So I know you're making a joke/statement about how the post-9/11 volunteers got royally screwed over and are (understandably) disillusioned with war but I just wanted to add something here.
This change increases maximum enlistment age. Maximum reenlistment age is something else entirely. To reenlist, you need to be able to complete 20 years of service by age 62. So if you joined at 18 and did 8 years then you can technically rennlist up to age 50. Not that you would or should but you can.
With all the posts lately about Karpathy's autoresearch, it remains unclear to me whether this name is intended to convey that this LLM-codebase should be useful for research across all domains - like molecular biology, aircraft control, sociological, ww2 history, etc. or is it intended only to discover new LLM capabilities.
Solar panels and the parts to build a hydro generator, then. A hydro generator would also be good incentive to plan around a reliable water source, without which all bets are off anyway.
Some part of my brain filtered out nuclear winter, which cannot reasonably be prepped for by individuals or small groups. However, that is just one, relatively unlikely, thing to prep for. Most other disasters are shorter-lived, and have a great deal of overlap in effective mitigation strategies. Prepping, in my mind, is not only practically useful for various classes of emergencies, but is good mental exercise for understanding supply chains and what's actually needed in the sort, medium, and long term. It can also be good for sharpening skills that benefit others and build community, which in many ways is more rewarding than knowing that you'll be the sole survivor. Prepping doesn't, and shouldn't, look like Burt from Tremors (as amusing as that may be).
I'm being a bit glib anyway; call it gallows humor to help me process currents events. Even worldwide, long-lasting nuclear winter must passes & settle eventually, and such sunlight-enabled microfiche files could be useful to subsequent generations if not earlier.
Fellow survivor on a bicycle which powers a lightbulb by which you can read the microfiche. Voila.
For anyone who never watched the original Soylent Green movie, it's worth a rewatch because it actually shows a future where people are having to make do without a power grid in cities, by doing things like riding a stationary bike hooked up to a generator to power their TV or radio long enough to get some news.
I mean, there's got to be a spectrum of nuclear winters, just like there's a spectrum of volcanic and impact winters. Assuming a full scale nuclear war, there's still questions of how much of the arsenals actually detonate(as opposed to failures to launch, getting destroyed by other nukes first, etc), how much the fires burn, the time of year impacting fires and dust and the state of the biosphere, and how much the aerosols are limited to the northern vs southern hemisphere.
Think about all the ways you could die from nuclear war + winter. There's some worth avoiding(slow painful death from radiation, moderate burns, trapped in collapsed buildings, etc), and others that you might be willing to delay/prevent in a hope of things getting better(starvation, cancer, civil unrest, etc). There are ways you can reasonably prepare to increases your chances, if you're lucky in the critical moment of nukes dropping, to survive long enough to attempt forming communities and farming again.
Stationary exercise bike, large hobby BLDC motor (or random PMAC motor from some AC appliance) plus some diodes (fullbridgerectifier meme goes here) to rectify the generated voltage. :)
Since I'm yet to seriously dive into vibe coding or AI-assisted coding, does the IDE experience offer tracking a tally of the context size? (So you know when you're getting close or entering the "dumb zone")?
The 2 I know, Cursor and Claude Code, will give you a percentage used for the context window. So if you know the size of the window, you can deduce the number of tokens used.
Haven't opened the article yet, but surprised the comments so far are generally in the framework of "oh, economy produces more X, but too much more of a proportion of X is going to [the fortunate few] and too little to [the unfortunate many]." where is X is some kind of fungible consumable. Rather what I see are asset holders and liabilities holders (same spectrum, some enjoy the positive side, some struggle on the negative side). Goods (the consumable, fungible sort) flow in, around, between, and all throughout them. But the only ledger that matters, the one that makes some stressed out and others feel empowered & satisfied, is the asset-liability spectrum.
Update:
And now I've read at the article. Decent, it might sa well be the GPT of "Update Das Kapital for the 21st century". (GPT here being a figure speech, i.e. irrespective of whether an LLM helped in composing the piece). Article still fixates too much on differential parceling out the flow of economic product, and not the asset-liability ledger which everyone is jostling around with each other on. (It almost touches on it in "Mechanism #3", but not quite).
Use flameview: https://news.ycombinator.com/item?id=25713858
Life-changing for HN users.
reply