shoguning

9 ms·

1.
by shoguning 5y ago
Advertising and media? Robots bid on the ads that billions see every day, and decide what videos billions of people watch.
2.
by shoguning 5y ago
I'm guessing low expertise programmers whose main contribution was googling stackoverflow will get less valuable, while high expertise programmers with real design skill will become even more valuable.
3.
by shoguning 5y ago
What you're saying is consistent with my hypothesis. Ie that casual public transmission on subways and in restaurants is low. Rather, it's "behind closed doors", at extended private gatherings, via domestic workers and s
4.
by shoguning 5y ago
According to a study in PNAS, shelter-in-place policies in the US did not work: https://www.pnas.org/content/118/15/e2019706118 Another study in Eur J Clin Inv had similar findings comparing countries: https
5.
by shoguning 5y ago
I think there is a proper level of conviviality/cordiality and fun that you can bring to work without getting too personally invested. A lot of first-time employees may have trouble with this balance, but after a tough break or two you
6.
by shoguning 6y ago
Is there any reason to use a 3rd party recruiter? The listings are all public, just submit a resume. All my offers came from applying direct. Always got the runaround from recruiters.
7.
by shoguning 6y ago
Or they believe (possibly correctly) that you don't have as strong a negotiating position because there are fewer options for you based on location.
8.
by shoguning 6y ago
> The reason excess deaths are only 3x higher is because of our precautions. It's a hard case to make. Sweden for example has had less excess death than harsher lockdown countries like Spain & UK. Lockdowns have done little if a
9.
by shoguning 6y ago
> We also outlaw speeding because speeding increases the odds that you kill someone else. We could go further and outlaw cars that don't have speed governors. That would save lives. Lower speed limits would save more lives as well.
10.
by shoguning 6y ago
It's not selfish to be out and about if you are young and healthy. The recovery rate for 20-30 YO is 99.99%+, and if you are healthy it's likely higher[1]. It's on the same order of risk as driving. Many young people die driv
11.
by shoguning 6y ago
California and SF are doomed, but the recent news of HPE and ORCL moving are not a big deal. IMO these are stationary phase companies trying to cut cost. They are not growing revenue, they are not really dynamic companies anymore. Who cares
12.
by shoguning 6y ago
> The bottom is still decently high even in Midwest cities Yes, and I also see the push for remote work also affecting the demand side of the software labor market as well as the supply. There will be even more push for good collaboratio
13.
by shoguning 6y ago
Yeah, that is mentioned as "one of the downsides of renewables". That is like saying, "one of the downsides of using a cooked noodle as a fork is that it doesn't hold its shape". It's disqualifying until that i
14.
by shoguning 6y ago
IMO, this is huge. One of the biggest applications of ML to science that I know of for sure. People used to manually crystallize proteins at great effort to solve for structures. Of course, there is a caveat. The static, crystallized struct
15.
by shoguning 6y ago
All I can say is, good luck. I discourage everyone from doing a PhD, but it sounds like you've made your mind up. Be very careful with your choice of advisor and department, and if it isn't working, be ready to walk away--sounds l
16.
by shoguning 6y ago
I was pleasantly surprised by this recently. Just worked using my touchscreen laptop. So rare on Windows.
17.
by shoguning 6y ago
Thanks! And yes, I agree. The ability to write ad hoc SQL is a great skill.
18.
by shoguning 6y ago
Pardon my ignorance--is inspections some concept that relates to database management, or are you referring to a query like "select * from inspections where status = 'open';". Honestly asking.
19.
by shoguning 6y ago
Most people don't care about climate change. They may say they do but the majority wouldn't do anything if it meant a sacrifice, even a very small one. 68% wouldn't pay even $10 per month [1]. Imagine how few would pay $25, $
20.
by shoguning 6y ago
> it's in the keywords This assumes that the source and target listings definitely have accurate keywords. There would be plenty of value in a model that can use either keywords OR images to effectively make recs. > Considering
21.
by shoguning 6y ago
Generating the vector representation is surely the hard part here, or at least the differentiator. The search algorithm, as you point out, is pretty much commoditized at this point.
22.
by shoguning 6y ago
I guess I was thinking of initial webpage load, which might include a bunch of network calls (DNS, CSS, JS resources). Each individual network call is faster tho, good point.
23.
by shoguning 6y ago
The latency is only reduced from tower to phone right? The tower still has to ping the rest of the network to find the service you were requesting to in the first place. If total latency ~ 500 ms, phone-to-tower is maybe 50ms of that in 4G.
24.
by shoguning 6y ago
Myself and many people I know bootstrapped technical PhD (earning potential in postdoc = $45k) into data science/ML jobs (100k+) using Coursera, textbooks etc.
25.
by shoguning 6y ago
On paper, of course they can do whatever they want. This just gives them more cover to do it, shielding them (at least partly) from internal blowback.
26.
by shoguning 6y ago
Exactly. For really large companies with many channels, it's very difficult to disentangle this. Here is a good discussion of the difficulty: https://thecorrespondent.com/100/the-new-dot-com-bubble-is-h...
27.
by shoguning 6y ago
Seems synergistic with what I said, no? Reduce budget and do some testing at the same time.
28.
by shoguning 6y ago
Maybe? Part of the leverage is going public and saying it's about some principle. For experimental control, executives tend to think simple is better. The more complex, the more room for fudging.
29.
by shoguning 6y ago
My interpretation of what's going on: CEOs finally have leverage against their marketing departments to do black-out testing on social media ad spend. It's difficult to actually measure impact of ad spending, and marketing teams s
30.
by shoguning 6y ago
My question is... do they really? They happen to be around when tech moves, but would it have moved without them? It's a historical counterfactual, and hard to argue about. I'm skeptical and I'm more skeptical of ARPA-e.

More ›