HNHacker News
TopNewBestAskShowJobs

qPM9l3XJrF

604 karma · joined June 18, 2020

submissionscomments
qPM9l3XJrF··on Warnings of AI doom gave way to primal fear of primates posting
The people concerned with AI doom stopped talking to journalists because journalists typically didn't report very well on the issue.

>...something I’m very concerned about is the use of AI for autonomous weapons. This is another area where we fight against media stereotypes. So when the media talk about autonomous weapons, they invariably have a picture of a Terminator. Always. And I tell journalists, I’m not going to talk to you if you put a picture of a Terminator in the article. And they always say, well, I don’t have any control over that, that’s a different part of the newspaper, but it always happens anyway.

https://futureoflife.org/2020/06/15/steven-pinker-and-stuart...

qPM9l3XJrF··on Are deep neural networks dramatically overfitted? (2019)
Thanks. The implication seems to be that if you restrict the # of parameters so it's equal to intrinsic dimension, learning isn't possible, but it is possible with this random projections method. Wonder why. It seems like with both methods, the number of possibilities being explored is the same, but the higher-parameter model space is richer with solutions for some reason.
qPM9l3XJrF··on Are deep neural networks dramatically overfitted? (2019)
The intrinsic dimension idea seemed interesting, but I didn't follow this bit:

>By searching through the value of d=1,2,…,D, the corresponding d when the solution emerges is defined as the intrinsic dimension.

Is the idea to increase the "intrinsic dimension" until the network is able to learn? I'm not sure I buy the part of the argument where because the network is able to learn when the "intrinsic dimension" is kept low, it therefore follows that the "intrinsic dimension" is also low even when we aren't forcing it to be low. It seems a bit like saying "because we can obtain 90% accuracy on MNIST with a 10-parameter model, it therefore follows that a 1000-parameter model for MNIST only has 10 'intrinsic' parameters". Seems like a dubious / handwavey argument to me.

qPM9l3XJrF··on A CO2 capture solvent with exceptionally low total costs of capture
"just reduce our energy usage" has been the plan for decades, and it's easier said than done. Also:

"Our hardest climate problems – the ones that are both large and lack obvious solutions – are agriculture (and deforestation – its major side effect) and industry. Together these are 45% of global carbon emissions. And solutions are scarce."

https://techcrunch.com/2019/02/15/how-to-decarbonize-america...

At the very least I think carbon capture makes sense for the long tail of industrial processes where the "CO2 emission" to "cost of reinventing process" ratio starts to get less and less favorable.

But even if you think we can decarbonize every last industrial process, we might as well develop carbon capture in case too much carbon gets emitted before that goal is achieved.

qPM9l3XJrF··on Humans display a few consistent behavioral phenotypes in two player games (2016)
Here are the player categories that the developers of Magic The Gathering came up with

https://mtg.fandom.com/wiki/Player_type

qPM9l3XJrF··on Google is accelerating reopening of offices and putting limits on remote work
Can you give specific instances of this happening? I feel like offshoring tech jobs has been the bogeyman that has never delivered in our industry for a really long time.
qPM9l3XJrF··on 'Fake' Amazon workers defend company on Twitter
Even assuming you can get rid of paid shills, you're still gonna end up with most messages being written by people who are extremely online.
qPM9l3XJrF··on U.S. Strategic Command response to FOIA request regarding gibberish tweet
5 years later... "Daddy wants to tell you a story about something you did when you were very young"
qPM9l3XJrF··on Female Founder Secrets: Men Clamming Up
Yes. If you aspire to "be somebody" then reputation management is important. But if your entire career will be spent working for employers who drug test and ask you if you've been in jail, but don't Google for your name, and you're not important to have the Google results of your name actually be about you anyway, then what is there to worry about?
qPM9l3XJrF··on Female Founder Secrets: Men Clamming Up
At a previous job, I worked under a woman who managed a practically-all-male engineering team. I think part of what made her successful at this was that people in our company had a great sense of humor, and she was an enthusiastic participant in our lunchtime bull sessions, ribbing each other and whatnot.

>Becca: What are you doing here? I figured they’d have locked you away in the psych ward for good by now.

