People go to music charts to see what others are listening to, and apparently they like listening to this AI generated song, but since most if not all of these charts are run by music trade bodies and other associations that want to promote their members, they set rules in such a way as to promote their members.

That's OK - just don't let this be marketed as 'Sweden's official chart'

The AI revolution is certainly driving additional capacity requirements, but even without it, all the cloud providers have been taking as much capacity as they could get.

As to benefits to society - modern data centres and clous solutions are a lot more energy efficient and sustainable then equipment sitting in server rooms cooled by building air conditioning systems.

Cheaper to operate at scale, and certainly a lot more reliable.

The two measures that define most data centres are commissioned power capacity and cooling capacity, which are obviously related. Data Centres are basically real estate with power and cooling assurance. Contracts with big customers, like hyperscalers - Microsoft, AWS, Meta, Google etc are then struck based on maximum power consumption, when contracts with smaller organisations that buy racks/pods/data halls worth of space are often converted from the power metrics. so i.e. you'll pay differently for a 15kW rack and differently for a 20kW rack. Obviously same amount of physical space. The hyperscale racks can go to 50kW, and can now be designed for 100kW for AI workloads, assuming the Data Centre can accommodate that amount of power density, generally requiring liquid cooling. This then requires different structural characteristics from the Data Centre, in order to accommodate the required piping and the right cooling systems. Yes, the DC provider needs to make sure there's enough power capacity available from the local DNSP or TNSP (if you hook up directly to the transmission network with a local substation build into or next to the DC), but generally power availability is only an issue with 3rd world countries and some really poorly performing networks. I find the top line, about 500MW added in one year in SV to be surprisingly low. We're talking now about building 1GW campuses by 2028 in places significantly less data hungry that I imagine SV is.

For many organisations not paying could equate going out of business, so the pay, to stay in business. This is now the main reason to have cyber insurance, which is then often required by customers. The âwe donâ(TM)t negotiate with terroristsâ(TM) stuff doesnâ(TM)t really work. Almost everyone negotiates, and those who donâ(TM)t, are able to accept the losses.

I was also initially surprised, but turns out itâ(TM)s a common practice in the industry. Shows you have âtaken reasonable stepsâ(TM) in court.

20% of 365 is 73, so 65 is in fact less than 20% of the streak. Since you're all about basic math.

Itâ(TM)s not that Microsoft canâ(TM)t do that. In some instances selecting Microsoftâ(TM)s public cloud offering may mean you canâ(TM)t meet YOUR requirements. That is not Microsoftâ(TM)s fault. Azure public cloud wasnâ(TM)t intended for that purpose. If youâ(TM)re trying to make it fit - thatâ(TM)s on you.

That is, in fact, incorrect. Microsoft is absolutely capable of building cloud services that meet data sovereignty requirements. Otherwise many governments wouldnâ(TM)t be able to store or process their data in ie Azure or MS365 environments.

Umm⦠what?? Youâ(TM)re mixing up data sovereignty with the fact that itâ(TM)s impossible to assure complete security of online environments. Data sovereignty doesnâ(TM)t aim to assure security, these are different concepts.

Youâ(TM)re attacking a strawman argument that you invented yourself . Parent post talks about using item differently than what it was intended for, youâ(TM)re talking about using an item not fit for its stated purpose.

No. If you use generic cloud services to implement a system that includes data that shouldnâ(TM)t leave your country, itâ(TM)s on you, and not the generic cloud services provider to assure that the data doesnâ(TM)t leave the country. Data owner owns the risk.

I think heâ(TM)s hoping to do similar to what Linus does with Linux kernel.

Craig Wright clearly doesnâ(TM)t have access to these bitcoins, he used to claim these were in escrow or otherwise will only become available to him at a later date. He canâ(TM)t cryptographically prove he is Satoshi, that much is clear, so heâ(TM)s trying to find other ways to get some legal recognition he could then use to pressure the bitcoin community.

Quite clearly, he doesnâ(TM)t have the keys to actually prove he is Satoshi. He has in the past provided obvious forgeries of signatures, like using old signatures done using satoshis keys that were publicly available. I donâ(TM)t understand how anyone can give him any credibility at this stage? It seems heâ(TM)s trying to use the law to convince a court he is satoshi, and then either force the bitcoin community to cut him in, or otherwise use the legal system to then assert some moral rights.