SSL cert for the web interface.
Implement access controls if you don't want it to be public - eg limited it to the address range of your mobile telco, put it behind a VPN etc.
Verify firmware updates before applying them to any devices - ensure they only fix bugs and/or add features, not strip out functionality you need.

Such systems are extremely widespread because a lot of people are stuck behind NAT and cannot self host, and there are too many legacy networks out there without IPv6. These devices hook people with easy initial setup.

Charging a subscription is actually the best case, and ceasing support is the second best. Continuing to run the service but selling your information is a lot worse. Think how many dodgy off-brand CCTV cameras you can get which send your video streams to a server in China... A country where voyeur porn is popular.

It's _ALWAYS_ going to happen like this. Hosting these servers costs money so there absolutely has to be an ongoing revenue stream to pay for the hosting costs.

You can do remote access yourself for free too, the subscription service is for users stuck behind NAT on legacy networks without IPv6 who can't self host.

My instance of HA is accessible globally, but only via IPv6.

This is a side effect of widespread NAT. How do you access your devices from outside?

If you're stuck behind NAT you simply can't without a third party service to relay the traffic, so device manufacturers are incentivised to provide that service because setting up something custom would be beyond the ability of most users.

But hosting such a relay service costs money on an ongoing basis - how do you expect such a service to be funded? Charging you a subscription fee is actually the most honest way, at least then you have a contract between you and the supplier and a clear revenue source that pays for the service. The alternative is that they seek *other* sources of revenue to pay for the service - this likely includes selling your data etc.

Some better devices use a simple dynamic dns service - eg fritzbox routers, synology nas devices and a bunch of others and even registers a valid SSL cert so you can connect back directly to your device. But this only works if you're able to open up ports to external connections. A lot of people can *only* do this via IPv6 so you then suffer problems if you go somewhere that doesn't have proper connectivity. There are thousands of people actively using these dyndns services in IPv6-only mode, for instance:

https://www.ev6.net/v6sites2.p...
https://www.ev6.net/v6sites2.p...
https://www.ev6.net/v6sites2.p...

Even this has challenges because most consumer routers block inbound connections by default, and the way to allow inbound traffic differs depending on the type of device, or in some cases isn't available at all.

DNS cache poisoning was around long before 2008. The non random source port reported in 2008 has exploit code available many years earlier, for instance this web archive shows it was available as early as 2001:
https://web.archive.org/web/20...

This version takes advantage of the non random source port, and brute forces the query id.

There were also other bugs in earlier versions of bind - such as non random query ids, and caching of additional records outside of the current domain etc.

After the publicity in 2008 the "patch" was to use random source ports to increase the entropy, but DNSSec was always the proper long term fix.
Bot now with the prevalence of NAT the random source ports are often broken, as many NAT gateways will rewrite the source port from the client to a predictable value.

IPv6 provides two benefits here - first the lack of NAT, and second the large address space means you can bind many addresses to source DNS queries further increasing the entropy (you can even dedicate a whole /64 to random source addresses of dns queries) - but this only works for domains with IPv6 nameservers.

Getting the LLM to write all the code is management's wet dream to replace expensive meat sacks with cheap LLMs, only it doesn't work that way.
Neither does the model of letting the LLM write everything and have someone else review it, as you're right it's better to just let that person write the code in the first place.

But there is a middle ground.
The LLM is a tool, used by the expert to make his job easier and more efficient.
The LLM gives suggestions, some might prove useful and some might not.
The LLM can complete tedious repetitive tasks. Sure a skilled developer could also write scripts to automate many tedious tasks, but that in itself takes time.

I doubt many of those engineers will just sit around. They're either going to be in the military fighting (and likely killed), or they will flee the warzone and claim asylum somewhere else.