>Scott: Nope. And what are you doing here? You haven’t killed off all your patients yet?

>Becca: Only person in this hospital I might kill is standing right in front of me.

>Scott: Be careful, I’m armed and dangerous picks up a central line placement practice set menacingly

https://slatestarcodex.com/2014/04/23/friendship-is-counters...

Not only did these kind of conversations make the company a more fun place to work, they also made it easier to speak critically to each other, because our critical feedback didn't seem like a big deal in light of the daily ribbing. There wasn't a lot of corporate BS at that company either--maybe not a coincidence.

I imagine there are other things you could do, like curse frequently, if you wanted to broadcast that words don't easily upset you. But telling jokes seems best if you think of one.

qPM9l3XJrF··on Female Founder Secrets: Men Clamming Up
>If someone asked me this, I'd probably think they used hashtags like #redpill or were into bashing Ellen Pao or something. It comes across as, "you can't say anything these days without being offensive, men in tech are soooo mistreated."

Wise moderates don't even join the conversation on social media. A wise decision for an individual, but it's harmful to all of us collectively. One negative side effect is people assume that everyone is an extremist of one stripe or another.

qPM9l3XJrF··on Groove Pizza
As a musician I'm surprised how much I like the sound of placing beats in a practically random way

https://apps.musedlab.org/groovepizza/?museid=xFlGGuYki&

Wish there was a "randomize" button that would let me input a probability that any given beat should be present and then randomly decide on/off for every node, I feel like that could be an interesting way to explore beatspace.

qPM9l3XJrF··on Sovereign writers and Substack
Just asking readers/customers to re-enter their payment details on a new platform could easily trigger 10% churn in subscribers no?
qPM9l3XJrF··on GPT Neo: open-source GPT model, with pretrained 1.3B & 2.7B weight models
Many people (correctly, in my view) criticized OpenAI for the name, saying that openness should be evaluated on a case by case basis. Glad they listened to critics instead of trying to maintain consistency for its own sake.
qPM9l3XJrF··on Google Analytics: Stop feeding the beast
I also find the claim itself rather implausible and it makes me think the author just has a generic anti-tech bias. (Know what else Google is responsible for? Forwards from Grandma! And reposts on your favorite subreddit! And people ghosting you on Tinder!)
qPM9l3XJrF··on How Vulnerable Is the World?
>I'm saying that the exponential rise of technology coupled with the same exact type of mind our ancestors had 2,000 years ago is a significant source of the issue right now. The solution to create even more powerful technology to solve this problem will likely only lead to a bigger problem in the future.

I don't think that's obvious. It could be that we're at a part of the techno-societal landscape that happens to be bad, but if we keep working on technology, we'll get to a new part of the landscape that happens to be better. For example, as VR tech gets better, maybe it will replace social media with something healthier. ("If you're going through hell, keep going" as they say.)

qPM9l3XJrF··on Block Party Aims to Be a 'Spam Folder' for Social Media Harassment
https://twitter.com/timnitGebru/status/1346940524603858944

She initiated the interaction by quote tweeting him

qPM9l3XJrF··on How Vulnerable Is the World?
The most important climate legislation has bipartisan support: https://techcrunch.com/2019/02/15/how-to-decarbonize-america...

Framing everything in terms of past injustices turns what should be a bipartisan issue into a partisan one.

We need less infighting and more creative collaborative problem-solving.

qPM9l3XJrF··on How Vulnerable Is the World?
This comment seems a little inconsistent to me. You complain about the fact that everyone doesn't have the mindset to solve problems. Makes sense so far.

Then you say carbon capture tech isn't going to work because... it's too focused on solving problems constructively? What is wrong with the mindset of carbon capture tech, exactly? What is the right mindset to have? (But also, the idea that we can figure out what solutions work based on diagnosing the "mindset" of those that propose them, as opposed to analyzing the solution itself in detail, seems pretty silly to me. See https://en.wikipedia.org/wiki/Genetic_fallacy )

qPM9l3XJrF··on How Vulnerable Is the World?
Social justice seems like a distraction from these issues to me. It is focused on the past, not the future.
qPM9l3XJrF··on Block Party Aims to Be a 'Spam Folder' for Social Media Harassment
In the case I have in mind, the woman of color who said GFY initiated the interaction
qPM9l3XJrF··on Block Party Aims to Be a 'Spam Folder' for Social Media Harassment
Alright... so my 3rd cousin twice removed did something to oppress your 4th cousin once removed, and that's your justification for abusing me on the basis of my gender/race?

