Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

So much of academia is about connections and reputation laundering

statmodeling.stat.columbia.edu

Post reply on HN
Print view View original

161–170 of 223 posts

Re: So much of academia is about connections and reputation laundering

#161

Post by akiselev »

giantDinosaur wrote: post #145
hn_throwaway_99 wrote: post #142

Earlier quoted context omitted.

I don't buy your arguments - at all. At the end of the day people have strong expectation about what software should do. If it doesn't do that, they will quickly notice. If they don't notice (now or in the future) things like your "off by one" errors, then that requirement wasn't likely that important to begin with. Now, I've seen plenty of "demo-ware" software that has been presented as something that was much more…

Perhaps we have different conceptions of 'fraudulent'. I don't think it necessarily implies 'fake', but something more along the lines of using inferior building material. Perhaps it 'wasn't important all along' or perhaps it'll lead to catastrophic failure in 25 years, who knows? Software 'engineering' may be hard to fake, but that's because you're probably defining it as the proper way to develop software. A lot of…

Building materials are judge based on the standard of "survives for at least three decades without crushing those inside" not "survives the next reorg or pivot." What would be the equivalent of inferior building materials failing over 25 year timescales in software? C isn't even twice that age and the oldest running mainframe software is probably only a little older whereas we've had millenia of experience constructing buildings.

Re: So much of academia is about connections and reputation laundering

#162

Post by em500 »

gowld wrote: post #123
CrazyStat wrote: post #115

Earlier quoted context omitted.

> By "cubic fit" I infer that they mean some kind of low-pass signal filter. It's literally just ax^3 + bx^2 + cx + d, optimize a, b, c, and d to minimize some loss function (probably L2).

And since cubics can't be nearly flat over two disjoint intervals separated by an extrenum, they added an unlabelled nonsensical bell curve continuation to the to the end where a cubic would predict the case rate plummeting to negative infinity

Not really, they just fitted the cubic to log-deaths and exponentiated them again.

Re: So much of academia is about connections and reputation laundering

#163

Post by roenxi »

atq2119 wrote: post #68
dragonwriter wrote: post #57

Earlier quoted context omitted.

> MMT especially is not at all a mainstream theory. AFAICT, the descriptive aspects of MMT are widely accepted, if deemphasized in prescriptive contexts, aspects of mainstream economic theory (not just [neo-]Keynesian, but across essentially the whole spectrum of descriptive economics.) The prescriptions that MMT adherents make based on those descriptive aspects are out of line with mainstream prescriptions, which te…

Is something descriptive really accepted if it is deemphasized in prescriptive contexts? Take the very basic thing that you mention at the end, which should be absolutely non-controversial: the US government cannot be forced into default. If your prescriptions are just going to ignore that fact, then have you truly accepted it? I'd argue that no, you really haven't. (That doesn't mean you'd have to follow the prescri…

If I lend the US government enough money to buy a sandwich, and get back only enough money to buy a half-sandwich it really doesn't matter to me whether they technically defaulted or not. I am down half a sandwich.

The US government 'can't default' but that is basically word games for a complicated tax where nobody is quite certain who is paying. The government is definitely consuming real resources and unlike a tax it isn't at all obvious who would have gotten those resources had the government not redirected them. I'd rather governments were straightforward and levied taxes to pay for things so we know who is supporting state spending.

The conversation really hinges on the semantics of 'default' - in real terms the Government absolutely can default. At some point the country has collapsed, there is nothing left to give (see classic hyperinflation cases) and the government will not make good on its debts. In nominal terms the government can't default but anyone who treats that as useful in their decision making is going to lose out sooner or later when it comes back to real goods and services. I don't want to be one of those people, and I don't want there to be any people like that because it seems dishonest at some level to pretend they aren't losing out.

People call US bonds 'risk free' - that is only in nominal terms. In real terms they are actually quite risky. Take on a 30 year treasury bond today and there isn't any certainty how much it will be worth in 2020 dollars as it matures. Is it a likely net win on the sandwich scale? Signs point to no, but it might be. There are risks.

