philipkglass

Yes, this does happen. Solar farms have an albedo modification effect. The good news is that compared to fossil fuel sources, the albedo modification is "paid back" in less than a year due to lower GHG emissions:

"Small reduction in land surface albedo due to solar panel expansion worldwide"

https://www.nature.com/articles/s43247-024-01619-w

If you download the supplementary information PDF, figure 19 shows a plot of GHG payback times by site. The most common time is about 1 month. The greatest time they found was about 9 months. These times go down as solar panel efficiencies continue to creep upward, because solar panels in 2026 don't have lower albedo than in 2024, but they do have better light-to-electricity conversion efficiency.

Ferrocyanide is not dangerous. Sodium ferrocyanide is used as an anti-caking agent in table salt. It's considered safe for this use in both the US and the EU, e.g.:

https://ec.europa.eu/food/food-feed-portal/screen/food-addit...

Dimethylmercury is (or was, at least historically) used as a reference material for nuclear magnetic resonance (a common analytical method used in chemistry):

https://sites.northwestern.edu/ohalloran/199hg-nmr-standards...

This MIT incident was in a chemistry building that has been in use since 1970, so it could have been accidental release of an old, unused container or sample tube of the material.

Yes:

"Crushed rock sand – An economical and ecological alternative to natural sand to optimize concrete mix"

https://www.sciencedirect.com/science/article/pii/S221302091...

The main problem is that even though crushing rocks is affordable, illegal sand mining is even cheaper in regions where law enforcement is weak.

The Soviet Union ran such a program:

https://www.nytimes.com/1973/11/10/archives/soviet-plant-to-...

https://www.cia.gov/readingroom/docs/CIA-RDP91T01115R0001000...

The term to search for is "single cell protein."

There are newer approaches to making single cell protein from carbon dioxide and renewable energy instead of fossil oil, like Solein:

https://www.solein.com/

There are a few problems with turning oil into single cell protein:

- Microorganisms have small cells that reproduce quickly, which means a lot of nucleic acids in the dry mass, which is trouble for animals (including humans) that can't tolerate a very high purine level in the diet.

- The abundant cell walls of microorganisms also can cause illness when they're a high proportion of dietary intake.

- Feeding single cell protein directly to people has to overcome traditions of cuisine, even in cases where it doesn't actually make people sick. Indirect usage of SCP (e.g. feeding it to salmon, and then having people eat salmon) can overcome that hurdle, but then the cost per human-edible calorie goes up too.

Urenco USA is a private enterprise that enriches uranium. It built its first American enrichment plant 20 years ago, and lately it has been expanding capacity:

https://urencousa.com/about/our-history

The parent Urenco company has been operating private uranium enrichment plants in Europe since the 1970s.

The Erlang day was on 2009-03-11. You can still see how the front page looked that day:

https://news.ycombinator.com/front?day=2009-03-11

The newest American nuclear reactors to enter service were Vogtle units 3 and 4, in 2023 and 2024:

https://en.wikipedia.org/wiki/Vogtle_Electric_Generating_Pla...

If Anthropic is the first to develop AGI, then everyone else becomes subordinate to Dario Amodei? And if DeepSeek is the first to get there, everyone instead becomes subordinate to Liang Wenfeng? I cannot believe that there is a threshold of intelligence where getting there first means a decisive and durable advantage. If you do believe that getting there first means ruling the world, then our viewpoints are probably too far apart for mutual understanding.

Copy the technology and give people access without trying to financially profit from it.

Even if the US federal government refuses to act, smaller government entities can deploy their own compute clusters running strong open-weights models like Kimi K3, and provide their own citizens with free access to strong AI models. There may come a time when even Chinese providers no longer release strong model weights, but then the previous generation of strong open-weights models can be used to accelerate the development of data pipelines and training setups to build the next generation. California or Australia (or many others) could fund university-led R&D efforts to build completely open models.

Despite the angry noises that leading providers make about their model outputs being used to train competing models, that's not illegal. States can use Common Crawl, Anna's Archive, Books3, commercial model outputs, and whatever other data sources that are useful for building models. States have sovereign immunity to ignore copyright complaints, if they decide to pursue model development.

