AWS revenues have been growing at + ~$6B per year for a few years and that has not changed. Just because MS is currently growing faster because they have a bunch of locked-in customers whose easy-path is Azure, doesn't mean AWS is "weak". Since Azure growth is coming from a largely fixed pool of customers, it will taper off at some point. There are plenty of things to not like about AWS' corporate culture, but the business seems to be firing on all cylinders when it comes to revenue growth.

It is going to take a rouge country or billionaire to unilaterally initiate a geo-engineering program like stratospheric aerosol injection (https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fen.wikipedia.org%2Fwiki%2FStratospheric_aerosol_injection) to save us all. Pakistan is one of the counties suffering the most from global warming, so maybe they will crack and undertake this mission to save the rest of humanity from their own stupidity. The clock is ticking. Our warming atmosphere is already demonstrating positive feedback on natural methane production and within 70 years natural methane will overtake human contributions to global warming if we do not act quickly to prevent this scenario. At that point, if humanity wants a cooler planet, geo-engineering will be the only choice.

Companies have farmed recruiting out to other companies that have no understanding of what is really needed and are just going off the job description given to them that often contains impossible requirements to meet. (e.g. 20 years experience in training LLMs). It is as if companies have forgotten that people who are good at something often know other people who might be good at the same thing. The entire process of hiring now revolves around companies like Indeed that create a tremendous amount of noise of false abundance in the entire process and are not providing value to either the job seekers or the employers.

Incompetence. They have redundancy, and unlike AT&T they participate in peering arrangements at many IXPs. They just do not have employees capable of adjusting routing and when they finally do come up with some new routes, they are bizarre and linger for months after the incident. Every time they have a cable breakage in Dallas, PARTS of Austin (200 miles away) and other areas of Texas also lose service. It is like they have no clue how their network functions until something breaks and then they start trying to figure it out. And before they are done figuring it out, the cable is repaired and since everything seems to be working they just stop managing the network until the next incident.

Companies cannot just "step up", the solution generally requires the national government that wrote the laws that effectively require a sovereign cloud to actually engage with providers to create sovereign cloud solutions. Most governments, including Britain's seem more willing to write laws than do the hard work of understanding the implicit impacts of those laws and in this case engaging CSPs to create working solutions under those laws.

The U.S., China, Germany, France have worked with Microsoft to establish sovereign clouds. France's Bleu Cloud isn't even operated by Micrsoft, but rather by Orange. The underlying problem here is that Police Scotland ignored British law and set up their "Digital Evidence Sharing Capability" (DESC) using Office 365 and SharePoint. This isn't on Microsoft this is on the British government which has ignored a huge demand signal and failed to engage Micrsoft and this is on the organizations willfully ignoring their country's laws. Given a reasonable contract, Microsoft will jump through hoops to establish a sovereign cloud for a given country.

Right now, it is a pretty short list to check to see if your country has a sovereign cloud. If it does not and you have data protection laws that effectively require a sovereign cloud for an approach under consideration, perhaps you should look at other options. Not complain after-the-fact that the chosen platform doesn't meet your needs.

They should focus on UPI-Euro if they want to make the Euro a centerpiece of global financial transactions. https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fen.wikipedia.org%2Fwiki%2F.... Of course I'm sure Visa, MasterCard and Amex are doing everything they can to convince these banks to pursue this foolhardy stablecoin approach instead of UPI which would be the end of the credit card industry.

I agree that stocks are currently overly socially valued, but disagree with the premise that they inherently do not have financial value. Both factors are always in play to a certain extent, but the current market seems to have many stocks that are as much as 90% socially valued -- which is certainly not healthy.

I just see this as "Idiocracy -- Wall Street Episode". The sheer extent to which sentiment drives many stock's values is a relatively new phenomenon in the market. Before 2000, the value-based assessments of stock analysts were taken seriously. Raiders like Carl Icahn used to buy up companies whose real value had been harmed enough by sentiment that buying up all of the shares and then parting out a company was a profitable venture. You will be hard pressed to find a stock that this is true for today.

It seems to have taken almost 100 years for the lessons of October 29, 1929 to be forgotten, but they clearly have been. We have seen these cycles previously in history where some collection of commodity prices came to be "Socially Valued" as you call it -- "Sentiment Valued" is what stock analysts call it. These cycles have never ended well.

I feel sorry for the plaintiff. She worked for half a year for fraudsters who did not pay her and then because her attorney is an idiot that filed 25 flawed claims, and zero legally valid claims, she gets nothing. The fraudsters who screwed her but hired better attorneys get away with it.

According to Pat McGee's book "Apple in China", Apple invested $275 billion dollars into upskilling Chinese labor between 2016 and 2021 while at the same time investing comparatively $0 (to three significant figures) in upskilling U.S. labor over the same time period. This shows that their plan and the plans of similar companies in China were about much more than just outsourcing production. This was not an "oops", it was a plan.

Entra ID is AD with OpenID--ish instead of Kerberos--ish. Because Microsoft's early cloud architects decided that they needed to jettison Kerberos to move to the cloud, which was really dumb. Also sure the infrastructure team standing up all of those Linux boxes did not want to have to deal with LDAP config being a pre-condition to everything. They would be so much better off if they had just doubled down on Kerberos. Now they have a giant mess of variations in how authentication works that have all been bailing-wired together and will forever be the source of security exploits.

The average for the entire Gen Z seems to be closer to 0.

I'm afraid the future is in companies like Loongson Technology which will be bringing forth the next generation of CPU designs. NVidia should be investing in the future, not the past. At a market cap of just $8B NVidia could buy Loongson Tech outright.

It seems like a majority of the movies ever made were never promoted or widely released for reasons that had little to do with the movies themselves. This indie movie https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fwww.imdb.com%2Ftitle%2Ftt0... was awesome but was sold for $1M and never released. I would assume that the person who wrote that $1M check thought the movie was worth having. This fun cheap movie https://ancillary-proxy.atarimworker.io?url=https%3A%2F%2Fwww.imdb.com%2Ftitle%2Ftt0... at least made it to VCR. IMDB lists over 10 million movies, the majority of these are available for next to nothing in relative movie making terms.

NVIDIA introduced CUDA (Compute Unified Device Architecture) in 2007 as a parallel computing platform and API that allowed developers to harness the power of GPUs for general-purpose computing. CUDA revolutionized computing by enabling massive parallelism, making GPUs ideal for scientific simulations, cryptography, machine learning, AI, and more. It provided accessible libraries and tools for developers in C, C++, Python, Fortran, and other languages. CUDA quickly became the backbone of GPU computing in academia and industry as a result of NVIDIA's open-source efforts.

18 years later, Intel has neither jumped on the CUDA bandwagon nor produced a compelling competing library. Intel in institutionally incapable of creating something like CUDA because it is too hard to tie future revenues to an effort like this.

NVidia bought into John Nickolls' vision. Intel completely missed this boat.