HN user

eroo

1,370 karma

#TODO: fix this later

Posts39
Comments38
View on HN
blog.openai.com 9y ago

Robust Adversarial Examples

eroo
166pts47
www.nytimes.com 9y ago

Cassini Moves Inside Saturn's Rings

eroo
2pts0
www.nytimes.com 9y ago

Cassini Moves Inside Saturn's Rings

eroo
2pts0
www.spacetelescope.org 9y ago

Observable Universe contains ten times more galaxies than previously thought

eroo
4pts0
www.nytimes.com 9y ago

Flooding of Coast, Caused by Global Warming, Has Already Begun

eroo
5pts1
www.nytimes.com 10y ago

MacArthur Foundation Will Award $100M for Solution to a Global Problem

eroo
4pts0
www.nytimes.com 10y ago

Google and Apple: The High-Tech Hippies of Silicon Valley

eroo
6pts0
blogs.wsj.com 10y ago

DoorDash Raises $127M in “Down” Round

eroo
4pts0
www.nytimes.com 10y ago

Philip Knight of Nike to Give $400M to Stanford Scholars

eroo
4pts0
www.bloomberg.com 10y ago

Faraday Follows Tesla to Nevada for $1B Auto Factory

eroo
5pts0
techcrunch.com 10y ago

Reid Hoffman to Teach “Blitzscaling” at Stanford This Fall (App Due 9/18)

eroo
1pts1
www.businessinsider.com 11y ago

Sam Altman says we're in a 'mega bubble' that won't last

eroo
8pts1
techcrunch.com 11y ago

Companies That Presented at Y Combinator Demo Day 1

eroo
108pts56
www.startalkradio.net 11y ago

The Future of Humanity with Elon Musk: Interview by Neil DeGrasse Tyson

eroo
79pts44
www.npr.org 11y ago

Why Many Smart, Low-Income Students Don't Apply to Elite Schools

eroo
109pts81
www.edx.org 11y ago

Get $1000 of AWS when you earn a certificate in Entrepreneurship 101 or 102

eroo
10pts2
www.newyorker.com 11y ago

Is Spotify the music industry’s friend or its foe?

eroo
31pts50
www.newyorker.com 11y ago

The Programmer's Price: Want to hire a coding superstar? Call the agent

eroo
374pts216
www.nytimes.com 11y ago

Texas Health Worker Tests Positive for Ebola + CDC Response

eroo
3pts0
www.macfound.org 11y ago

Meet the 2014 MacArthur Fellows

eroo
1pts0
arstechnica.com 11y ago

Amazon–not Google–reportedly buying Twitch for $1B

eroo
6pts0
www.khanacademy.org 11y ago

Why I'll Never Tell My Son He's Smart

eroo
266pts64
www.nytimes.com 11y ago

Transplant Brokers in Israel Lure Desperate Kidney Patients to Costa Rica

eroo
19pts10
www.npr.org 11y ago

Ebola Outbreak Is A Global Health Emergency, WHO Says

eroo
4pts0
www.newyorker.com 12y ago

New York’s Shadow Transit

eroo
4pts1
www.npr.org 12y ago

Safety Feature For Pedestrians Has Undesired Consequence

eroo
76pts93
www.newyorker.com 12y ago

Living the Fitbit life

eroo
116pts74
recode.net 12y ago

Google Ventures’ Bill Maris on Venture Capital in Silicon Valley

eroo
6pts0
www.theatlantic.com 12y ago

Why Audiences Hate Hard News–And Love Pretending Otherwise

eroo
3pts0
www.cloudflare.com 12y ago

DDoS attack protection for at-risk public interest websites

eroo
13pts8
Welcome, ACLU 9 years ago

Where was the outpouring of funding [...]

Build the thing that needs funding. Then, seek support and apply.

perspective: finished first year @ wharton mba and taking a year off before second year to do an ms in cs (background is quant but not super computational - I'm really enjoying diving deeper into the latter now though).

1) Only do the eMBA if you can get BIGCO to pay for it. Price is unreasonable because it's expected to be expensed to an employer and it is a serious time commitment. I don't think it'd be worth it unless BIGCO is invested in you enough to sponsor.

