How so?

I think there’s two different approaches to this. This chair is obviously designed as a luxury experience, as the process takes a full 15 minutes.
My idea is designed for efficiency, to reduce the amount of time it takes to bathe in the morning without reducing cleanliness.

Interesting idea. Seriously over-engineered though.

If you want a ‘human washer’ you don’t need a $350k fancy chair with heart rate monitors. Just take a page out of the automatic car wash.

Human stands in a stall. Shower allows human washing of hair and face. Then just hold arms out making a diamond in front of you (think TSA body scanner position, but with arms forward instead of upward) and a 360° robotic sprayer starts at the neck and goes down spraying soapy water, then back up again with a slight up angle to get the groin and armpits. Shower comes back on to de-shampoo hair, then the same 360 robot does two passes with clean water to rinse everything off.

If you get fancy with machine vision and body position sensors, the 360 wand could flip 90° to do the hair and would be angled backward a bit so it doesn’t get water or soap in your face.

You could build this for a lot less than $350k. And instead of $1500 worth of body sensors you have a $50 waterproof emergency stop button.

Never said they were my friend. They might have been once, in the ‘Don’t be evil’ era, but that era is long past.

They are however somewhat more interested in open standards than Apple. Android for example uses OGG a bunch under the hood.

Nobody knows what to do with it because it’s proprietary and requires a license. If it was not encumbered, windows would ship with a decoder built-in for free and nobody would have a problem. If Apple devices didn’t use it by default, no one would have a problem because they just wouldn’t use it for anything ever.

If Apple got sick of paying the fee, they could switch to AVIF or JPEG XL or anything else. It wouldn’t be hard, just bake native support into the next OS of everything, and have the next iPhone take pictures in that format by default. The rest of the world will catch up right quick.

Actually come to think of it I’m kind of surprised Google doesn’t do that. Make the native Android camera shoot in AVIF by default…

Yeah but look at the AV1 hardware support matrix. A lot of current mobile silicon supports decode, not nearly as much supports encode. To have AV1 truly replace MP4/MP5 a hardware encode is necessary so you can do video calls in AV1.

The one who could really make this happen is Apple. If they decided to move away from MPEG-LA and embraced open codecs (AV1 / VP9 / Opus / FLAC / AVIF / JPEGXL / JPEG2000), supporting them in software, hardware, and their services (imessage/ichat/facetime, music store, video store) that would single handedly push the industry.

They did that with HEIC- before iPhones switched to HEIC by default nobody bothered with the encumbered format. Now it’s become de facto standard. That SHOULD have been something open like AVIF, JPEG XL, etc.

I’m kinda in the same boat. My main box was used, only a couple hundred bucks and it’s served me pretty well. Same thing with all my other personal PCs for the last 5-7 years.
If you’re not gaming, the benefit of going new is pretty limited. The hardware outpaced the software.

Perhaps, but if that involves desoldering BGA chips and remounting them on new cards, do you really think someone’s gonna bother to do that?
Especially if the chips are optimized for AI and don’t perform well for games?

Sadly I think you may be disappointed. The GPUs used for AI are not typical graphics GPUs, they aren’t on PCIe cards with video ports. They are set up in configurations designed to cram as many chips as possible into as small a space as possible while still providing power and cooling for 100% output on all of them.

The stock is down because the Overton window has shifted. Two or three months ago, AI was the future and question that was lunacy. Now it is a mainstream point of discussion that AI is a giant bubble, that it is entirely likely to pop at some point, and that the efforts pushing it so hard are bordering on irrational given the actual capability of the product.

If AI bubble pops, Nvidia loses. No more big tech companies with blank checks and open orders for AI chips that basically amount to ‘please send us as many AI chips as you can manufacture whatever it costs we will pay’.

And, if anything there may be a surplus of AI chips and hardware in the market as companies that have built entire data centers for AI suddenly realize that paying millions for electricity so online idiots can generate videos of cats racing Roombas down F1 tracks is not a trillion-dollar business model.

Yes this is absolutely ridiculous.

This is also a good reason to avoid proprietary codecs. H.265 may be a great codec, but the licensing fees are basically a tax on the world.

The best solution would be an overall switch to AV1. But silicon support for that is not nearly as widespread.

A nice thought.
Unfortunately the only ones who switch will be people 1. Know that it is even possible to switch/that they have an option to switch, and 2. Have the technical knowledge to back up their data, then download and install Linux, that 3. Don’t have to run Windows for other reasons and/or haven’t already switched.

Unfortunately with all three qualifiers you aren’t left with a giant number. Certainly no mass exodus.

Problem is that Samsung is like Apple- a shitton of people just blindly buy the latest Samsung whatever with zero research.
So you have a bunch of other companies trying to stand out in one way or another- Motorola for example just released a phone that brings back the 3.5mm headphone jack. And you have a ton of cheap Chinese companies that may or may not offer any software support after purchase but have interesting form factors.

That makes it hard for the little guys to get the kind of sales volume needed to justify the development and tooling for a really cool flagship phone.

