This is exactly what I want: To communicate with me, have your LLM expand your message to include relevant parts of _your_ context, then I will have my LLM summarize it as briefly as possible with respect to _my_ context.
1,299 karma · joined September 18, 2015
As a non-native English-speaker, I would say that the `v`/`w` pronunciation distinction would be a major challenge: Asking someone to pronounce "Darth Vader" is a pretty good check, at least in the Nordics.
This might be a Nordic thing, but being able to handle rear wheel lockups is a necessary skill when driving on winter roads. Once you overlook a section of black ice, not even the nicest hydraulic disc brakes will be gentle enough.
I sometimes put out the question anyway and added an answer. Never quite figured out if this was considered bad style.
The second and third examples are safety lockouts [0] working as intended: Some system is locked in the off state to ensure safe access for technicians.
Especially the padlock lockout is simple and effective: As long as you have the key in your pocket, you can be sure than no one is going to turn on the meat grinder you are cleaning.
With some agreement on mapping (maybe just `%HH` for anything outside `A-Z a-z 0-9 . _ -`), this could be completely standardized and made part of standard library argument parsers.
I could see a bunch of my utility scripts replaced with a python script and a `uv` shebang if this was in argparse.
I can sort of understand why beepers where a cool sales gimmick back when the microwave was the only appliance with a micro controller, but really -- it doesn't make sense: Firstly, immediate attention is not critical when the time is up: unlike a stove or an oven, energy transfer stop the moment the magnetron is de-energized. Secondly, the microwave (at least my microwave) is not exactly silent: if you are not deaf, chances are you can easily tell when it is done.
Maybe I should apply the Joe-treatment from my old lab: whenever there was a new shipment of frequency meters for the lab (we always needed more), Joe would meticulously unbox them and stick a pointed screw-driver through all the piezo buzzers to make sure the would never make a sound.
[Edit] microtron (sic) -> magnetron
[0]: https://www.researchgate.net/publication/321225042_A_novel_w...
My hope is that in 5 years, I will not have anything in my feeds that have not been signed in a way that I can assign a trust level.
Here in the Nordics, we are already seeing messaging apps such as [hudd] that require government issued ID to sign in. I want this to spread to everything from podcasts and old-school journalism to the soccer-club newsletter, so that I can always connect a piece of information back to a responsible source.
[hudd]: (https://about.hudd.dk/))
To truly record an appearance without reference to the sensory system of our species, you would need to encode the full electromagnetic spectrum from each point. Even then, you would still need to decide on a cutoff for the spectrum.
...and hope that nobody ever told you about coherence phenomena.
Put differently: If you used the same amount of energy to heat one bucket of sand by 200C (A) or two bucket of sands by 100C (B), you would be able to recover more electric energy from case A because of the fundamental Carnot Limit. This is why sand is a good storage medium (as opposed to e.g. water), and why some solar power systems work with molten salts. Also why steam-based power plants need to operate at high pressure to be able to obtain high-temperature steam.
And I completely GP, having been in or in contact with academic research since the late 90's, there has been a very strong shift from a culture where the faculty had means for independent research, and were trusted to find their own direction, to the system we have today where a research project has much tighter overlook and reporting than most corporate projects.
A professor with a 4-5 person group will typically need two staggered pipelines of 4-5year funding projects to run risk free. In the EU it is virtually impossible to get funding for projects that do not involve multiple countries, so you need to set up and nurture partnerships for each project. Coordination the application process for these consortia is a major hassle and often outsourced at a rate of 50kEUR + win bonus. And you of course need to run multiple applications to make sure to get anything. When I talked to mentors about joining academia around 2010, the most common response was "don't".
[0]: https://en.wikipedia.org/wiki/New_public_management [1]: https://en.wikipedia.org/wiki/Financialization
My understanding is that this is true for all the Trump handouts: otherwise the ten-year economic outlooks would have cratered. The Economist had a couple of nice analyses on this.
Of course this means that the next administration will need to start with tax increases just to get to neutral, but maybe that is a feature?
