1,685 karma · joined September 17, 2015
(Though I used to use the em dash with spaces on either side — like this — and only recently converted to the more common, tighter spacing—like this. I might go back to my old ways since it's different from how AI uses them, like some sort of weird shibboleth.)
Is this not a major problem with YC, specifically? Our beloved orange site funded and accelerated these guys.
https://www.youtube.com/watch?v=EsUBRd1O2dU
TL;DR... cooling in space isn't passive, you're on the "inside" of an enormous vacuum flask. And radiative coupling with space is possible from the ground, if that's what you're interested in:
https://www.skycoolsystems.com
But god bless crazy entrepreneurs. Don't ask whether we can, definitely don't ask if we should, just ask whether it makes for good headlines...
I did, and I have previously read an article in (iirc) The Atlantic about the screwfly program when it was working. It was a monumental effort to push screwflies back to the Darien Gap, involving widespread coordination of cattle ranchers and government workers, constant flyovers of planes dropping huge numbers of sterile flies, and a massive breeding program.
Two university students in a van can get about as much done as it sounds like. Some problems require coordination, government involvement, and yes, the organization — i.e., bureaucracy — that implies.
(Edit: it was The Atlantic, not The New Yorker. Comment further down has the link, but here it is: https://www.theatlantic.com/science/archive/2020/05/flesh-ea...)
This is one of the few carbon capture approaches that appears to be able to approach global scale, so I'm rooting for it. Even enhanced weathering suffers from needing to move billions of tons of rock, but scattering trace minerals seems pretty high leverage. The sheer mass of material that must be removed from the atmosphere is otherwise very intimidating.
Here in the US, you can buy publicly traded companies in the US for a song; in fact, that was the source for the "acquiring" company in a lot of those reverse-acquisitions of the late 2000s / early 2010s: publicly traded companies that had a ticker, but were basically defunct. A company that wanted to go public without the scrutiny of an IPO process could "buy themselves" with the tiny, publicly traded company, thereby getting a ticker and access to retail investors. (This was not a good phenomenon, fwiw, though it definitely anticipated the SPAC craze... which was also not good, for similar reasons.)
The best-positioned lawsuits to win, like NYTimes vs. OpenAI/MS, is actually based on violating terms of use, rather than infringing at training time.
Emitting works that violate copyright is certainly possible, but you could argue that the additional entropy required to pass into the model (the text prompt, or the random seed in a diffusion model) is necessary for the infringement. Regardless, the current law would suggest that the infringing action happens at inference time, not training.
I'm not making a claim that the copyright should work that way, merely that it does today.
For fun, look up how Brazil tamed inflation by creating a fake currency it never actually issued. (Taking off so no time to look it up myself.)
“Milei has implemented these reforms legally and constitutionally...”
Argentina is what Trump’s team imagines the US to be. They’re not the same.
EDIT: fixed some typos.
Argentina has a nearly century-long history of economic volatility alternated with stagnation, massive inflation crises, a spell of military dictatorship (after a flirtation with fascism)…
Wait a second…
(In all seriousness, Argentina is truly unique. As economists say, there are only four kinds of economies in the world: developed, developing, Japan, and Argentina.)
I’m not making this up, camera manufacturers have told me to my face that focus is open loop, period. They can’t guarantee repeatable focus.
Notably the measurement isn’t of the state of the motor/gearing. Furthermore, being “in focus” means the subject matter’s out of focus blur is below some threshold; there is a range of focus states that qualifies — but those seemingly small differences can affect camera calibration, with >pixel-level differences in effective focal lengths.
It’s open loop.
This is why camera calibration can be tricky, and you often don't want to touch the cameras after you've done a calibration pass.
https://www.bitsaboutmoney.com
And subscribe to Bloomberg's Matt Levine's "Money Stuff" newsletter and podcast:
I trust a Nobel Prize winner has thought through it more deeply and integrated the causes and effects more than most random internet commentators. I am happy to report that “low interest rates” do not cause inflation by themselves on any scale less than a decade. Source: 1995-2021.
What actually causes inflation is more a function of overall demand and supply in an economy. Interest rates have an indirect effect on that balance that can also have higher or lower leverage depending on a lot of other factors.
