The problem is, Cheato has already proven himself to be an "oath breaker".. which means China has no reason to accept a "deal" from them.. oh, they will make nice..but ultimately they will focus on divesting themselves of the US market as much as they can.. They won't ignore it completely.. but it won't be a priority.. Deals work best when both sides are known for KEEPING TO THEIR WORD.. But... well..... you know..

Costs for Cloud infrastructure RISE over time.. as well as their needs I'm sure grew.. the problem is, adding another TB on-site is relatively cheap (assuming you purchased with capacity to grow) and don't usually go up annually.. vs. cloud infrastructure continues to rise in cost as companies like AWS/Azure dice the pie finer and finer on what is included in base costs vs. addons and escalating costs in initial growth. (23% annually).. vs. on-prem tends to be relatively fixed cost with annual increases being in the 3-5% range not close to 30.

So sure, it sounds great initially but overtime cloud can be its own trap.. (since customers also have to consider the cost to exit which sometimes exceed the cost to continue (ie: trapped).. in their case, the annual cost far exceeded the exit cost so its a no brainier to cut and run), but most customers are close or slightly exceeding the "break point" and thus risk adverse CTO's tend to stay..

At the size of MS there is never really innovation because innovation costs money and often doesn't lead to a marketable product or is in competition with another product they either already own, or are partnered with.. Small companies innovate all the time because they NEED the "new" or "different" to distinguish themselves from the MS of the world.

Large sized companies CONSOLIDATE.. (which by extension means the control/elimination of the SMB competition)
Medium companies expand the market by merging products into solutions..
Small sized companies innovate for new products/solutions because its their way into market.

You do know there is a huge difference between the "cheap" stuff (everywhere) and the "good stuff" (everywhere, even in China).

The stuff that's imported into the US from China is mostly the lowest cost (and thus the lowest quality) goods because companies in the US focus on price first and quality second. The domestic products in China have about the same high standards as everywhere else. In China, reputation matters and people talk QUICKLY.. so a bad product in many cases will DOOM a company (legally, financially, and socially).. vs. in the US, the attitude is very much "buyer beware and if your building blows up, shame on you for making it so flammable"

As a general rule, almost all "things" can be done in the US.. its just the cost (economic, human, environmental) has been high enough to push it off to someone else so THEY bear that cost while domestically the US reaps the profit..

Now all that "cost" is going to come home to roost.. so people are taking a look at at least softening the blow.. but no matter what, someone is going to pay it... China is large enough that it can absorb some of that human/environmental cost.. but even they long term have been looking at alternatives because its finite. And when a sizable number of people live well below the poverty line, the economic cost (as it impacts humans) is less critical.. (but China has national health care so technically that poverty line is closer to the US line if you factor in free education and national healthcare

I think people are ignoring what's going on.

Facebook (most especially) but apple, and others have been essentially saying:

Well, our HQ is not in the EU and while the servers are in the EU and we serve EU customers, because the HQ is in the US (and/or the leadership is in the US), they are NOT subject to the rules/stipulations that the EU has put forth for firms servicing EU citizens.

Which is almost the same as when the US tries to claim sovereignty for any data that's being accessed by US citizens (regardless of point of origin).

If anything, this is very much tit for tat.. (if the US can claim they have rights to data housed in the EU because US citizens/business accessed it).. then the EU certainly has the right to say "well, then you must comply with our rules regardless of where your HQ/Leadership is if its available and servicing EU customers"

Its essentially the same thing.. either both parties need hands off their respective data when not in their territories.. OR it all applies and businesses/customers must comply with both issues.

The biggest problem he (and many others have) is the assumption that their "brand" carries a lot more weight than it does.. the US "brand" carries as much weight as its willing to invest in the brand by its actions and other's ability to see them as a RELIABLE and TRUSTWORTHY partner that is focused on PROGRESS.. The second that goes away, the brand is dead... especially in IT we have seen this story play out many times with firms that never invested in innovation and instead focused on their "existing product portfolio" only to be displaced with other companies that instead showed others WHY they are better.. (3COM, Novell, etc...)

The US is pretty much the Novell of the geopolitical world.... clinging to a legacy in a world that has changed is trying to exert pressure in a market to drag others down and back towards itself. Ignoring that nothing they are actually doing will benefit others and no one really NEEDS them because there are alternatives.

Humans are notoriously vulnerable to selection/information bias.. so why would we assume that AI models (which are in many ways, modeled after how kids learn) would be except from the notion that all data that's available is not honest/truth.

Arsenic is pretty common in most rocks and soils. But cooler temps keeps it "locked" away so its a very slow leeching poison.. but as you heat it up, that process speeds up and dumps a LOT more into the surrounding water where rice paddies are growing.

Think of it this way.

Get some rock salt (like big crystal salts).

get cold water and drop some in there.. It will dissolve slowly
then
get warm water and drop some in there, it will dissolve quicker and more diffuse in the water (if you taste the water, it will taste saltier quicker).

