HN user
Anecdotally, lots of intelligent people I know like to pretend to be completely uninfluenced by advertising, which I think is delusional.
In my view advertising is in big parts (your category 3) basically a constant side-channel attack against peoples minds, eroding decisionmaking and purchasing decisions (mostly!) to peoples detriment.
The net-beneficial aspects (product information, and, arguably, boosted general demand) could be better achieved otherwise.
Is this excluding maritime shipping?
According to eurostat (https://ec.europa.eu/eurostat/statistics-explained/SEPDF/cac...) that is two thirds of freight volume and it probably still eclipses even rail transport in CO2 efficiency (until full CO2 free rail electrification), so this probably distorts things completely...
If you actually compare the emission fractions, you can see that the US uses much more CO2 per capita on transport compared to Europe (about 3 times as much!):
https://ourworldindata.org/grapher/per-capita-ghg-sector?cou...
How much of that "decoupling" is actually just "exporting"?
Some fraction certainly is. Here is trade-adjusted CO2 per capita: https://ourworldindata.org/grapher/consumption-co2-per-capit...
Note how this still trends down for most western industrialized countries despite positive economic growth over the last two decades.
The reality is obviously more complex, and many countries have already managed to somewhat decouple economic growth from CO2 emissions (very good!), but as a crude approximation: More money => everyone gets more stuff => more CO2 emissions (producing and running that stuff).
If you want meaningful CO2/capita comparisons, you also have to be very careful with countries that get ("free") hydro power opportunities for electricity because that distorts the picture massively (same for e.g. Norway).
Big producers of hydrocarbons, on the other hand (like the US) have to work hard to resist the allure of cheap & convenient fossils.
Switzerland is so far ahead in this comparison because they get a lot of CO2-free hydroelectricity (>50%), have to import most hydrocarbons (=> incentive against) and also save massively on transportation because density is much higher.
Agree on a lot of this but not AI; it's already aged like milk.
In my view, you don't get to dismiss LLMs as "stochastic parrots" unless you can argue with precisely defined terms why you ain't one.
Outsourced manufacturing is much less of a factor than people typically assume:
https://ourworldindata.org/grapher/imported-or-exported-co-e...
US numbers are insanely high because cheap hydrocarbons are locally available (=> bad incentives) and everyone is wealthy (that correlation is very strong; just compare Luxembourg, which is much wealthier and more polluting than surrounding nations) and also population density is rather low so more energy wasted for transport.
Sure, but if you want dynamic loading from your c stdlib (which is defensible IMO), and you want the behavior/implementation to match with the loader, then you need some kind of coupling somewhere no?
You could have a very slim libdlopen that is used by both loader and libc, but I don't really see how that's any improvement/much different.
Yes. Would currently be "free" (<1c/kWh) around 1400-1530 tomorrow, and go up to ~15c/kWh at night. Fixed costs might distort everything significantly though.
Quick numbers that I found suggest lifetime healthcare costs of >$300k even if you're 65 already (with current yearly spending trending more towards $600k-ish).
So even with a (single) catastrophic event just paying yourself might save you money still (which seems insane to me).
Isn't that kinda to be expected if you want to provide dynamic loading functionality (dlopen)?
Is the windows situation really all that different/better (with GetProcAddress in kernel32.dll)?
What would you want to hear from a non-American biased model in response?
Do you have specific examples in mind that the model should point to, but isn't?
If you circularize the orbit at 550 AU (which is costly but possible, at ~1km/s of delta-v), then you end up with an orbital period of >10000 years (to complete your whole-sky survey in one plane).
But first-- the price to get that circular orbit is steep, you end up imaging empty space for a lot of the time (no real way to "skip ahead") and your orbital speed vector makes you leave the observation cone for any interesting target pretty quickly.
In conclusion: You probably really want to stick with "one telescope probe per target system", and to give them one-shot orbits because those are cheap, reasonably fast and inside the observation cone (for your single target) for longer.
First: Orbital period out there is over 10000 years.
And if you circularize (which is expensive to do in delta-v), you minimize the time window you have for observation (because you're basically pointing your speed vector straight to outside of your observation cone).
This is indeed an interesting perspective, but "constant 1g rocket acceleration" is not even an engineering pipedream, it's strictly fantasy territory.
Adding to this:
Those 190km/s of the Parker solar probe were, crucially, periapsis speed.
This is a bit like bouncing a rubber ball from a building, measuring its speed at ground level and then going: "Given our fastest achieved speed, we expect to hit the cloud level in <10s".
~200km/s sustained speed is already insanely optimistic for anything we could realistically build in the next half century, so your position is even more ironclad than it looks at first glance.