2) 11-12 year experience upon entry is late for full-time MBA (mean is 5, max is 13). The MBA won't give you the boost to executive management - no one hires fresh MBAs for exec roles. MBA or not, that'll only come from hustle.

3) No empirical data, but my working theory is f(work incredibly hard, be kind to people and foster both friendships and partnerships, always be seeking out opportunity)

Misc: the coursework can be really useful/fun depending on the program. For example, Wharton lets me do the coursework equivalent of a grad degree in statistics w/i the mba. Lastly, the ready access to such a diversity of smart, ambitious people mid-career is really cool.

Happy to chat further if you'd like or put you in touch with MBA or eMBA students/adcom folks.

New Image of Pluto 11 years ago

~10 years in flight.

~20 times the speed of a bullet.

~5 billion kilometers away.

And there are pictures from this. What a feat.

Fitbit Files IPO 11 years ago

Devices in millions [2011,2014]: 0.2, 1.3, 4.5, 10.9

Revenue in millions [2011,2014[: 14.5, 76.4, 271.1, 745.4

Paid active users in millions [2011,2014]: NULL, 0.6, 2.6, 6.7

I realize they sell more than just pedometers, but their core product is a 3-axis accelerometer that calculates a dubious step count w/o really actionable insights. Imagine what this can be once there start to be useful applications/more of the data.

Click bait with misleading headline. A better headline would be "Comcast plans to roll out personal broadband service at a higher fraction of current business plans."

"Comcast's new Internet service absolutely crushes Google's offering, with speeds twice as fast." It won't be symmetric like Fiber, so it's going to be (at best) 2x download and probably 1/x upload. This speed also only appears to be coming to Atlanta in the near future.

"For those who don't live in Atlanta, Comcast plans to offer 1Gbps plans to almost all of its Internet subscribers in 2016." This is bigger news, though it's still quite far away and not only will this not "crush" Fiber, where available, but will be strictly worse since it isn't symmetric, features Comcast customer service, and will surely cost many times what Fiber does.

I would love to know how the conversation about setting up her email went. Does she tell someone and they say "I'm on it," and then no one follows up or audits the infrastructure? Or was the setup one of her people's decisions and she didn't really know/care?

I'd consider that a corner case. And even then, it is possible to get something out. With private investments, that just may not be true or involves high ad hoc transaction costs.

My original reply was intended to point out that what was the top article comment at the time completely missed the point. Cuban is arguing that severe liquidity restraints are bad, especially so for small time investors. For scenarios like you describe, public exchanges aren't perfect either, but they are very, very good at facilitating near-instantaneous liquidity and they strictly dominate the current set of private crowd funding vehicles.

Sure they can except in the most extreme corner cases. You an contrive situations in which it's difficult to sell stock on the NYSE or NASDAQ, but it's basically always quick and easy (e.g., you can move $100K of FB stock in seconds using your phone during just about any market hour and often even outside of them). That is definitely not true of private markets, at least at the moment.

This isn't an indictment of valuations being necessarily inflated more than they were in 1999. He's critiquing a new preference for private investments which are very, very difficult to liquidate. That lack of liquidity, combined with easier access to investments via crowd funding has the potential to crash very similarly to mortgage backed securities.

It's easy to get in and impossible to get out. If things start falling, investors are locked in for the whole ride down. That structure combined with a heady appetite for putting it in the first place primes the pump for a painful crash.

[edit for question] He implies the SEC is restricting mechanisms for adding liquidity to private/crowd funded investments. Any idea if he has a specific proposal in mind?

He isn't arguing firms are overvalued by a greater degree now relative to 1999. He's arguing that investments in private firms, which are far more popular now, are worse for small players due to their lack of liquidity.

If things start going south in a private investment, a share holder may not be able to exit even at a large loss.

>"Enter the agents. Solomon describes himself as an equalizer. In creative industries, he told me, 'there’s always this pattern that the creatives start out at the bottom of the food chain and are exploited.'"

