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://www.imdb.com/title/tt0... 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://www.imdb.com/title/tt0... 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.

Hundreds of companies exist that are doing exactly this. I worked for a company trying to get into this market almost 2 decades ago and I doubt it has changed much. This is one the most fragmented market niches there is. The sales cycle is extremely long -- measured in years because police departments do their budgets at most once a year and then whatever governmental entity they are associated with gets to tell them "no". Which is why the software too many police departments use is home grown written by somebody's relative, only used initially at one department and then spreads organically regionally based on personal connections. Which is probably what is going on here. Somebody's relative cobbled something together in A365 and the police department started using it because it solved a problem they had. The compliance status of the platform was probably not even considered. It is a horrible market to try sell into with a "we built it come buy it attitude".

The way our governments hire contractors pretty much requires every employer that wants to compete for a given contract to round up a bunch of prospective employees -- but not actually hire them unless they win the award. There is lots of hurry up and wait when trying to work for a government. One contract I was part of the pursuit on did not announce a winner until 2 years after the RFP submissions. At which time the winner went back to all those people they had resumes for and tried to bring them on board -- good luck with that. For every contractor currently working for the government there were 10 or more -- possibly over 100 almost identical job postings before RFP-time from all of the companies that competed for the work. If you come across almost identical job postings from Lockheed Martin and Northrop Grumman what you are seeing is pre-RFP job postings.

was find a way to give a billion or so of the CHIPS money to Trump ... for instance by "investing" in Trump coin. They did not give Carrot Man his carrot, so now they get the stick.

I am sure there are many factors that lead to these failures, but two that come to mind immediately are:
- The apples and oranges issue between APIs like LangChain and traditional business process workflows found in tools like CRMs, and exiting IT data pipelines.
- SQL -- If the LLM has to generate correct SQL every single time for the integration to work, you may be doomed to failure. This isn't just syntax, it is correctly choosing all of the constraint values. One incorrectly chosen constraint value (e.g. "airplane" substituted for "aircraft") and the integration just does not work reliably.

Every year they publish a report attributing bad corporate processes to "insider threat" helping to drive an entire industry through this nonsense. If a company has a worst practice corporate procedure such as "HR personnel opening emailed resumes on their desktop computer" there is no amount of phishing training that is going to protect a company from being hacked by an exploit ridden resume. Almost every even hacking event Ponemon attributes to "insider threat" is really the result of worst practices. CIOs are drinking each other's Kool-Aid which results in the "need" for this training.

The Ponemon survey is of CIOs -- and CIOs are never going to acknowledge that they were hacked due to worst practices. Instead, they throw some lower level employee under the bus. The Ponemon entire methodology is flawed by how their survey is crafted and who is answering the questions. Then this flawed methodology is used to sell training that we all have to suffer through that we all know is ineffective.

In Texas, a relatively small capacity equivalent to around 100 PowerWalls is sufficient after a lot of paperwork to become a wholesale power generation company. Then you can charge 30X normal rates for the power you contribute to the grid during Texas' routinely declared power emergencies. Get the whole neighborhood in on it under an LLC or non-profit charter and profit rather than being exploited. Or set it up as a MUD which can help make your neighborhood harder to annex if you are in a neighborhood under threat of annexation. And for neighborhoods that already have a MUD and a place to put a bunch of batteries, it makes much more sense to do this at neighborhood scale since the ROI is so much higher.

From figure 3 we can see that the signal is fundamentally parabolas. Reality is not composed of 4 well spaced objects and solving for multiple parabolas is hard and often simply cannot be done if you have no idea of how many you are looking for in the first place (like when trying to detect underground objects). Some cryptographic algorithms are even based on the difficulties of solving parabolic problems (but not in a "parabola detection" mode) This relatively recent paper https://pmc.ncbi.nlm.nih.gov/a... is addressing simpler scenarios but shows that methods of identifying parabolas in reality are still being sorted.

This is a cool use of Rydberg atoms, but some of the use cases they call out are not compatible with the nature of the fundamentals of their signal. They need to stick with cases where they know the number of things they are looking for. They didn't list radio-telescope, but this seems like a better use case for this (if built with a shielded, narrow field of view like a traditional telescope) than ground penetrating radar.

Good deal for me: Everyone really pays +3% or more, but my family gets more than 3% of everything we spend back in air travel and card benefits. However, this entire business model is falling apart in near real time. We have things like UPI https://en.wikipedia.org/wiki/... that started in India and is spreading around the world because it is so inexpensive. We have all of the apps like Venmo and if you believe crypto bros, crypto is going to cut into Point Of Sale (POS) transactions. And then there is the fact that the credit cards companies literally created the credit card fraud industry as a side effect of their effort to maintain control over the POS transaction business. They could have made credit cards FIPS-140 compliant security tokens that would have been readable by any PC making "Card Present" assertions possible for all web transactions. Instead they made sure PCs couldn't read credit cards. So though I benefit from the shenanigans of these companies financially I can't wait to see their long deserved demise. And when they go they are going to take industries like the airlines that have become dependent on them with them.