HNHacker News
TopNewBestAskShowJobs

teaman2000

177 karma · joined March 4, 2014

Product manager for lots of stuff
submissionscomments
teaman2000··on Clickbait, engagement, and credibility in YouTube's medical ecosystem
It's always useful when a study can quantify the obvious. But it is depressing.
teaman2000··on Reasons veterans are especially hard-hit by federal cuts
1. Doing things like randomly firing the staff who were hired recently to address gaps in care is NOT the way to improve things. And it takes staff, time and money to make new processes.

2. The article points out ways, besides VA cuts, that veterans are being harmed. For instance, many of those government workers fired without cause are veterans, and the veterans' suicide hotline (already understaffed) is going to hemorrhage more staff when the edict against remote work goes into effect.

teaman2000··on Facebook is blocking access to data about how much misinformation it spreads
I am constantly amazed at how reputable news outlets like the New York Times misstate data from Facebook's CrowdTangle - usually in a way that overstates the spread of the questionable information. But Facebook seems to deliberately want to prevent outsiders from getting access to information that would help us judge the scope of the problem.
teaman2000··on Remove left turns for less dangerous city traffic
There are lots of roundabouts in the Boston area (where we call them rotaries). People encounter them regularly and STILL don't know how to use them. Although recently there have been great strides made in using lane markings to steer drivers through more effectively.
teaman2000··on Kids with a desk and a quiet place to study do better in school, data shows
I guess the question to ask is: if you gave the kids a desk and a room of their own by itself, would academic performance improve? Or would academic performance improve more if you raised the parents out of poverty?
teaman2000··on Is YouTube embedded playlist broken?
That didn't work for me, alas. But adding a random query parameter did: https://www.youtube.com/embed/?z&list=YOUR_PLAYLIST_ID
teaman2000··on Feeding cows seaweed daily could reduce their contribution to climate change
I know, the answer to the climate toll of eating meat is vegetarianism. But I like to see this more as a question of harm reduction - hundreds of millions of people will not give up meat tomorrow.
teaman2000··on What happens when guys add their cats to their dating app profiles
Ummm, in the article it says: "Each participant only rated one man, with and without a cat."

The top photo in the article is a posed photo. But the second one is one that was part of the study.

teaman2000··on How kids in a low-income country use laptops: lessons from Madagascar
I got my start with Logo (which I think is a direct ancestor of Scratch) from MIT
teaman2000··on A month in, Tesla’s SA battery is surpassing expectations (inc. charts)
What's unclear to me is given how much extra electricity (20%) is consumed in charging up the battery, is this more efficient than just firing up another turbine?
teaman2000··on When You Call 911 and Nobody Picks Up
The credibility of this article is undermined by the statement that Boston's 911 calls are sometimes unanswered - backed up by a link to a 2006 article from a student newspaper.
teaman2000··on Uber’s Didi deal dispels Chinese ‘El Dorado’ myth once and for all
A lot of this is just the problems in any foreign market. It's the same in the US - how many Americans shop at Baidu's US ecommerce site? Or use Weibo or vKontakte?
teaman2000··on Is it time to eliminate tenure for professors?
Since the tenure system is so inflexible, universities instead rely on semester-to-semester adjunct faculty.

Sounds like someone needs to make Uber for higher ed.

teaman2000··on Ask HN/YC: Please suggest me some good books I can gift a 8 year kid
My Weird School, any of the 60 volumes or so, all of which are silly.
teaman2000··on Test your “free will” at the Aaronson Oracle
Well, it's less about your 'free will' and more about one's inability to generate a random number.

I'm not going to take the time to test it, but if you flipped a coin before each key press, this algorithm would eventually get to 50% accuracy. But in the short term it would probably be <50%, because it assumes it won't actually get random input.

teaman2000··on Game theory explanation of why journalists don't call out lying politicians
Or, to put it in terms of the article, there are costs to being honest - both in terms of numbers of viewers and readers, and in reputation.

But readers are a lot less interested in serious news than entertainment.

teaman2000··on Universities Are Becoming Billion-Dollar Hedge Funds with Schools Attached
The issue is not so much that schools invest in hedge funds (the endowment managers are just trying to maximize returns), but that the endowments pay out such a small percentage of their funds each year to support research, financial aid, and everything else. Harvard pays out only 4-6% of the endowment yearly. On their huge endowment, if they increased the payout by .5%, they would have a whole lot more to spend on worthy activities instead of saving for the year 3000.
teaman2000··on Even if Netflix is serious about blocking VPNs, it's unlikely to succeed
I think the average VPN user IS pretty tech savvy.
teaman2000··on TechCrunch has 2M Likes but they only get 300 Likes per Post. Fake Likes?
No, I think it just reflects that Facebook put the average post in front of a tiny percentage of fans of a page, and of those only a small percentage will like, comment or share.

If a larger than usual proportion of users interact with the post, Facebook will start showing it to more people - possibly starting a snowball to full-fledged virality.

teaman2000··on Boston's Basic City Services Dashboard
Boston's previous mayor, for 20 years, didn't even allow voicemail on phones at City Hall. But he did fund an office that build the Citizens Connect app and made great strides in freeing up public data. It's a refreshing change to have a mayor who will now look at the data.