Or are you referring to the fact that he said it didn’t show up as having caused growth in the underlying economic data? I think he was pointing out something quantitative at the time, and in recent years he speculated that the strong growth in US GDP was finally reflecting efficiency and productivity gains from IT and the Internet. (The US’s strong growth until very very recently was an outgrowth of strong and rare increases in worker productivity.)
Krugman actually generally felt it went too far, and was not a proponent of it.
There is some similarity to his making fun of people worried about “bond vigilantes,” iirc. But I don’t think your characterization of his positions is accurate.
Conflating that with Keynesian principles is false. He didn’t propose that the trillion dollar coin would ever affect actual spending.
You don’t need a conspiracy, you just need a philosophy of small government and low regulation to let grifts expand through the financial parts of the economy. The history of financial regulation in the US is the banking industry finding new ways to evade regulation, which lead to moral hazards and privatized gains / socialized losses.
But it’s not that the “system” is explicitly designed that way; instead, it’s an adversarial system and the US in particular (for cultural reasons) has long favored the vigilante over the regulator.
The story works even without simple Hollywood villains. The system isn’t malicious, it’s just stupid.
Which is a long way of mostly agreeing with you.
(Oh, except on printing money. That’s a surprisingly deep subject that’s often misunderstood. You’re spot on about government-sponsored wealth redistribution, though. Fully on board with that.)
And I'm not even an economics expert. I've just read more about it, or read more critically, than geohot. And I can tell you he's wrong, not because he's an outsider but because his analysis does not make sense.
Copernicus had well-reasoned, self-consistent arguments that incorporated previous models, and produced better results. This doesn't. There is a rubric for how one should score outsider contributions, even if you want to allow for revolutions in a particular field. This essay doesn't rise to that level.
In fact, it matches the first pattern -- smart person outside their field, makes ass of self -- much better.
China has a serious debt problem in its economy. Not government debt like in the US, but shadow banking debt. My fear, based on reading Pettis primarily, is that the Chinese people will end up paying for that debt. Households that have expensive second homes as investment vehicles will suffer those losses (since there isn't sufficient population to fill those homes), unless those losses are socialized or redistributed.
I like Michael Pettis's analysis, though I'll grant you that the distortions he cites in the Chinese economy have gone on longer than he predicted possible. But the health of the Chinese economy is not just the visible things you cite, but also the distribution of assets and balance sheets in the economy.
Per Pettis (and other sources), the banking system is dealing with a large debt overhang in society. Pettis's proposed policy fix is to assign that debt to the state, to state-owned-enterprises, or to the corporate sector. Because the long-term growth of China's economy depends on being less reliant on exports, and having a more "normal" mix of domestic demand. There are nuances there, but Pettis does a better job than I do in explaining it.
The trouble is that it's not politically expedient to assign the losses due to (unrealized) bad debt to those entities. And the result is a slow, grinding failure in the housing sector. The government does have a tight lid on dissent, but I understand that the Chinese people are in a more precarious position than they should be due to many households having overinvested in nonproductive real estate projects. How that unwinds will be fascinating to watch, and extremely relevant to world economics and geopolitics.
To be clear, I really am hoping for the best for China and its people. The right policy is sometimes well-known but hard to execute, and that's the tragedy. (It's playing out now, in spectacular style, in the US.)
Ironically, that has created a populist backlash that has vaulted even worse economic ideas into power.
Anyone read "Capitalism in the 21st Century"? It's long and dense, but it does make some excellent, well-supported points about the structure of an economy that explains the current situation very well with historical antecedents.
This won't do it justice, but: inequality always increases, by definition, if the returns on capital outstrip the overall growth in the economy. (Think through that if it's not obvious; it's literally a mathematical truth.) The book argues that history is a series of long, slow accretions of wealth, with increasing inequality, until some shock (usually a violent revolution) overthrows the old system, resets inequality, and the cycle starts again. The latter half of the 20th century, for the US, was one notable exception: growth outstripped returns to capital, so the benefits of that growth were widely shared: inequality decreased. That trend ended in the ~1980s, and the widening inequality gap has fueled unrest.
And the book points out that we've seen this show before. I think we're in the "history repeating itself as farce" phase, though.
I mean, we've sort of long assumed that increasing trade ties led to peace, and the detente between the US and China was long taken as proof of that axiom. But morons with half-formed ideas about trade policy can blow up detente in less than 100 days, as it turns out.