pbmonster

As long as you have well suited oil/gas wells to put compressed CO2 into, you'd much rather do that.

DAC of a ton of CO2 costs around ~2 MWh. Compressing it into a well costs ~0.5 MWh, but low temperature reduction to solid graphite costs ~10 MWh.

Solid storage in soil is much more interesting in bio char applications, because the char is full of nutrients and a good water absorber.

Of course, direct air capture of CO2 is extremely energy intense. You can only do it if you've built excessive renewable energy overproduction. And I mean unimaginable excessive.

We've emitted more than a trillion tons of CO2 since 1990. If we want to put just that back into the ground (pump it into old oil wells), the final storage will be several thousand cubic kilometers of high pressure CO2. Just the direct air capture will cost more than an Exawatt hour of renewable electricity (currently, humanity produces around 30 PWh of electricity per year from all sources).

You can use some tricks. Turning all agricultural waste, forestry waste and organic garbage into bio char (instead of letting it decompose on the fields or in compost facilities, which emits CO2) captures a few billion tons per year, but that's far more labor intense and requires a lot of automation work.

Not a massive hurdle. BMW got the i3 to be street legal in the US without much trouble. And I know, the i3 was only a smaller car, it wasn't a cheap car (I think they went for around $40k new)... but that's just BWM.

Getting Renault e-Twingo / Citroen e-C3 street legal in the US wouldn't really rise the production cost much. But nobody is going through all the trouble, because (obviously) almost no American would buy such a car, even though they start at <$20k new.

> For solar, seasonality would be best addressed by thermal storage

Still, adds a lot of cost and complexity onto cheap solar + 12h of chemical battery storage.

> This is being pursued most interestingly by Standard Thermal's "hot dirt" approach.

Pity they want to do only 400°C. Because once you go towards firebricks and >700°C, you can run practically all thermal chemical processes in a refinery or a polymer plant off the heat, and you can make electricity with a supercritical CO2 turbine at >50% efficiency (in a comparatively tiny package). At the cost of much harder engineering problems, of course.

But its probably worth solving those, because you can do interesting stuff with ultra low cost hot gas. Preheat the charge for a steel mill, design a Hall–Héroult aluminium smelter that keeps the electrolyte bath molten via a gas heat exchanger instead of electricity, ect.

And once you have several GW of guaranteed heat consumers, adding a couple of CO2 peaker turbines isn't such a big leap anymore.

I always wish maps like that would include some sort of seasonality score. Maybe a worst week / best week ratio.

As it is, the map is useful until solar goes to a double digit percentage of total yearly generation. Any higher, and seasonal places start needing backup generation capacity that is only running in winter (or during monsoon), which does violent things to electricity markets.

No, that's the Sichuan Basin. The climate is subtropical with monsoonal rain all summer and dry and cloudy winters. It basically has persistent cloud cover all year and heavy fog. Less than 30 days of sun per year.

> Micro and nanoplastics are also a problem that urgently needs to be addressed.

Bio-plastics won't do this automatically. You can make today's polymers from biomass feedstock. Those will have the exact same micro/nano plastic problems.

You can also make polymeres, that are much more bio-degradable, with the associated particulate matter decaying instead of accumulating forever - but you could make those from crude oil just as well, and those degradable polymeres bring an entirely different zoo of problems with them.

> Mercedes is certainly one of them which has lost it's "our engines are nicer" brand

Mercedes is doing something interesting with its level 3 DRIVE PILOT. As far as I know, it's the only autonomous driving system/auto pilot on the market that actually takes full legal responsibility for the vehicle and its actions. If you put a new Mercedes in self driving and it crashes, Mercedes takes full liability.

Yes, it currently only goes up to 100 kph and it will buzz you to take over if it can't guarantee prefect safety anymore. But this approach is really the only one I think is interesting. All this self driving shit is worthless to me if I'm still liable in the end. So I respect Mercedes for putting their money where their mouth is.

How did you get that number? Is this just a quip about batteries?

Most of India has a solar capacity factor >0.2 and low seasonality. That means even 5kW peak rooftop panels and a 10 kWh battery will get you over 95% almost always. And panels+labor is so cheap in India, there's really no reason not to just tile the roof in panels.

Things get more difficult in dense cities, if you have multi story apartment blocks going 90+% locally is impossible. But then just transmit power from lower density rooftops, or actual solar farms.

And this already assumes widespread AC (really no way to increase electrification this much and not add AC in that climate).

> "that's on me, I'm owning that, I'm sorry, that shouldn't have happened"

Does it auto-update now? I like its UI better than PipePipe, but I got so sick of hand-updating Vanced (including hand-downloading the correct apk and waiting several minutes for the patches to run against it) every time Google breaks Youtube again. There was a period of Warfare last year where Google would make breaking changes more than once a week...

PipePipe is so nice with the fast auto-updates. If I ever even catch it breaking, I just have to use Firefox for a few hours and then F-Droid ships the fix and PipePipe works again.

But you would need maglev if you wanted to compete on energy per ton of cargo per mile. Ships are several times more efficient than classic trains.

> At 2% of sales, the opportunity cost is the reason to kill the product. Keep in mind this still needs design, manufacturing, marketing, and distribution resources.

All this is true for many other products, specifically cars. Ford makes the Bronco Sport anyway, and they make money on it.