What is being accomplished here?

qPM9l3XJrF··on Block Party Aims to Be a 'Spam Folder' for Social Media Harassment
My great-grandparent harmed your great-grandparent, and now I am convicted in response?

At the very least this sentence would appear to violate statute of limitations :-) But the bigger problem is that the punishment has no deterrent value--there is no way for me to change my great-grandparent's behavior. And it is straightforwardly unjust to convict me for a crime I didn't commit.

I think this is actually a pretty racist idea--that because you share genes with someone, you are partially responsible for crimes they've committed. It would be like if someone in a small town committed a crime, and the town responded by having the police begin always harassing every member of the criminal's race.

qPM9l3XJrF··on Block Party Aims to Be a 'Spam Folder' for Social Media Harassment
Fair enough! Sorry to hear about that.
qPM9l3XJrF··on Block Party Aims to Be a 'Spam Folder' for Social Media Harassment
>women and people of color are particularly targeted for online abuse

I think you have to be careful with the vocabulary here. If a white man is being abused online, and he complains about it, his complaint may be rendered illegitimate on the basis of being "tone policing". He may be told that "freedom of speech does not mean freedom from its consequences". He may be told that the people abusing him are under no obligation to educate him and are doing "unpaid emotional labor".

"I call out bad behavior, they are trolling, you engage in targeted online harassment."

If anything, I've noticed that women and minorities seem to have more of a blank check to get really crass and say things like "go fuck yourself", etc. I see people being mocked on the basis of being "mediocre white men" far more often on Twitter than I see them mocked on the basis of being "mediocre women of color". Gendered/racialized equivalents of "toxic masculinity" and "whiteness" aren't as widely employed.

(Agreed that social media can be terrible; I hope this initiative makes it better.)

qPM9l3XJrF··on Statement on New York Times Article
I've come to respect The Economist and The Financial Times. Their articles don't go viral as much--I suspect because they aren't as optimized for virality ;-)
qPM9l3XJrF··on The Making of Facebook’s Supreme Court
>Sophisticated questions about “automatic tools for takedowns” and the “equity principle with diverse communities” were soon overtaken by a joke about “Game of Thrones.” Posts were initially anonymous, but users quickly found a way around the system; “Harold” wrote, “I figured out how to identify self,” which provoked laughter. The moderator shouted to regain control of the room. In the midst of the chaos, someone posted, “Can we abandon Slido and talk?,” which quickly accumulated likes.

This passage made me appreciate the culture of decorum that has evolved around our court system. My inclination would be to disband that group and keep retrying until you can find a group that takes their duties seriously.

qPM9l3XJrF··on The Queen has more power over British law than we thought
Here is Vox, of all places, arguing in favor of constitutional monarchy

https://www.vox.com/2015/9/9/9294955/queen-elizabeth-constit...

qPM9l3XJrF··on Fueling Reddit’s Future
They said they are doubling their staff. Maybe you should be the change you want to see in the world and apply. It's an opportunity to have a highly leveraged impact, because even though reddit is one of the world's big social media sites, the staff is small.
qPM9l3XJrF··on U.K. coronavirus strain doubling in the U.S. every 10 days
What if the people who call out the names they don't like are the ones trying to score cheap political points? That seems more likely to me.

We could have controlled this virus a lot better in the beginning if it wasn't for cheap political point scoring of the form "the real problem isn't the virus, the real problem is anti-Asian racism"

https://www.cnn.com/2020/02/20/us/coronavirus-racist-attacks...

https://www.washingtonpost.com/health/2020/02/25/shen-yun-co...

Almost half a million Americans have died from the virus. To my knowledge, 0 have died from anti-Asian racism from the virus.

← PreviousPage 6 of 9Next →