Re: So much of academia is about connections and reputation laundering

#164

Post by akiselev »

mettamage wrote: post #159
o10449366 wrote: post #154

Earlier quoted context omitted.

FWIW, this wasn't my experience at all, but also there may be differences between the US and Europe. I went to a unremarkable top 50 state school in the US, graduated with a 3.0 GPA BS in CS, had a single internship at a no-name local company, no research experience, no referral, no significant personal projects, no TAing experience, and I got interviews at Amazon, Facebook, and Google by cold applying online. It was…

The reality hurts so much, and it's upsetting to me I can't be more grateful (some people that I know are homeless and hungry, relatively rare in The Netherlands but it's there). I should've done my own research when I was young, but I was young. I didn't know that going to the US would be so tough, no one tells you. 18 months of not even being looked at in most cases, especially US companies (some Dutch companies di…

That almost certainly has nothing to do with you - hiring a fresh graduate through the H1B program is practically impossible in my experience. H1B applicants have to show "significant work experience" in a relevant field and unless you have a PhD (and often times even then), any work experience gained during your education will not count towards that requirement. Most companies won't even bother with the interview; 18 months sounds like the bare minimum to even have a chance at meeting the requirements.

Re: So much of academia is about connections and reputation laundering

#165

Post by mettamage »

mettamage wrote: post #159
o10449366 wrote: post #154

Earlier quoted context omitted.

FWIW, this wasn't my experience at all, but also there may be differences between the US and Europe. I went to a unremarkable top 50 state school in the US, graduated with a 3.0 GPA BS in CS, had a single internship at a no-name local company, no research experience, no referral, no significant personal projects, no TAing experience, and I got interviews at Amazon, Facebook, and Google by cold applying online. It was…

The reality hurts so much, and it's upsetting to me I can't be more grateful (some people that I know are homeless and hungry, relatively rare in The Netherlands but it's there). I should've done my own research when I was young, but I was young. I didn't know that going to the US would be so tough, no one tells you. 18 months of not even being looked at in most cases, especially US companies (some Dutch companies di…

@akiselev: I haven't only been applying to US locations of US companies (e.g. Google Switzerland). But a SWE from Google gave me a referral link to which I'm grateful. I wonder to what extent it will help though since it went through HN and we just met. I want a phone screen and be told that I am or am not good enough. I can have peace with this if one FAANG company shoots me down on my D&A skills. For my own sanity, I've resigned to the idea that I'm not going to get a phone screen as I have little faith in the Google referral to help me past the resume check, but we'll see.

Re: So much of academia is about connections and reputation laundering

#166

Post by tprice7 »

Gatsky wrote: post #148
tqi wrote: post #112

Earlier quoted context omitted.

Definitely agree, this whole affair has a real rashamon vibe to it. it feels like everyone involved (the article, philipson, and furman) are talking past each other. As far as I can tell, everyone agrees forecasting and curve fitting are different things, and that a cubic fit (red line, allegedly) would be a bad representation of what the future will hold. The disagreement seems to be: - furman believes (or claims to…

https://en.wikipedia.org/wiki/Rashomon "Rashomon (羅生門, Rashōmon) is a 1950 Jidaigeki psychological thriller/crime film directed by Akira Kurosawa... The film is known for a plot device that involves various characters providing subjective, alternative, self-serving, and contradictory versions of the same incident." Now that is an esoteric reference!

It's one of the most famous Japanese films, top ten perhaps.

Re: So much of academia is about connections and reputation laundering

#168

Post by atq2119 »

roenxi wrote: post #163
atq2119 wrote: post #68

Earlier quoted context omitted.

Is something descriptive really accepted if it is deemphasized in prescriptive contexts? Take the very basic thing that you mention at the end, which should be absolutely non-controversial: the US government cannot be forced into default. If your prescriptions are just going to ignore that fact, then have you truly accepted it? I'd argue that no, you really haven't. (That doesn't mean you'd have to follow the prescri…

If I lend the US government enough money to buy a sandwich, and get back only enough money to buy a half-sandwich it really doesn't matter to me whether they technically defaulted or not. I am down half a sandwich. The US government 'can't default' but that is basically word games for a complicated tax where nobody is quite certain who is paying. The government is definitely consuming real resources and unlike a tax…

You are right. But note that in order to contradict me on the surface, you are in fact agreeing with my point by starting to shift the conversation away from nominal terms like the deficit and towards real resources -- which is the correct framing!