I get it, Apple has much better margins than Ford. And maybe a iPhone 18 Mini would reduce total margin for them. But if they move a couple of million units, it absolutely would make money.

> enough ammonia in the air will suffocate you.

It will suffocate you on account of all that blood pulp in your lungs. And eyes, and nasal mucosa...

Having a high pressure ammonia leak indoors is a death sentence for anybody in the same room. Same building, if it's large enough...

Technically, you could also just pull refrigerant lines from the heat pump to the bath tub - you probably already run those to the indoor AC units you have in different rooms anyway - or connect the bath tub to the hot water lines you run to the radiators. Both imply a grey water heat exchanger built into the tub.

Depends on setup, of course. If there's a central AC blowing hot/cold air through ducts, you're stuck using a grey water loop. But it is the least attractive setup...

Total consumer-price inflation has been about 27.7% since 2021.

Inflation is multiplicative, so: 1.277 * 1.26 = 1.61 giving us the 61% subscriptions went up.

Of course, they never even brought it to the US. It has no chance there, there's barely a market for lowslung station wagons, and close to no market for EV wagons. Always was a niche in the US, while it was always the primary family vehicle in Europe.

And yes, VW is fighting for survival. The things they are killing are the low volume models specific to the Chinese market (because absolutely nobody in China is buying any VWs) and a lot of the variants and overlaps they've accumulated over the years. They'll continue making less than 100k iD.7s, and will be happy to do so (it's modestly profitable).

I don't get that mentality. Millions of units, shares many parts with other models, OK margin. This is no reason to kill a product.

For comparison, that's roughly like Ford killing the Bronco Sport, or VW killing the ID.7.

That's not correct for a SOTA model with trillions of parameters. Those have immense amounts of knowledge trained into their parameters. Try it. It "remembers" the ending of random books, Goosebumps and otherwise.

That's the entire point of why knowledge cutoff is so important (if you use them offline).

I find it entirely believable that the Navier-Stokes conversation was auto-flagged as high value training data, and burned-in the models knowlegdge base.

But this is different, right?

The equivalent would be taking a (fully offline) LLM and asking it about the ending of one specific Goosebumps book, and it revealing the twist. And although that specific book was (probably) only once in the training data, a high parameter LLM can usually "remember" the twist.

> Is that not self evident by the insane revenue from frontier labs?

No...? Of course not?

Because revenue is only one side of the equation. Did you ever look at total cumulative OPEX and CAPEX, and how long it will take them to even just break even at current growth?

> You would think that scientists would spend more effort trying to create "super" rhizobacteria that can colonize even more aggressively and also fix many times more nitrogen than ordinary ones

I think the problem with that idea is that the rhizobacteria needs energy to break the nitrogen tripple bond. A lot of energy. It gets that energy by getting carbohydrates from the host plant - which in turn are not available to the plant to grow.

I'm sure there's plants where not fertilizing with nitrogen at all would make the trade-off worth it, but the vast majority of our optimized crops can't spare double-digit percentage points of their total yield to feed bacteria. You're much better of just using natural gas (or green hydrogen) to fix the nitrogen.

The N-gram parameters can be fetched from SSD, with maybe the hottest ones staying in memory.

Compete with Garmin instead of Apple.

Good news: nobody does.

From the hardware side, extremely precise. Most discussions assume the sweet spot for ISAC is in "Frequency Range 3" (7 - 24 GHz, decent penetration and range, very good resolution), and in that range you can pin down objects with millimeter precision (both position and velocity). So, not only can you detect people walking down the street, you can detect them breathing when they stand perfectly still.

This, of course, requires 6G base stations to be spaced pretty tightly (hundreds of meters apart) and to feature sophisticated phased array antennas. But both can be assumed in cities (they're planning to completely replace WIFI with 6G, so every current WIFI access point will become a 6G micro cell). Walking humans can be detected much easier, but radar SNR is funny, that only gives you 4x-6x the detection range.

Software (signals processing) for all that will be... a lot of difficult work. Object separation/tracking, multipath reflections, clutter suppression, ect.

It's a drop in a bucket. Global warming is due to ~5e18 Wh of excess heating. A nuke dumping 5e12 Wh of additional heat into water makes no difference at all.

The article poses the question "What do you want your phone/computer to do that it can’t", and suggests that there really is nothing left - software is done, all remaining problems are hardware.

I can't think of many things, but there are a few: I want an LLM teaching 1 on 1 language lessons. Good ones, with tailored vocab, tailored reading, tailored listening, grammar lessons, and (most importantly): conversation training including it correcting my pronunciation and grammar.

Current LLMs can do most of that when prompted correctly (and supplied with textbooks and podcasts), just the last part is missing.

But what else is there?

> the issue is that when a phone is unlocked, they can just plug in a USB device and scrape everything off it

That's useless if the backup account and the main account use different keys for the home partition.

VeraCrypt (used to?) have this. It was called a hidden volume. One volume, but two keys, two passwords, and two different containers full of data. Technically, the second volume is written into the partition "from the back" using key 2, while the first volume is written "from the front" using key 1.

Fun fact, there's no protection against writing over data in the other volume if volume 1 + volume 2 exceed the size of the partition - and there can't be, otherwise the volume wouldn't be hidden.