The big AI companies don't have much of a moat. Let the big companies spend a fortune figuring out what works best, then openly build imitations having 95% of the capability at 5% of the cost.

As far as I'm concerned, the right choice is that the US government learns to live with remotely secure devices in the hands of everyone. No new laws are passed to force hardware/software makers to insert remote backdoors. Law enforcement and intelligence services have to investigate targets using metadata, publicly posted information, the numerous online service providers who are already subject to warrants, and physically proximate surveillance.

I'm using Qwen 3.5 for OCR, and reasoning_effort is supported there too. I found that it can be loop-prone (though somewhat less so) even if you set reasoning_effort to low.

Amazon is also investing in nuclear power for data centers:

https://www.world-nuclear-news.org/articles/amazon-updates-s...

However, building new nuclear reactors is a slow process. These reactors aren't expected to be built before the 2030s.

China is currently the fastest at bringing new reactors online. Even there it takes 6-7 years from construction start to commercial power generation, e.g. https://www.world-nuclear-news.org/articles/chinese-reactors...

The 32 oz "Big Gulp" was introduced by 7-11 in 1976 in Southern California:

https://en.wikipedia.org/wiki/Big_Gulp

Even bigger sizes arrived in the 1980s:

Called Big Gulp because it was initially the largest such drink available at any retailer, 7-Eleven eventually introduced larger sizes. In 1986, they introduced the 44-US-fluid-ounce (1,300 ml) Super Big Gulp, followed by the 64-US-fluid-ounce (1,900 ml) Double Gulp in 1989 (later reduced to 50-US-fluid-ounce (1,500 ml))...

Google may not have written about it in that document, but they're funding what will be the highest storage capacity battery system in the world for a renewable-powered data center in Minnesota:

https://www.utilitydive.com/news/worlds-largest-grid-battery...

It's unfortunately worse than coal subsidies. There are direct Federal orders that obsolete coal plants are not allowed to shut down. One example from Washington state:

https://www.opb.org/article/2026/06/18/federal-order-keeps-w...

Yes, bleach could destroy the complex odoriferous compounds in fish sauce. Ordinary chlorine bleach in liquid form is so dilute that you wouldn't be able to run the process in-place in the original fish sauce storage tanks, though. There are more concentrated forms like granular calcium hypochlorite that would take up less volume, but the heat generated by oxidizing organic matter on this scale makes it a sort of chemical engineering problem.

I read the whole series on the Kindle about 10 years ago and the obscure nautical words were not in the Kindle's internal dictionary. It would try to connect to an online service to look up those words, but that doesn't help if you are on an airplane. Has the dictionary grown more comprehensive in newer Kindle models?

The German THTR-300 reactor operated with air cooling:

https://en.wikipedia.org/wiki/THTR-300

But it was admittedly a huge cooling system for a reactor that only produced 750 megawatts of thermal energy.

I have an uncle with Parkinson's disease. He has a daughter who's looking after him. But he is a tall man who weighs about 200 pounds and she is a petite woman. She literally doesn't have the strength to help him up again if he falls. She has to call emergency services.

Google was scanning millions of books before the LLM craze, and they didn't burn them. They partnered with libraries and built their own non-destructive scanning hardware. I'm guessing this comment is really about Anthropic, but they are distinct companies.

https://en.wikipedia.org/wiki/Google_Books#Scanning_of_books

I know a lot more about chemistry and physics than I know about biology, so biological risk is still at least a "maybe" reason to be worried about model safety for me. I don't worry about other weapons of mass destruction developed with LLM assistance. Nuclear weapons are already controlled much more with political pressure and the physical difficulty of e.g. enriching uranium than they are controlled by secret-keeping.

For all I know (I don't know a lot about biology), it's possible to develop a virus that spreads as easily as measles, but has the lethality and long induction period of HIV. If that's true, and if LLMs would make it easy to develop such a virus, I see how no-guardrails LLMs could be truly disastrous.