Even recognizing that the current hiring model has major inefficiencies, it's hard to not see this as awfully ironic.

>"part of our goal is to de-risk freelancing and make it more viable. [...] She also appreciated that they had been vetted for interpersonal skills. At one point, they had to speak directly with the health-care company’s New York offices. 'They were good,' she said. 'And it wasn’t embarrassing to let them out of their cave.'"

The value proposition of the agent, pushing both technical and personal professionalism of candidates, should be addressable through a reputational system that doesn't take 15% and require ad hoc negotiations. It would, however, have to be complex enough to address how well certain talent is at addressing specific projects. How much of that is a lack of proper metrics and how much is the hiring party's inability to frame their needs?

Crypto 101 12 years ago

[1] is starting next week and I was considering spending time on it. I'd really appreciate any feedback from those who have taken it!

[Edit] PLEASE DONT DELETE YOUR REPLY IF PEOPLE ARE GIVING THOUGHTFUL FOLLOW-UPS.

That is a very mature sample given the short duration of the relevant time frame. Admittedly I'm ignorant about the specifics of several of these 39, but I recognize most as public, acquired for $1B+, or pretty clear "winners" (e.g., Uber).

Do any in particular seem suspicious or are there examples you think are missing? I'm not arguing an ideological position, I'm genuinely interested in the "right" answer. The definition of "success" is nebulous, and data are limited, but what's available would suggest you're strongly overreaching.

2 more recent ~$1B ones that just came to mind: Wayfair and HubSpot, both of which are MBA founder teams. The latter is also a dual technical founding team: ee and cs undergrad majors.

This is empirically false. A recent look at 39 $1B+ companies founded in the last 10 years: 33% had an MBA founder, to say nothing of the MBAs that have played significant early-stage exec roles (e.g., Facebook's Sandberg, Evernote's Gullicksen)[0]

Amazon also does not appear to be "anti-MBA."[1]

A broader misunderstanding comes from the fact that MBAs are all a particular "type." Top MBA programs are going to have everything from engineers to Peace Core volunteers. Is there any reason to suspect comp. sci. majors who spent 3 years at [insert reputable tech, consulting, banking firm here] and then went to get an MBA are systematically worse for tech jobs than those who did not?

[0] http://bostonvcblog.typepad.com/vc/2013/11/unicorns-and-mbas...

[1] http://www.amazon.jobs/team/university-mba-graduate

Steven Pinker offers a very convincing dissection of the referenced Deresiewicz article [1].

[1] http://www.newrepublic.com/article/119321/harvard-ivy-league...

TL;DR:

- Deresiewicz is willfully ignorant/misleading in his pursuit of being controversial.

- Ivy league students are smart, but there are smart students everywhere.

- Admissions to Ivy league schools is not as objective as it should be. There is too much focus on fluff and not ability/potential (e.g., meaningless "volunteer work").

- College isn't for everyone, but if you're going to go, an Ivy tends to be the best place because they have the most resources (and therefore are often counter-intuitively the least expensive).

"What it costs" shouldn't be represented as a percent of mean household income. This is very misleading. For example, if you drive out poor by increasing prices it may well make a city "more affordable" because both the numerator (prices) and the denominator (income) are increasing. Despite what Figure 6 would suggest, New York is not considerably "more affordable" than Detroit.

I wasn't aware of this research. Reflecting on my own schooling experience, however, there is something pleasantly intuitive about it.

I'm always impressed with Salman Khan's work.

There are two problems that too often seem to be conflated as a single issue: (1) we need more STEM students/workers, because they will build the future, and (2) there is persistent discrimination (notably sexism) in some of the sub-specialties that can't be tolerated.

Is there a reason to suspect that tackling both of these at the same time (e.g., demographic specific initiatives) yields significant benefit over alternatives?

In Boston, an artificial scarcity of housing is the unfortunate result of laws that effectively limit new development. You either can't build new, can't build too high, or have to meet a neighborhood's aesthetic to get approval. The result is higher prices, more sprawl, and a generally depressing force on the kinds of great things that come from sufficient population density.