Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
nwiswell
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
24 ms
·
271.
▲
by
nwiswell
4y ago
> then the rich will hire skilled fighters to go be legislator-duelists for them I'm going to be honest: considering the smorgasbord of possible dystopias, this one sounds pretty rad. Can we do cyberpunk corpocratic gladiatorial par
272.
▲
by
nwiswell
4y ago
Definitely, but it doesn't solve the reliability problem
273.
▲
by
nwiswell
4y ago
> I don’t see a fundamental reason why one couldn’t replace one ASML EUV machine with, say, 1000 e-beam machines and run them all in parallel. Fab floorspace is also very expensive, nevermind that's not even close to a realistic pri
274.
▲
by
nwiswell
4y ago
I don't think being an ML chip means the defects are necessarily less fatal. These often interfere with the actual functioning of the chip, cause shorts, etc -- it's not just a matter of the TTL being very slightly messed up somew
275.
▲
by
nwiswell
4y ago
> Also, while identity theft sucks, getting your password stolen is painless, whereas someone chopping off your thumbs to squish them up to an ATM machine is a less pleasant experience. This is not a real distinction if they are physical
276.
▲
by
nwiswell
4y ago
Sure. Chief among those dimensions is the fact that it's not used as a serious production technology, so the manufacturing of these systems doesn't benefit from economy of scale. E-beam certainly does provide a bounding limit on h
277.
▲
by
nwiswell
4y ago
The machines themselves couldn't be "super" cheap, that's impossible. You still have to deposit the e-beam resist while keeping the wafers extremely clean. This is non-trivial. The only route to economic viability is abs
278.
▲
by
nwiswell
4y ago
https://hbr.org/2014/04/pikettys-capital-in-a-lot-less-than-...
279.
▲
by
nwiswell
4y ago
You should probably read some Piketty. There is absolutely nothing which structurally ensures that the wealth created by capitalist pursuits will be distributed sufficiently to ensure that standards of living will increase for everyone, and
280.
▲
by
nwiswell
4y ago
You're still confused. The EMH is understood to imply that securities prices are the NPV of future cash flows (or more specifically their best estimate, using all public information), but that has nothing to do with ROIC. (As an aside,
281.
▲
by
nwiswell
4y ago
> If you believe that it's costless or near-costless to hedge your inputs, you should do it literally everywhere to everything that you possibly can ... no? Just because you've eliminated price risk doesn't mean you'v
282.
▲
by
nwiswell
4y ago
> A small amount of seaweed added to a cow's diet massively reduces the amount of methane they sell. Beggar thy heifer!
283.
▲
by
nwiswell
4y ago
A lot of businesses actually do hedge their inputs, this is not a crazy idea I've just come up with. For some industries it's de rigeur (jewelers typically hedge even the value of their inventory and WIP), in others it seems to
284.
▲
by
nwiswell
4y ago
I think in general the transaction costs associated with hedging wheat price risk are so small via futures (0.05% or less) that it proves the larger point that profits cannot possibly be due principally to assumption of risk.
285.
▲
by
nwiswell
4y ago
> If you actually believe that a residential REIT is zero-risk and a casino is high-risk but they have the same ROIC, you should buy REITs and short casinos and you can make a big profit. That illustrates why there's probably a flaw
286.
▲
by
nwiswell
4y ago
That's because the answer is 'no', unless there's an underlying reason to expect that the expectation price of wheat will be less in the future. The only impact to profits is the actual transaction cost of the wheat fu
287.
▲
by
nwiswell
4y ago
That's kind of absurd as a construction. It may be literally true that a bread factory accepts various kinds of risk in the ordinary course of business, but it's plainly not the primary reason that a bread factory turns a profit.
288.
▲
by
nwiswell
4y ago
> On top of that, the pollution is dramatically reduced. A modern natural gas turbine at a power plant is dramatically cleaner burning than an open flame in a kitchen stove. Not to mention the pollution from the power plant is not going
289.
▲
by
nwiswell
4y ago
> The rules would essentially require all new residential construction to have electric water heaters by 2027 and heat pumps instead of furnaces by 2029. By the time this is fully in effect for a significant number of people (i.e., basic
290.
▲
by
nwiswell
4y ago
It's also trivially defeated by the kind of financial centralization that would necessarily follow in this kind of mass-adoption scenario: the payment processor (BitPay, Visa, MasterCard, whatever) can have a pool of Bitcoin transactio
291.
▲
by
nwiswell
4y ago
> Currently, these technologies are approaching parity on cost fairly quickly. It's basically at the prototype stage presently, so much more expensive, but widespread adoption and economies of scale could change that within several
292.
▲
by
nwiswell
4y ago
That's very interesting, but is it cost effective to get carbon out of the atmosphere rather than to drill it out of the ground? Put another way, you can either build 10 wind turbines with atmospheric carbon fiber, or 11 turbines with
293.
▲
by
nwiswell
4y ago
Not that unreasonable a guess. It's very late in Europe right now (currently 4 AM GMT, perhaps 2 AM when that was posted), and you know YC is an American company... right? https://news.ycombinator.com/item?id=30210378
294.
▲
by
nwiswell
4y ago
Alright, but can I run Doom on a thermostatic shower valve?
295.
▲
by
nwiswell
4y ago
For an airship of the size proposed in that post, I don't think I agree. Presumably there would be some measure of compartmentalization, ventilation, and fire suppression so that even if a portion of the craft was stricken, the entire
296.
▲
by
nwiswell
4y ago
> From that link it sounds like there are alloys that maybe preserve the desirable properties of the metal but aren't as flammable as elemental magnesium. Anybody know more about that? Why it works? How well? It's been a long t
297.
▲
by
nwiswell
4y ago
> Again, you have to manufacture and assemble all the components for the machine to work. And, yet again: that is not the hard part. The reason why ASML is the only company which sells EUV tools is because it is the only company which sp
298.
▲
by
nwiswell
4y ago
Has it occurred to you that I might work in this industry? The other guy had a six month stint in manufacturing two decades ago.
299.
▲
by
nwiswell
4y ago
In fact it is quite easy! This is a solved problem, the stage is moved those minute distances using piezoelectric actuators. Moving hundreds of microns accurately is actually a harder problem, since that requires traditional lead-screw styl
300.
▲
by
nwiswell
4y ago
That is a development system. The purchase prices of the actual manufacturing systems are never disclosed, since the negotiated price differs between customers. In any event the HVM systems will cost significantly less. This similar article
More ›