And that's basically what;s going on..

And its the same for just about all toxic chemicals (toxic to humans) in rocks and soil.. as the ambient temp rises, it speeds up the erosion of those rocks and chemicals contained in soils..

It costs 100M a year..

And the problem with thinking like a business is not everything is a "profit" line item. Government's job is never about "profit".. yes, there is a balance between expendiatures and intake. but its not exact..

Like for instance.. how most people think is:
Direct Income Taxes = Income
Everthing else = Expenses

But the reality is more complicated than that.. Like the CVE program... that 100M that the US government spends, helps to reduce cost by many companies (US and otherwise) as well as agencies by reducing threat exposure and vulnerability costs (from threat actors and breaches) which in turn make those agencies spend less money (or allow companies to generate more profit) which in turn for companies earn more in taxes (income taxes).. Same for non-US companies that deal with the US.. most have provisions and contracts in place to mitigate risk by doing audits against such CVE's which means they earn more money, and have a greater interest in doing business with the US, because of the mutual benefit), which generates "soft power" because is a tangible benefit/perk for such an arrangement..

When everything becomes a transaction, there is simply far less of an incentive to do business since everything becomes a process and cost.. many have to weigh the penalty vs. benefit.. and it starts tipping more against the US than for it. (its one reason why a lot of companies simply don't focus on specific geographic markets.. there is very little benefit so they may deal with a larger region as a whole but simply ignore the sub locations because of that "cost".

Basically 100M in the federal budget is practically free, but the benefits to US companies and others is immeasurable... and just like "insurance" everyone bitches about it UNTIL they need it and then they kiss their agent, up until that time, they curse the name as an expense.

The EU has the tech.. its just a matter of who's at the forefront..

On average in terms of technology pacing:
China and the US jockey for the #1 spot (so they go back and forth between 1 and 2).
Number 3 tends to be Europe (on average they are about 2-3 years behind the curve).
Number 4 tends to be Australia/NZ (on average they are about 5-7 years behind the curve).
Number 5 tends to Russia and the various "stans" (on average they are about 4-6 behind the curve, but because of political and military aspirations their ability to capitalize on it is low, thus putting them in the #5 spot).)

So it just depends on how the EU wants to play it.. they could tell both China and the US to bugger off.. but the problem is then they are in competition with two groups that are going to strong arm others to comply and thus creating a global fragmentation (which from a domestic security standpoint is good, but from a user/market/international security standpoint is bad) and thus both China and the US are going to use it as justification for more "issues".
OR
They could align themselves with China (and we KNOW they are going to spy/intrude) but would be polite about it..
OR
Or they could align themselves with the US (and we KNOW they are going to spy/intrude) but are going to be dicks about it..

There isn't really a good answer here.. the best answer would be to stick with domestic but there are long term consequences that they have to make sure they understand.

I mean basically that's just giving money to Cheato enablers... so why would NATO (as a collection of sovereign nations that are pretty much under "attack" by cheato) essentially invest in a company run by someone that's supporting him?

That makes little sense.

Yes, but those "older and wiser" students are usually going to be more "conservative" in their mindset. (Its been widely studied and while some people maintain their beliefs overtime, most do become more socially (which often translates to politically) conservative overtime as they age (https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fwww.theguardian.com%2Fcommentisfree%2F2015%2Fnov%2F03%2Fdo-we-become-more-conservative-with-age-young-old-politics) as they are more focused on "job" and not on finding out their skill sets/interests and other social learning that most younger college kids do. Basically its a backdoor way to turn colleges conservative by simply changing the make up of the students that can/do attend.

Its a little tricky..

DP can handle it.. but DP is open spec.. which means different devices tend to implement the pieces they like and ignore the pieces they don't. HDMI is NOT open standard (requires licensing), but it does assure a level of cross compatibility and functionality (depending on the level you are at).. But because its open spec and unlicensed, ARC and CEC (which are licensed features) are "expensive" to implement.

This GPMI is something of a hybrid designed to supply similar features like ARC/CEC in the Open Spec model of DP while retaining some "backwards compatibility" with HDMI. Its basically a frankenstien spec that should in theory be better than both DP and HDMI, but in practice will most likely be less than the sum of its parts. And as GPMI is licensed, but "Free" for now, most vendors (outside of China) are not excited to jump into a spec that could become expensive down the road. (free is often the most expensive thing to deal with)

That and also the actual demand.. right now, companies are still dipping their toes in the AI waters so its not like EVERY company and person is requesting AI capable VM/Infrastructure.. Which means its the classic demand vs. supply (read: cost). in a healthy economy, they could float some on the "if you build it they will come" plan.. but in a global recession that's right around the corner (thanks Cheato).. everyone is pulling back, curtailing major spending, and playing a wait and see approach before spending.. and of course efficiency training as well (How to squeeze performance while reducing hardware requirements)