Reading some more, they do acknowledge this:

"The bismuth electrode is inexpensive and can robustly modulate the pH of chloridecontaining salt water regardless of the type and concentration of cations. On the other hand, although silver has the advantage of large capacity and low energy penalty, electrode dissolutionwill need to be overcome by modifying the silver particles or replacing silver with another material; this is to be the subject of future studies."

Seems like a pretty relevant caveat for New Atlas to skip over.

"Methods to overcome the fouling of electrodes during the regeneration step owing to local high surface pH conditions that promote the formation of precipitates such as Mg(OH)2 must be developed, and several engineering and electrochemical approaches are currently being considered to alleviate the fouling issues; one such method was discussed in which the degasified water was mixed with fresh ocean water before being fed to the regeneration cell to reduce the overall pH increase in the flow channel."

I hope they figure it out.

It seems to me that the recapturing of protons won't work perfectly, and some will wash out to sea. Simularly, anyone with a boat knows sea water has a lot of gunk in it, and some of it will be attracted to the electrodes, or just naturally stick to them, further degrading performance.

The New Atas article talks about this briefly, but doesn't say anything about the material the electrodes are made of, and what it costs to replace them:

> And mineral precipitates are fouling the electrodes on the alkalinization side, so there's plenty of progress yet to be made.

The research paper does shead light on this:

> The material and electrochemical characteristics of the Ag and Bi electrodes employed in the electrochemical system will determine the efficacy of the process.

Ag is silver. 933 euro per kilogram.

Bi is Bismuth. $390 per kilogram.

Sounds expensive to just dissolve into sea water.

> devices don't require any cloud access to work

Don't require is not the same as "will not use". Check the Matter promo videos, they all talk about how it also becomes much easier to send data to the cloud. In practise even onboarding will likely use cloud-based certification/authentification. https://csa-iot.org/newsroom/m...

> destroy their flash memory

There is a middle ground between storing data on the end-point devices and in the cloud: local smart home controllers. IKEA has done it this way for years, as has Home Assistant. Matter is a wasted opportunity for enforcing data to be stored in the home only.

> "trivially easy with even a consumer grade firewall"

Only if you want to block everything, which you don't. In practice you'll have to unravel which connections you'll allow through and which you want to block. Good luck separating firmware update requests from data exfiltration for tons of different vendors. And you're also ignoring that consumers wouldn't even know how to engage the 'nuclear option'.

> carefully designed to allow that to operate

Again, coulda woulda shoulda. Most consumers will simply buy commercial devices and have their data shipped off to the cloud. When a few scandals later they realize they want to have more control over where data flows, they'll realize they need to follow an IT course to make that work. Matter's designers could have built consumer-friendly control over data flows into the spec, but they didn't.

What's baffling to me is that Zigbee was designed to mitigate most of these risks.

> Have you even read the spec?

It's 900 pages. I'm currently adding Matter support to a project. I've read the relevant parts, watched all the videos, and had lots of discussions.

It does not do the opposite, it does both.

Bingo. There are so many (privacy) issues with Matter.

- Devices can talk locally, which is great, but they can also more easily connect to the cloud. Since "hubs are evil", according to the Matter creators, storing data logs is implied to happen in the cloud.
- As you say, any border-gateway will allow even very simply low-power devices to connect to the internet. You have to secure all these border-gateways somehow, which will be extremely hard. There is no single point of control anymore, as "Matter sees privacy as damage, and routes around it".
- Vendors have even more options for custom extensions than they had with Zigbee, where already companies like Xiaomi and Tuya made incompatible devices. The expectation is that most Matter devices will expose some basic features locally, but still require specialised apps for extra features.
- A security best-practise was to move your IoT devices to a separate Wi-Fi network, so that any trojan horses could not easily access users' personal devices. With Matter that credo seems to have gone out the window. Where Zigbee still has built-in separation of networks, they moved away from that.

A lot of these boil down to "Matter makes compartmentalization harder".

There are other red flags. The Matter website boasts of "improved privacy", but when you look closer at that claim, it's about having improved encryption stength. A classic misunderstanding of privacy. https://csa-iot.org/developer-...