[0]: https://lyra.horse/blog/2025/08/you-dont-need-js/#fn:10
As long as the controller is made sufficiently conservative, there is no fundamental problem: you limit the current according to the cell that heats the fastest and shut down once one cell is near depletion.
Maybe they have even gone a more aggressive route and build a balancing circuit that can route significant current around a low-capacity cell. Or maybe just charging logic to keep as many cells as possible in the 20-80 regime if they will be limited by low-capacity members anyway. There are so many options here.
Shor's paper on polynomial time factoring is from 1997, first real demonstration of quantum hardware (Monroe et al.) is from 1995: Yes, quantum has had decades -- but only barely, and is has certainly only now started to have generations.
To look at the kind of progress this means, take a look of some of the recent phd spinouts of leading research groups (Oxford Ionics etc.): There are a lot of organisations with nothing but engineering to go before they reach fault tolerance.
When I came back to quantum three years ago, fault tolerance was still to be based on the surface code ideas that floated when I did my phd ('04). Today, after everyone has started looking harder, it turns out that a bit of long-range connectivity can cut the error correction overhead by orders of magnitude (see recent public posts by IBM Quantum): The goalposts for fault tolerance are moving in the right direction.
And this is the key thing about quantum computing: you need error correction, and you need to do it with the same error-prone hardware that you correct for. There is a threshold hardware quality that will let you do this at a reasonable overhead, and before you reach this threshold all you have is a fancy random number generator.
But yes, feel free to be a pessimist -- just remember to own it when quantum happens in a few years.
I get the impression that maybe you are not aware of what is implied by the 'buy, borrow, die' tax strategy that parent is referring to: in effect it completely eliminates taxes on capital gains (in the US). [0] and [1] below are first two Google hits.
[0]: https://www.reddit.com/r/BuyBorrowDieExplained/comments/1f26... [1]: https://www.wsj.com/finance/investing/buy-borrow-die-how-ric...
Source: worked on CAPEX and yield estimates for major player operating in this sector for a decade.
My initial take was to share this with my son who used to built all kinds of things in Scratch, but I know that the lack of simple install instructions will be a deal-breaker for him.
Given that (part of) your audience will be persons like my son graduating out of graphical programming in Scratch, it might be worth spending a little time on non-dev install instructions. Even more so as you can leverage the rust toolchain and just suggest something along `cargo install --git ...` [0], without even publishing as a crate.
Apologies if this comes across as entitled: I just want to communicate that a single README-sentence on top of the work you already put into this would make it significantly more accessible.
[0]: https://doc.rust-lang.org/cargo/commands/cargo-install.html
My wet sourdough is 1 part flour to 3 parts water. As noted in a sibling comment, this favors the sour parts (lactic acid?), but compared to a dry starter, there are significant advantages the the starter/yeast combo:
1) Feeding the wet starter takes 10s: pour flour and water onto leftovers and stir quickly with a spoon. No sticky stuff to deal with.
2) The starter seems exceptionally stable, maybe because of the water layer: I only wash my starter jar every two or three months, and the 10g or so that I put back in the fridge after starting a dough, will last for weeks and consistently restart overnight when fed
3) Being able to independently adjust yeast-levels in a predictable way, means that I can easily play with sourness levels and adjust leaving times when I have to match the timings with other activities.
There is the downside, of course, that I need to keep bakers yeast in the house as well...
Also, nobody in the field disagrees that in the more distributed grid we are seeing today, more endpoint communication and control could lead to more resilience. Whether pricing signals are the best path is a more open question, but they certainly appear to be a feasible option.
Their approach is PV + DCC (Direct Carbon Capture) and then simple carbohydrate synthesis, with the goal of establishing standalone autonomous systems that can generate valuable resources on their own in remote areas with ample sunlight.
They have a great blog where they go through their motivation for the approach from first principles [1].
[0]: https://terraformindustries.com/ [1]: https://terraformindustries.wordpress.com/home/
[0]: https://github.com/wcshds/typst-math-to-mathml-converter