Intel has quite a few foundries in the US at various ages, ranging from 300mm up to the most current 20Å. They've now opened their foundries to other clients. While the capacity is not sufficient for true independence, I'd be interested to know why Huang didn't mention them once in his diatribe.

No, adding friends is opt-in, but they changed what information is shared with friends, and that change was opt-out and done without notice.

This mission is part of a program that is all about low-cost projects that are willing to take on more risk. The total cost of the mission was $80 million including $20 million for launch. A Falcon 9 launch costs around $70 million, which would have been more expensive than all the other hardware and operations costs combined. So if it works NASA get a great bargain, and whether it works or not they are helping develop competition in the launch market.

The trick has always been to extract heat from cold air or push heat into hot air. If an A/C doesn't work when the OAT is 90F (or 32C), it's worthless. Conversely, a heat pump that can't pull heat from 30F (or 0C) air has a very limited range.

Because a refrigerant's phase change occurs in two physical locations (expansion valve and condenser coils) separated by tubing, it works well for central HVAC. This device would need to cycle through those phases in one location, probably an outdoor unit. I think it would need to cycle the electrocaloric material through 4 temperature zones. Eg., when heating during old weather, starting with non-energized at ambient OAT:

• - heat the material to indoor temp
  - energize and extract super-heated fluid for heating indoor air
  - cool the material to outdoor temp
  - de-energize and extract sub-cold fluid for cooling outdoor air

The first and third steps could use a closed loop heat exchanger so you're not just wasting the heat extracted in step three, but it adds more complexity to developing a solution that's commercially viable. Plus, thermal stresses would be significant for a ceramic material that's cycling rapidly between say 90F in step 2 and 0F in step 4.

I hope it works. A/C is a huge energy sink that's only growing.

To state the obvious, a compressor is a type of pump. They both forcefully move a substance from a low pressure side to a high pressure side.

A typical A/C / heat pump has three electric motors that operate: the compressor that circulates the refrigerant between the condenser and evaporator, a fan on the outside unit (condenser), and a blower on the inside unit (evaporator), plus various reversing and expansion vales. I don't see how this device reduces the number of motors, as I don't expect it's feasible to size the electrocaloric material so it interfaces directly with blown air. Pumps moving fluid through of the electrocaloric material won't have as much pressure differential as a typical refrigerant compressor so would draw less current, but any savings there would need to be measured against the added draw needed to form the electric field.

You could say it's solid state, but no more so than the coils of the condenser and evaporator of a conventional unit are. The moving parts are the pieces that circulate refrigerant or air and the valves that control those flows. Those might change, but wouldn't go away with this system.

This will trend to the top of several sites, generate a slew of bad PR for Microsoft, and they will backtrack and defer it to some undetermined later date that will never come around. They will recover from the bad PR with a small uptick in official accessories from those that are afraid that error will come around again. Their next console will only allow official accessories from the start.

You heard it here first.

When Johnny's mom caught him chasing Sally with a hammer, Johnny sheepishly said, "But, Mom, I was just kidding. I wasn't going to really hit her."

Something along the lines of "I wasn't really going to do it" is what ALL the bad actors say when they are caught red-handed.

"I don't intend to get rid of a single US Postal carrier. I intend to hire more as new cities are built and need postal services. What I do intend on eliminating is the Pony Express and all its horses, stables, and stable hands, now that we have procured this new fleet of horseless carriages to deliver mail."
-- US Postmaster General, circa 1860

(BTW, that's a fictitious quote, made to point out that each new disruptive technology that improves all of our lives comes at the expense of disrupting some careers. The Pony Express was never actually part of the USPS. In the end, all those excess stablemen were no longer adding value, but there were new jobs opening as auto mechanics and gas station attendants.)

This theme repeats many times throughout history with inventions and technologies that we now consider essential but were first viewed with fear as entire industries were thrown into upheaval. The best way to prevent AI from replacing your job is to learn how to control the AI doing your job.

I would love to have more control of my lights, but the first thing I look for is local control WITHOUT the need for a hub or WiFi/Zigbee/Zwave/etc. Insteon mostly fits this bill, but is expensive and proprietary and its future is cloudy.

Here's what I would love to see:

Switches control the lights directly over Romex wiring. Even if there has to be some signaling over the line for dimming and color selection, it has to work without RF. This allows easy retrofitting of existing homes and won't confuse electricians who are accustomed to their way of wiring a house. It also allows an easy fall-back to conventional switches and bulbs.

Switches can update or be updated via RF by each other and/or a hub, but do not require it. The technology for that can be local or RF, or both as Insteon does. However, if the hub is down or inaccessible, local control via the switch MUST still be possible.

The hub must also be locally controlled; no cloud-based crap for me. HomeAssistant, OpenHAB, or something similar are the likely candidates at this level, but there is room for commercial providers or even cloud-based services for those that are ok with that.

I go to sell my house, and the elderly couple willing to pay cash look in the server closet, realize it's a necessity for controlling the lights, rip up their earnest money check and leave.

I go to sell my house and realize the extra $50k I spent on commercial-grade wiring, DMX controllers, and lights is all sunk cost that I'll never recover.

People expect residential goods in their residences. The likelyhood that you'll find a buyer that is as excited as you were to rewire your entire house for DMX lighting is next to nil.

Local control and the "Luddite father-in-law test" are key criteria for me when evaluating a potential HA product. In the USA, UL-certification and fire insurance clarity are a close second.

So the moral of the story is early buyers will pay full price while getting a buggy, unbalanced, unfinished product. Meanwhile, those who wait will generally get discounts, see fewer bugs, and more polished content.

This is why I almost never buy anything as soon as it's released.

Keep it up, shitheads. People will eventually come around to the idea of self-hosting again.

When Jeff coined the "disagree and commit" phrase, what is he asking for in a commitment? If I disagree, should I commit to a belief that my stance is the correct one, or commit going with the flow even though my stance is being overridden?

I can see the value in saying if you're going to disagree with what's happening, be vocal and committed to your stance. That's how you ensure you're not surrounded by sycophants. However, once your disagreement has been aired and discussed, it sounds like he's asking everyone to commit to the RTO even though it seems it's driven entirely from the top.

Have you tried Zig? It claims to solve the most common memory-safety issues but without many of the blockers you seem to associate with Rust. Coming from a C background, I personally find it closer to my ideal, but my ideal isn't the same as most rustaceans...

Also not (necessarily) true. Dissolved CO2 can be consumed by algae and seagrasses in the ocean, converting it to biomass -- where carbon remains largely sequestered when the organism dies -- and oxygen.