The next steps are to recognize and integrate into the conversation that:

* Whether and to what extent the government consumes resources that would otherwise have been consumed by somebody else cannot be determined purely by looking at the deficit as a single number. A government deficit, when done right, stimulates the economy which means that it causes the creation of resources that otherwise simply wouldn't have been created.

* Inflation, which is what you're really getting at, is complicated and has many potential drivers. A government deficit can be one of them, but isn't necessarily. There are many other potential drivers: overly lax monetary policy, excessive bank lending, entirely internal mechanism such as genuine supply shocks, businesses' price hikes in an attempt to increase profits, strong unions helping drive up wages broadly across the economy, and so on. Note that some of these effects, especially the last one, are actually desirable for most people, meaning that inflation can actually be a good thing for society overall! Admittedly that happens rarely in practice, but that's really a function of workers having too little power. It all depends on the details.

In fact, on that last point there's reason to suspect that we'd have had quite a bit lower "effective" inflation (meaning higher purchasing power of wages / salary) for the majority of the population today if governments had decided to address the Global Financial Crisis by direct job creation and handing out money to the population at large, rather than leaning on monetary policy which really only caused asset prices to balloon.

This alternative policy wasn't even discussed seriously, because people largely do not understand that the government cannot default. Discussion was shut down with slogans like "making sure that the US is not going to be the next Greece", which are complete nonsense given that the US and Greek governments operate under very different currency arrangements (sovereign currency like Japan, vs. the effectively foreign currency of the Euro in Greece's case).

So anyway, the point is that the framing in real resources matters significantly, precisely because there is no ironclad correlation between government deficits and real resources. People need to learn to go into the real resources framing and then stay there.

Re: So much of academia is about connections and reputation laundering

#169

Post by _8j50 »

kbos87 wrote: post #40

So much of everything is about connections and reputation. Tech in particular holds up a guise that people are chosen and ascend because they are “good”, meaning, they are good at what they do. That’s part of it, but it’s an insufficient part, and you can do pretty well in a career being solidly average at what you do. The real currency in my experience is one thing - relationships. Do I think this person likes, resp…

Mixed experience here. Technical breadth has helped me a lot,but relationships helped me more. They're not mutually exclusive. If I had barely enough technical skills though,no relationship can save me.

Re: So much of academia is about connections and reputation laundering

#170

Post by akiselev »

mettamage wrote: post #165
mettamage wrote: post #159

Earlier quoted context omitted.

The reality hurts so much, and it's upsetting to me I can't be more grateful (some people that I know are homeless and hungry, relatively rare in The Netherlands but it's there). I should've done my own research when I was young, but I was young. I didn't know that going to the US would be so tough, no one tells you. 18 months of not even being looked at in most cases, especially US companies (some Dutch companies di…

@akiselev: I haven't only been applying to US locations of US companies (e.g. Google Switzerland). But a SWE from Google gave me a referral link to which I'm grateful. I wonder to what extent it will help though since it went through HN and we just met. I want a phone screen and be told that I am or am not good enough. I can have peace with this if one FAANG company shoots me down on my D&A skills. For my own sanity,…

I don't know how many satellite offices you've applied to in the EU but I bet your sample size is very small. Don't get discouraged! Hiring processes at most companies are byzantine, multinationals even more so. It's really hard to tell who's pulling the strings or what the constraints are unless you're directly involved in the hiring decisions. I've passed on at least one once-in-a-lifetime hire (brilliant engineer from the Indian space program) just because I thought the H1B paperwork would interfere with a project deadline since my employer was a startup and didn't have the HR infrastructure a big company would have.

More importantly: fuck them. Don't judge your self worth based on the latest shiny megacorp's inscrutable interview process.

Post reply on HN