Similarly, the Development goals of Matter, as listed on Github, don't even mention privacy:
https://github.com/project-chi...

I'm pretty worried.

It's not that black and white. You can make it less seductive to spy on each other, as well as cultivate a culture of trust. Take a look at:
https://www.candlesmarthome.co... for an example in a smart home context.

Governments telling industry "what to do" has often succeeded. Often the government dares to invest in projects where the marker is too risk averse.

Take for example Denmark and robotics
https://www.universal-robots.c...

Or how Angela Merkel pushed for battery cell production in Europe when the German automakers were dragging their feet.
https://www.reuters.com/articl...

And let's not forget where the internet came from.

My point being that blanket statements like that make me wary of your otherwise valid arguments.

As hype cycles go, I suspect drones will have quite a large plateau. Especially in the military. For example, I believe one of the largest worries right now is "swarming attacks", which are one of the only real threats against aircraft carriers.

In short, I think that is a smart move. Smarter than investing in the Metaverse, blockchain or quantum tech anyway.

I just installed it, opened it, and the first thing I see is a warning from my outgoing firewall that it's trying to connect to improving.duckduckgo.com.

Come on. How hard is it to ask people if they want to opt-in to that? It's calling home before even showing me a privacy policy.

My point is that Matter lightbulbs will be able to autonomously communicate with the internet, where Zigbee lightbulbs could not. That makes them more dangerous.

Zigbee was well designed, as devices could not autonomously connect to the internet. For example, a zigbee light bulb could not autonomously communicate with a cloud server.

That protection is now being removed, as Matter introduces an IP protocol layer. Matter is designed to increase data extraction, and as such it shouldn't be a suprise it is being pushed by Google.

It's also likely to increase security issues for the very same reason. Most IoT security issues are with devices that use WiFi.

I once visited an evening about Ethereum relatively early in its development. I was surpised to learn that most of the available "dapps" on the ethereum blockchain were about gambling. Many even refered to themselves being pyramids in the name or description. You were basically 'gambling' that you were one of the early adopters. That there would be more suckers coming in after you. What struck me most was that this was deemed perfectly normal.

It's one of the things I only later realised: if you have lots of crypto currency (say a billion dollars in bitcoin), you're not rich yet. At some point you will need real world people who will want to buy all those tokens from you for actual money. If trust in the token diminishes you might not be able to cash out.

So yeah, it really is a big pyramid scheme in that sense, reliant on the influx of new suckers.

The link is paywalled.

Curiously I couldn't find a mention on the rover's blog:
https://mars.nasa.gov/msl/miss...

Also no mention of this in the Nasa press release from two days ago.
https://mars.nasa.gov/news/924...

Both AppLovin and IronSource seem like surveillance nightmares...

IronSource:
"Increase your app’s ARPDAU and maximize the value of your users with powerful ad mediation and high-impact ad formats"

AppLovin:
"Receive the best value for each user through our real-time auction that optimizes based on your parameters."

*le sigh

I've always wondered if this would create strange shadow patterns on the earth. I wonder if the bubbles let light through in a specific pattern. Or are these things far enough away for light to bend around it?

Isn't a fundamental downside of bubbles that they are material ineffecient? It's basically two layers, with the side parts it perpendicular to the sun, so not very effective. The sides of the bubbles might also bounce light towards their neighbours, not helping with heat management?

I always imagined a slit-design. Think: window blinds, with paralel beams of material that can together rotate to block more or less light. It seems this design could vary the amount of bubbles instead? Imagine if we could allow more sun through to the hemisphere where it's winter, while at the same time blocking light on the hemisphere where it's summer.

Aside: odd that this project seems to be spearheaded by MIT's Sensible City lab, which is normallty engaged with smart cities research.

Interesting, as Google's research pointed to something else: psychological security.

https://duckduckgo.com/?q=goog...

(I remember because of the irony, where Google's surveillance businessmodel is about measuring everything, which as a managerial style is often harmful to the sense of psychological security)