Look into placebo a little more. You will find that a large portion of the effect isn't even "expected to feel better, felt better" but rather effects that are completely useless to the patient. Two big ones are reporting expectation (you expect me to feel better, so I'll say that I do) and the fact that most studies have other components (often general monitoring and support) which help both branches of the trial. Neither of those benefits anybody outside of the trial.

Battery swap is an interesting idea but I think there's been enough progress in high-speed charging that we can consign it to the dustbin, outside of some niche applications. The biggest problem is that it forces you to design all of your vehicles in a way that allows the same type of battery to be swapped in and out easily. That limits both your design choices for batteries, which are evolving quickly, and how you attach them to the vehicle, which could significantly impact the overall cost and performance of your vehicle. Secondarily, it effectively forces you to lease one of the more expensive components of your vehicle. You're talking about swapping a $10,000 battery which contains $10 of electricity. As if all of that wasn't enough, you would need to automate moving heavy and valuable objects quickly and at high volumes in a way that works for typical consumers, not a tightly controlled Warehouse environment. While those problems are soluble, we're still talking about a theoretical solution with some big obstacles, which is only a little better than the best real fast chargers today. That said, I would love to see some sort of swappable range extender standardized. Not the sort of thing you do every time you fuel up, but the sort of thing you rent for a week when you go on a road trip. Imagine a small trailer with a 10 kilowatt generator and a little bit of extra cargo space for luggage. Such a trailer could probably be built for $3000 and rented profitably for a couple hundred bucks a week.

Excellent point. Free Will isn't some magical attribute of the universe, it's an emergent property of the complex systems known as people. Yes of course that's governed by the laws of the universe, the initial state of your brain and the inputs and outputs it gets. That doesn't stop it from being a useful and meaningful concept. Particularly when you consider that some of those inputs are information about what will happen if you break certain rules and knowledge of what happened to others who broke those rules. I do think that this is a argument against retributive Justice, but that is only one reason for punishing people who break the rules. Discouraging rule breaking in the first place, and in some cases preventing immediate repeats are also sound reasons.

You've identified the right problem but underestimate the extent to which it's being worked on. In your post you referenced GPUs and not CPUs because we understand it's a parallel problem. Lots of smart people are working on it and gpus are incredibly parallel compared to what was possible in such a small space 20 years ago. On the other hand we're still working with silicon because it's the best we've got. It also has some massive advantages: even if it was power efficient, no one would be very interested in a neural network technology they took a decade or two to train. Even with silicon, the most extreme parallel architectures tend to be inflexible in the software that they can run efficiently. While we can come up with better technologies, I believe this is an inherent trade-off: when you build a massively parallel system you make hard wired assumptions about where the data is going to flow and the type of operations that happen in between. If there is a mismatch, large portions of that hardware sit idle. In contrast, a serial system might have lower peak efficiency, but you have much more flexibility to reorder those operations without taking a huge performance hit.

The catch is there's relatively little difference between 30 year payback and 50 year payback unless you assign the time value of money to nearly zero. Of course, I believe Japanese interest rates have been effectively zero for a long time now. That creates its own problems since unproductive companies will continue to be viable on paper.

People aren't upset because they're selling a product consumers want to buy. People are upset because selling a product that has massive externalities both in production and consumption. And the companies are not paying those externalities. Externalities aren't magic, they are real costs that happened to be borne by people not directly party to the transaction. When a large organization imposes costs on you without your agreement or consent, it's completely reasonable to use the power of government to either stop them or at least demand compensation in the form of taxes or other levees.

While various uses are possible, the most obvious advantages of local processing power are to provide features while preserving a higher level of privacy. If they want to upload stuff most phones have enough bandwidth to just upload it. On the other hand, if you have significant local processing power you may be able to run models like image enhancement or pretty good voice recognition right on the phone. With enough hardware it can even become possible to train customized models on the phone itself. Many of these use cases require a lot of processing power to take them from cool toy to something you would use everyday. Voice recognition is probably at the top of the list, they had demos running on relatively inexpensive Hardware in the '80s. With orders of magnitude more Hardware now it's still barely good enough to use to dictate this reply. Of course, I'm using Google's Cloud to do it. The on phone recognition does work but still sucks.

Energy required does not scale with heat moved, but rather with heat*(deta temperature). Since air has a much lower heat capacity than oil, they have to move a LOT more of it, or cool it more, or both. Those take extra power. In particular, consider a server farm in a relatively cool area, like Virginia. The chips themselves might be operating at 40-60c, a higher temperature than the air outside. With perfect heat transfer you would need no energy to cool them, in theory you could even reclaim some power. In practice transferring that heat outside is difficult and it may actually be more efficient to run an air conditioner than to try to push that much ambient/passively heat exchanged air through your system.

This sounds brutal, but I suspect you have much higher than typical power demands. My typical consumption is only 30-35kwh/day for a 3 bedroom, spacious (not palatial) home built in the 90s - ok but not exceptional insulation. I have friends who are off grid with a tiny power system, but step one was very efficient insulation and appliances. Admittedly, going to just a couple kWh /day does require sacrifices, but there's a lot of room between 1-2kwh and over 60. You also mention an electric car. Are you charging while the sun is up, or charging at home from those batteries in the evening? Hoping not to sound too contrary, it sounds like you got shit contractors. Just trying to put it in context.

There are still phones with an audio jack, or you know, just get an adapter. They appear to be available on Amazon for $10 to $20 CAD . No one is being forced to throw anything away. Offering a competing project certainly doesn't force you to upgrade.

Have you done a deep dive on nickel iron? Yes those suckers can be maintained for decades, but their round trip efficiency is terrible which means you need more solar panels. They also have high upfront cost and they need regular maintenance. They have some very cool niche applications, but even factoring in replacement cost, lead acid or lithium ion is going to give you a better dollar for dollar system.

I see plenty of comments about how fast the human eye is. These generally miss the mark by considering only the retina or, roughly eqivalently, a fixed gaze. Human vision is a system that includes rapid eye movement/tracking and (neural) post-processing. Since the eye can track a moving target across the screen, you're going to get motion blur at any frame rate. The point of diminishing returns is limited by spatial resolution and the speed of the movement. An easy to understand rule of thumb: if something is moving at more than one pixel per frame, there would be at least a marginal benefit to running at a faster frame rate.
Not to say it's going to be a large benefit, your eye can't possibly be tracking every single thing in the scene, however there is a huge difference between good enough and truly eye limiting.

True enough. Those are the contexts where you are used to seeing smooth motion on-screen. However, I can almost guarantee that if you saw the next dozen releases at high frame rate, you'd never want to go back.

As screens improve resolution (4k is approaching, but not at, eye-limiting for most people), colour, and frame rate (120Hz seems to be the point of diminishing returns) it's less like watching a picture book and more like looking through a window, or even a live scene. For most movies, the more you can give the feeling of "being there" with the action just in front of the audience, the better.

I did as well, but I will say that seems part of the non-replaceable battery move has been to make them a little bigger. None of my last 3 phones were replaced due to battery life: Two failed due to the GPS becoming unusable (the Sony apparently due to a shitty driver after a major os update and the old Samsung most likely due to a weak internal antenna connection). The third mostly because I wanted a better camera.

24 to 48hz is generally not a software upgrade. Perhaps on a few models with software locks. Cinema projectors are very cost sensitive. In general if the hardware can handle 48hz and you sold 24, you could have used cheaper hardware. The expensive part isn't the part that is like a tv. Look into the data rates that feed a high end DLP chip, it's astonishing.