Personally while phones today are far more capable, I think phone designs peaked in the mid 2000s. Mainly because you had actual innovation in design– wildly different form factors. There were a few phones that flipped open like a laptop with a physical keyboard, a handful that slid open to reveal a blackberry-style keyboard, many had SDIO ports or other ability to clip on expansion modules, etc. Phones had fun features- there was one that could do an early ‘google pay’ type thing by pulsing a magnetic field to pretend to be a magnetic credit card stripe for a swipe reader. A lot of the early Samsung phones had IR blasters so you could turn TVs on and off. There were a couple designed for gaming that were laid out like a game pad. Manufacturers weren’t afraid to experiment and the result was some really cool stuff.

Sadly that’s all gone today. HTC (which made many of those cool phones) was driven out of the market by Apple and Samsung, so now virtually all phones are identical flat bricks.
I see a glimmer of hope with flip phones and foldables, but not much. They’re all just excuses to

FSD has routinely plowed into children, emergency vehicles etc.

You are using this word ‘routinely’, but I do not think that it means what you think it means.

Can you give me, say, 10 incidents of this? Of a Tesla confirmed to be on FSD driving full speed into a child, emergency vehicle, etc?

FSD used to ‘routinely’ be overly cautious and slow down when not necessary, but I don’t think it’s driving into things.

I’d also point out the driver remains responsible for the car and an eye movement camera prevents distracted driving, but I digress.

Other companies have implemented these more limited systems (that often include better sensors such as lidar) not because they can’t do it but rather because they are more cautious about brazenly lying to people about the capabilities of their system.

Other companies simply have less capable systems.
If I go and buy a current product Tesla, I can have it drive me home and chances are I won’t have to touch any controls. In a few cases, new production Teslas literally deliver themselves to the new owner’s driveway. Can any other automaker say the same?

I don’t think MS is out of touch with what the customer wants, I think they just stopped listening.

The fact is nobody is asking for this stuff.

I think the hope is that they build it anyway and then people will use/want it. AI is the big buzzword of the decade, just like ‘cloud’ was the big buzzword of last decade.

Significantly changed. Even in the last few months. I would encourage you to go do a test drive. Night and day from the type of experience you have.
The driver monitoring now uses a camera. If you are looking at the road, it doesn’t ask you to jerk the wheel at all.
Speed control is much more organic and considers turns, hills, etc. The machine vision on the cameras is different as well, it uses a processing technique called occupancy networks to produce 3D data out of the 2D camera images.

The one concern is you list speed in km, the current full self-driving software is not available in all countries and may not be available in yours, which might mean if you do a test drive you are still on the same very basic system you had before.

The core issue, IMHO, is a mixture of lack of critical thinking and intellectual laziness, reinforced by algorithms and echo chambers. You see it in almost any contentious debate these days, including things like politics, but it’s pretty much everywhere.

Whatever my opinion is, various algorithms will figure that out and feed me a solid stream of crap that agrees with me because that’s what I will click on and engage with. Every time I see an article that reinforces my opinion it gives me a little hit of dopamine that I am right and so I conclude that I am right and everybody smart agrees with me because my position is obviously the right one.
Meanwhile the guy on the other side of the issue has the exact same experience and thus is convinced that he is right and everybody smart agrees with him.

Combine this with an educational system that is teaching the test rather than teaching to think, and the very simple thought process of ‘what if I’m wrong? What if I don’t have all the details?’ simply doesn’t occur in an awful lot of people.

Elon Musk is a perfect example. A few years ago, he was a genius eccentric billionaire working to make the planet a better place with green technology and electric cars. Then he joined up with Trump, and suddenly he is a fraudster using Daddy’s money to bully his way into companies and taking credit for their success. The rockets are bad, the cars are bad, the tunnels are bad, the brain chip is bad, and all these things always were bad from the beginning because it’s easier to retcon than to acknowledge your position changed because of politics.

The fact is, in this age of information there is really no good excuse for ignorance. The information is always out there, if you put even a little effort into finding it. Yes it requires waiting through a lot of crap and slop, But it’s out there. And as you say you can just head down to your local dealer and ask for a test drive, and then you have real empirical data to base an argument on. Not that anyone would do that, because to them, their opinion is just as valid as my first hand experience.

Consider the difference between supervision and intervention.
All production Teslas need human supervision, this is enforced with driver monitoring systems as a safety procedure. But the current versions of FSD, released in the last few months, can often navigate through most or all driving situations without human intervention. So the computer will make sure you are paying attention, but will in most cases execute the drive perfectly without making mistakes that require the human to take over.

There’s plenty of videos on YouTube check some of them out :)

Curious when this was? And did it have FSD?

Every time my car gets an update and FSD gets better. Every time I get in my car and hit the FSD button.

Serious question- have you ever driven a Tesla? And if so for how long?

There is a significant difference between lane control and FSD. Lane control just keeps you in the lane so you don’t have to actively steer. FSD actually drives the car, changes lanes, makes turns, stops for traffic lights and stop signs, navigates intersections, etc. With the current v14, you can get in your car, type in a destination, and then not steer or push the pedals at all and the car will take you to a parking space at your destination. Lane control does not do that. I’m not aware of any other company that does that.