Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
hooande
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
61 ms
·
481.
▲
by
hooande
6y ago
The point is that the majority of people live boring lives and don't have enough wealth to be a valuable target. I certainly fall into this category. I'm not important or interesting enough that anyone is going to look into me. It
482.
▲
by
hooande
6y ago
...Google Reader still exists?
483.
▲
by
hooande
6y ago
can't this variance be explained more accurately by relative income? the US is more ethically diverse than many western nations, but also has high income inequality. This inequality, unfortunately, is distributed ethically
484.
▲
by
hooande
6y ago
this seems like it might add capital to existing investment markets. at the least, it should make it easier for people to raise early investment rounds?
485.
▲
by
hooande
6y ago
if twitter is so bad, why do you keep using it?
486.
▲
by
hooande
6y ago
There are tradeoffs to every system. We chose first past the post and our executive structure for historical reasons. It's largely working for us (in a broad historical sense) and we're not going to expend the effort to change it
487.
▲
by
hooande
6y ago
It doesn't make sense to compare the US to "other countries" that have different political systems. We don't have a parliamentary system. The executive doesn't necessarily represent the majority party in the legisla
488.
▲
by
hooande
6y ago
This is a great idea, great product. Pricing is very reasonable. I'll recommend this
489.
▲
by
hooande
6y ago
is there a source for five attacks, other than this nitter post?
490.
▲
by
hooande
6y ago
Are you just dumbing down humans to match the model? LeCunn's post is very sparse on detail, but the point is that humans can easily reason about a vast number of things that any form of sequential language model cannot. That alone is
491.
▲
by
hooande
6y ago
The problem with ML/AI is that it almost never provides people with a significant new capability. It can improve and augment, but rarely lets people do something that they couldn't do before. There is no "killer app" It&
492.
▲
by
hooande
6y ago
wouldn't this leave you in a worse situation? say twitter becomes a publisher (whatever that means) and gets sued out of business. then you still won't be able to follow @NYPost or retweet content to followers. I'm curious, h
493.
▲
by
hooande
6y ago
make this service. "tune your own algorithm" social network. not sure how broad the appeal of that will be, but you can do an MVP to find out
494.
▲
by
hooande
6y ago
Twitter has made themselves clear on this. They're going to censor content they don't like, even in direct communications. Use another platform if you don't like it. If you can't do that because all of the content produc
495.
▲
by
hooande
6y ago
then don't use twitter and convince others you know to do the same
496.
▲
by
hooande
6y ago
Why couldn't twitter say "we will delete any tweets that are critical about a certain person"? is there a law against that? is it morally wrong? It seems that if they wanted, they could say "we want Candidate X to win an
497.
▲
by
hooande
6y ago
The issue here is that the company HAD a political mission, and decided to change that. Some employees may have joined specifically because of the company's politics and now they can leave. Church and state are separate because you don
498.
▲
by
hooande
6y ago
You would do better to come up with innovative new jobs for humans to do. It's easier to come up with a new service jobs than it is new materials to build cranes with Humans can always invent new forms of labor. Limited only by imagina
499.
▲
by
hooande
6y ago
I loved, loved X-COM. It had an ability to capture imagination in a way that few turn based strategy games do. The randomly generated soldier names still loom large as heroes in my mind. I was affected when they were killed in action. "
500.
▲
by
hooande
6y ago
to follow up on the point above, how do you know that leaving earth is the only way to escape extinction? that's only what you can imagine given the technology available, just like someone living in 320 couldn't conceive of space
501.
▲
by
hooande
6y ago
I think the problem with your approach is that no sane person would publicly defend racism. Because they would be outing themselves as a racist to all of their co-workers. I believe that I understand very well WHY people have racist points
502.
▲
by
hooande
6y ago
I would very much like to know for sure which of my co-workers are willing to defend racist points of view
503.
▲
by
hooande
6y ago
"nerds are socially awkward" is an observation that borders on being trite. There's a lot of truth to that of course, just because of the definition of the term "nerd". But it's something that sounds true as
504.
▲
by
hooande
6y ago
The federal budget is hundreds of tiktoks. and where would the money even go?
505.
▲
by
hooande
6y ago
Is there a difference between "counting" and "determining relative size"? for example, I can distinguish between a group of three people and a group of ten people without being able to count at all. There are big differe
506.
▲
by
hooande
6y ago
It looks like the source study [1] was based heavily on IBM Waton's personality traits API thing. "Personality" is a difficult thing to quantify and my experience with Watson is that it was good, but not great. Not sure I wou
507.
▲
by
hooande
6y ago
I spend hours per day browsing tweets. Nothing about this algorithm seems to be groundbreaking. TikTok has a strong brand and is benefiting from that
508.
▲
by
hooande
6y ago
You made a very compelling case about why it would be difficult for everyone to switch between software like the root comment describes. But the article goes a step further and says that Zoom specifically is of critical importance. I don&#x
509.
▲
by
hooande
6y ago
last I saw, roughly 80% of people in the US have employer provided healthcare. In a democracy a 20% minority has to be VERY vocal edit: I was incorrect, this is 80% of people get healthcare from their employer OR the government. Only 20% of
510.
▲
by
hooande
6y ago
I'd estimate that this is at least 15-20 years away, shortly before or after we have fleets of autonomous vehicles
More ›