Sounds like you're one of the two Bobs and want to promote him?

Agreed, this is the natural solution. A simple process that can be handled by a local service station, something you can do a few days in advance of longer/unusual trips, then switch back afterward.

Manufacturers choose their position based on perceived market desires (partially) and cost. Ultimately they need to make something that sells.

Because we are in the shake-out period of a new market, the chargers are sparse and non-standard. Fast forward 15 years and most likely there will be a better charger network with better standardization which allows the manufacturers to design their product around that landscape. Most likely the range thing will be a standard option per model so the consumer can choose which range works best for them.

Two points:
1-A paper from Oct 22 that was a compilation of research about carbon fiber for deep sea basically concluded that the math to properly model it is not currently known, existing math for this type of stuff don't apply to carbon fiber, experimental results not always the same as the expected result.

2-Viewport: Only rated to 1,500 meters, Rush thought there was too much safety built in to the charts for that material and it would be fine, others disagreed with him.

If the sub is heavier than water then you need a method to bring it back to the surface (propulsion and energy). To get around this issue, titanium subs get coated in syntactic foam which is expensive but allows the sub to float to the surface after dropping balast.

That was the only employee, but there were many other deep sea specialists that told him carbon fiber not so good (36 signed a letter to him, plus he consulted directly with a few different people that told him carbon fiber either not so good or there simply isn't enough info about how to make it safe/how to understand when it's safe).

If you read their paper it will help you understand how their system works and where it has applications. Here are a couple key points:
1-" First, the data transfer between digital and analog domains is fully eliminated until convergence to a solution is achieved."

2-" Second, there is no separation between compute and memory since the variables are computed as light traverses through the optical modulation matrix"

Their system is specifically designed to solve optimization problems like traveling salesperson in a manner that avoids key inefficiencies with our current typical computing model.

Before I respond, just let me state that I'm not claiming any kind of certainty about QM in the brain or even that NP-complete problems need to be solved to produce consciousness. I have no idea and neither do the researchers that I read.

Having said that, there is active research in this area, and, just following the slow accumulation of data in the same way that has happened with the now current understanding of astrocytes performing a functional role in computation (non-linear intra and inter cell calcium wave signaling etc.), it sure looks like there is an interesting bit of info being added that could reasonably end up at the point that QM is a key functional tool in our mental processes.

Maybe.

It could be the secret sauce from an efficiency perspective. Meaning it's the tool used to quickly and efficiently solve key NP-complete problems required to produce consciousness. Maybe.

that generative AI...

Exactly, there is real value for this type of product for this use case. In addition, when all cars have this type of system, it should make the commute more efficient by reducing the process of large groups of cars speeding up to 60 and then hitting their brakes hard and coming to a complete step only to do it again over and over. The efficiency of a smooth flow should cut the time down by some percentage.

These portable computing devices are being called a "personal" cloud, or PC.

But the GPU is broken down into "cores", they look different from CPU cores, but it's a reasonable unit of measure to compare the M3 to the M2 (for example).

An additional critical constraint is keeping those cores fed. Memory is slow and having many cores increases the need for larger cache.

I was wondering the same thing. Generally microservices style relates to separating data and functions per domain, as opposed to designing around centralized/shared data and/or functions. For example, if multiple domains require customer master data or sku data to perform their functions, they are each required to create+maintain their own local copy, typically by consuming add/change messages from other domain(s) as well as maintaing additional attributes locally that relevant to the local domain.

The benefit for enterprise systems is the ability to adapt local systems more rapidly due to changes having a reduced impact on other systems (compared to centralized/shared). The downside is increased complexity and effort to build out the local systems compared to centralized/shared approaches.

I'm not sure how that applies to a framework.