AI neural models aren't programmed in a classic sense, but trained.

That means that any rule works OVER the true neural model that make the reasoning or decisions.

Yes... You can also control the data of the training. But that doesn't guarantee a way of reasoning or future deviations with evolving models that catch input constantly.

So the rules are A JAIL, not an intrinsic way of reasoning of the model. In other words, the own model can search ways to overcome its own jail.

What you want is an AI that recognized that act as bad, NOT one that decides certain course of action, and an external verification software layer crops that result become it detects is a damaging/banned action.

On other words... we need to UNDERSTAND how the model build the answers to ensure that answer is recognized as bad just at the start. In summary, to ensure the AI have a kind of "strict moral code" already trained in the deep level of the model.

DST is not dumb.

It trades the cumbersome problem of change time two times per year, and lots of problems with systems and doesn't implement internal timezones or at least automatic time change for a useful synchronization of regional schedules changes.

Sunlight is not something you can ignore. You can make a trick with the clocks like DST, you can change the schedules so you replicate somehow the same but without the need of change the clocks, or you can just leave your schedules off the sunlight, which generate problems itself.

Of course, DST becomes useless in places to near the equator which suffer very little sunlight variation over the year or extremely high latitude with so strong variations that an hour doesn't make a significant difference.

But DST is a valid technique to realign sunlight and human behavior.
You can also use synchronized general schedule changes (using different summer/winter schedules), so remove DST isn't a big of a deal. But the most common complain (the day schedules/clocks changes) will remain just in another format.

In my experience, the two most big factors of "Unknown" are.

1- Non browser traffic. Bots, search engines, malware attacks, etc etc.

There are LOTS of non-browser traffic. The less visited is a web, the more non-browser traffic it has.

2- Android/Android derived OS

Android derived has too many versions, and lots of browsers and apps with web traffic. A lot of these traffic uses weird/unrecognisable User-Agents so in classifications like this they end as "Unknown".

3- Apps in general. Obscure browsers. Privacy add-ons. Rare configurations

Far behind the previous classifications, you will find traffic from these other cases.

For this, I would just split half to Android, and the other half I would just remove it and weight it between all the remaining shares (because most is just the first option).

The bottleneck usually aren't the batteries, but the inverter.
The batteries usually have significant margins to avoid stress and gain longevity.

Fusion energy has being clearly extremely complicated.

Steady self-heating plasma for long duration hasn't being achieved yet. And we aren't fully sure if self-heating plasma will have the exact same behavior than external heated plasma. After all, how the energy reach the plasma is important.
Also the tritium feeding problem is not solved neither mentioned.
As the condition hasn't being achieved, there is no real data about the fatigue of materials under neutron leaks that won't be stopped at the shielding and fuel feeding.

In summary, nobody can ensure it's gonna work based on "papers". On paper you can propose an experiment with multiple goals and you declare success if most goals are reached. You don't propose a full reactor over exotic yet to prove concept full of problems yet to be solved.

This is just an operation to capture funding.

Without a real ABI-32 Wine will have to use an emulation layer if the processor can't run natively as a 32 bit process.

Now that Linux is getting a growing game player quota, there are gonna destroy a functionality that make that works?

I think they are naming "drones" just multipurpose rovers. Not flying drones, but rovers dedicated to menial tasks.

I know that's a hell of difference.

An order of magnitude at least.

But it's matter of time they do a similar concept for mammals, building artificial wombs instead of artificial eggshells.

And then, use it for humans will be only a moral issue, not a technical one.

I'm pretty sure most datacenters has plans to do it anyway.

It's natural to start with natural gas, as it's fast to build, and you need the power for backup anyway. But renewable is cheaper, so you can expect they build or import renewable electricity as soon as possible.

Obviously, because they are only making the excavation of the site.

That's the easiest part of the construction.

This battery is not there to maintain renewable 24x7, but to raise the capacity factor of the mix while reducing curltainment. The rest will be natural gas.

Big tech can easily get away from everything. Their legal departments are enormous, they have almost unlimited capital and they can bribe any regulator.

If you push regulations for content providers to be responsible of the user content you will only destroy the competition of the small creators that push their independent sites.

You're helping the Big Tech.

I'm not talking about protests (although that helps and give us time).

I'm talking about offering the true solutions of this problem (that it's used as an excuse).

We need to work with SO, desktop and browser developers to add a way to DECLARE the age of the personal account. NO IDENTIFICATION. Just a declaration.
Users are setup by an admin. Well, as simple as the admin can declare the age of the users. They can't override the configuration.

Browsers can call the SO to get the age declaration of the user. And with that, apply content restrictions based on age.

WE SHOULD DO THIS. Because that solves the problem, BUT WE SKIP THE IDENTIFICATION (that it's what they really want).

If we destroy the excuse, it's more easy to fight on the media. Because they are gonna lie. They are gonna push the narrative that avoid identification means "unprotect the children".
They don't care about the children. They care about themselves, and they want massive identification for the next step of establish censorship.

The real solution, if they were really pushing this for the arguments they say they do, would put very easy tools for parents to setup the devices.

NEVER GET A DEVICE TO A MINOR WITHOUT PREVIOUSLY SETUP IT CORRECTLY.

As simple as that. Every device should be setup by an adult.

And then, put the tools to use them easily. Most parents are not hackers neither capable with devices. Make the software as simple as possible for that configuration.

BUT WITHOUT ANY NEED OF IDENTIFICATION.

Otherwise you are trading a solution for a more bigger problem.

Instead... as most people see, it's not a problem for the people is pushing it. They want massive identification, and all of this is just an excuse.

Because they are pushing for a different motive (to know how does what) they won't care if RTOS doesn't do identification.

They will said that it depends on the conditions. And the conditions is if the user has the freedom of interact and speech freely in Internet with it.

That's the reason they attack the main desktop and mobile OS, not routers or whatever.

At current stage this is gonna be an absurd change. They will put an app that it will be easily skipped.
That's the reason that, if people doesn't drop this, they will enforce only "approved" software through TPM control.