There is no "building" such a thing. All we could do right now is send the "telescope probe" >500AU away, on the opposite side of the sun from the observation target, then hope it still works 80 years later or so when it gets there.
Edit: My point is that you can't "build" such a thing and later point it somewhere-- you have to fly the camera part of the "telescope" about 3 times as far as voyager 1 went, exactly opposite of your observation target, and it is not gonna stay there for too long either.
As long as we improve rapidly at both drone-building and exoplanet target selection, it is not really gonna be worthwhile because both the drone hardware and the target will be hopelessly obsolete before we even get halfway to the observation point.
I think its fair to argue that modern silicon easily makes up the difference with clock speed; neurons can't do Ghz things by design (signal propagation in biological brains is also really slow compared to copper).
I'm pretty confident that algorithm/hardware design insights over the next decades will allow us to build human-rivaling cognitive abilities on basically todays consumer tech (analogously to how computer chess improved over time).
Some aspects of the conjecture make sense and are observable:
Consider e.g. Steam (digital video games): Prices are discounted over time because of "greed" (=> desire to sell the same product to customers that value it less than the first wave).
Customers do adapt to this, and expect future discounts (sales) at release date already, and defer their purchase accordingly (despite valueing it higher!).
But in reality, customers are not 100% rational, don't have perfect information (on seller strategy), and the product value (to buyers) changes over time too (typically mostly downward), so the base assumptions are difficult to find in reality.
Not really.
The main "purpose" for such information is to be published, and at that point it makes no difference who exactly requested it.
Personal view:
The current mainly governing party in Germany (CDU/CSU) is a bunch of incompetent, nepotistic gerontocrats, and this change is mainly intended to make it harder for independent press to air their dirty laundry.
Just for reference: Approval rating for Merz (chancellor) is under 20% (!!); even Trump is >35%.
Unlikely.
This is him, at least according to that account:
AI psychosis.
No need to be so dramatic, might just be a bit of an early midlife crisis thing.
I liked the post, some interesting takes.
I concur. It's interesting that this seems so difficult to get rid of; despite actively trying to overlook the taste of LLM writing, it still grates, and this does not happen to me in the same way even with communication from half-illiterate human idiots.
It's kinda fascinating how far LLM coding/problem solving has come without making much progress on the annoyingness of their writing. Any theories?
The regulation crucially results in your recruits not dying for no good reason during training just because some random piece of trash equipment predictably failed.
Preventing that is much more important than the exact dollar efficiency of said equipment during peacetime.
Yeah, but you can always buy a slightly smaller bird, strap the warhead directly to it, and fly that into your target 100+ miles away.
Guided missiles specifically are insanely pricey by comparison to the warhead alone; just for the possibility that your slow, vulnerable drone might be able to return (and be used again) you have to make very expensive engineering tradeoffs, and even when the thing comes back you have to repair and service it, too (and stock it up with more expensive missiles).
First: microwaves are only safe if you don't mess with them, a bunch of people get killed by tinkering with their high-voltage transformers (at least 35 US deaths from one specific usecase alone in the last decade, see https://www.woodturner.org/Woodturner/Resources/Safety-Mater...).
A big potential concern with small reactors are highly toxic radionuclides; those can be much more dangerous (with LD50 far under 1mg/kg) than "ordinary toxins" like bleach or even nasty stuff like methyl isocyanate. That means expensive disposal and protection measures.
All of this is a non-concern though because there is no realistic path for nuclear reactors to compete with PV+batteries, ever. With cells already <$100/kWh and the panels being cheaper than glass windows, we will never be able to build, maintain and dispose of nuclear based reactors tech at a competitive price point, especially not with the insane current battery demand (automotive) driving technical optimisation and price competition.
The modern (UA) take is to strap the warhead directly on the thing, make it just as big as needed and directly fly everything into the target.
If you insist on firing guided missiles at ground targets from a drone that returns to base you're never gonna be able to compete on cost.
If you onshore production via tariffs while having low unemployment, you are basically forcing your population to build basic stuff despite them having better things to do.
This is somewhat justifiable with vital sectors like agriculture, but if you do this in an arbitrary way just for the sake of it you just make stuff more expensive and your workforce less productive for no gain.
Those dollars are not just vanishing abroad, you are getting actual stuff for them, and your citizens then don't have to spend their own time building it and can do something more productive instead.
It's gramatically correct, spend is a noun here (but you expected a verb).
"won" is a very strong word for a country that sits economically somewhere between South Sudan and Congo, with a primary "enemy" that started out similarly and is now easily 20 times richer.
Self-sufficiency and exclusion from global trade can be very expensive for a nation long-term (a cautionary tale, since those ideals are a bit of a siren call nowadays to many Americans, after getting quite wealthy by doing the exact opposite).
Just like a medieval barber surgeon.