The other justifications for restricting model training don't seem nearly as compelling. Worse, restrictions on model training have an infectious quality if they mean something more than "leading AI labs will voluntarily slow down together." (Like if the signatories mean that better models and software for building them should be regulated.) It's too dangerous to publish stronger models. So it's also too dangerous to publish training software. So it's also too dangerous to publish tools that allow you to build training software. (Claude's Fable won't help you write generic model training pipelines for this very reason.)

If you keep following the implications of this safety argument, it's as broad an assault on the distribution of software and computing as has ever been proposed.

It was a front page story here on HN back in December 2025:

https://news.ycombinator.com/item?id=46313991

The biggest problem is the infectious nature of restrictions that start narrowly. Fable is too touchy about helping people with biology and chemistry problems. It was initially released with an even more insidious safety mandate:

https://simonwillison.net/2026/Jun/10/if-claude-fable-stops-...

In light of the ability of recent models to accelerate their own development, we’ve implemented new interventions that limit Claude’s effectiveness for requests targeting frontier LLM development (for example, on building pretraining pipelines, distributed training infrastructure, or ML accelerator design).

...

Unlike our interventions for cybersecurity, biology and chemistry, and distillation attempts, these safeguards will not be visible to the user. Fable 5 will not fall back to a different model. Instead, the safeguards will limit effectiveness through methods such as prompt modification, steering vectors, or parameter-efficient fine-tuning (PEFT).

(And although the "silent" downgrade part was quickly dropped, Fable still won't help you here.)

Anthropic won't teach you how to build bioweapons, or enable you to make your own software infrastructure so that you can train your own biology model. That's where lawmakers may arrive too if they buy Anthropic-style safety arguments. It's too dangerous to publish models that understand biology. It's too dangerous to publish training software. It's too dangerous to publish tools that allow you to build training software.

If you keep following the implications of their safety argument, it's as broad an assault on the distribution of software and computing as has ever been proposed. Worse than the Clipper Chip proposal of the 1990s era Crypto Wars. I have seen how "children must be protected online" has in practice turned into an attack on adult privacy affecting a wide swath of services and devices. I'm taking a maximalist position on openness now because I think that I can anticipate the next steps on the safety side, and I reject those steps.

I wouldn't object to a government advisory body that tests models for safety so that users can make informed decisions. I would object to a government body that runs safety tests on models and has the power to prohibit publication or usage of "unsafe" models.

That's a good example, but I'm unsettled by Anthropic's growing refusals in the areas of chemistry and biology. If they think that scientific assistant models should be as unhelpful as Fable, because applied scientific knowledge is inherently dangerous, I don't want Anthropic or like-minded thinkers setting the standards for model safety.

How much biogas per year would a site need to produce to justify installing a unit? How much on-site labor is needed to run it?

The reaction N2 + 3 H2 => 2 NH3 is thermodynamically spontaneously but kinetically forbidden by the strength of the N-N bond. That's why the key breakthrough of the Haber-Bosch process is the use of a doped iron oxide (magnetite -- is that what you meant?) catalyst that allows the reaction to proceed at an industrially useful rate. This wind powered ammonia plant gets the hydrogen from water electrolysis instead of steam-reformed methane; the plant is otherwise using a conventional Haber-Bosch process.

Using electrolytic hydrogen to make ammonia was fairly common in the first half of the 20th century, because a lot of remote hydroelectric plants didn't immediately have uses for all of the electricity they could generate. They turned low cost electricity into pure hydrogen and used it to make ammonia and derivatives, which were easier-to-transport high value products.

We would probably have large scale use of renewable ammonia by now if fracking hadn't made natural gas so cheap. Natural gas is the largest variable-cost input to making most ammonia (it's processed with steam reforming to generate hydrogen), and in the United States it cost more 20 years ago than it does now:

https://fred.stlouisfed.org/series/DHHNGSP

The cost of wind power and solar power have fallen dramatically, but without accounting for the externalities of CO2 emissions, natural gas is still the cheaper way to make ammonia.

The only chemical inputs a Haber-Bosch plant needs are nitrogen and hydrogen. The nitrogen always comes from air and in this plant the hydrogen comes from electrolysis. The plant also needs electrical power to run compressors, but that can also be supplied by wind and it's minor compared to the embodied energy of the hydrogen input.