Y
HN Search
Hacker News Search
new
|
comments
|
top
|
jobs
datsci_est_2015
searching PlanetScale…
1.
▲
2.
▲
3.
▲
4.
▲
5.
▲
6.
▲
7 ms
·
91.
▲
by
datsci_est_2015
2mo ago
Almost nothing like what is seen in police procedural television series, that’s for sure. Although there is (allegedly) a fun feedback loop that cops more and more expect their jobs to be like police procedurals and try to solve crime as if
92.
▲
by
datsci_est_2015
2mo ago
Local (to us) firm. Two main reasons I didn't: - Have never seen it mentioned on this forum, in any context. - We're potentially pursuing legal action.
93.
▲
by
datsci_est_2015
2mo ago
It is interesting, how much money is being lost during this outage? My significant other was just let go from their job as a scapegoat for an organizational error: 3 layers of failure - IC, manager, director, and the IC was let go. The er
94.
▲
by
datsci_est_2015
2mo ago
I'm reminded of the scene in Independence Day with the people on top of the tower welcoming the alien spaceships. Except instead of celebratory fanatics holding poster boards saying "Welcome", it's all the singularity AI
95.
▲
by
datsci_est_2015
2mo ago
Effective altruism may have had a positive reputation at some point, but it's definitely jumped the shark and almost everyone influential who invokes in the context of AI it seems to have fallen into the trap of the "bottomless pi
96.
▲
by
datsci_est_2015
2mo ago
It's likely a mixed bag. If you're diligent about which organizations you choose to support, you're more likely to have a positive impact. If you want a simpler example that doesn't involve biotech research, I would just
97.
▲
by
datsci_est_2015
2mo ago
Is there something like a meta-Dunning-Kruger, where someone invokes Dunning-Kruger in a way that itself is a potential Dunning-Kruger phenomenon? I'm inclined to believe that the thing - that everyone intelligent says is complicated
98.
▲
by
datsci_est_2015
2mo ago
No, but to roll with my comparison for humor’s sake, writing an email to hire a hitman to kill someone for you isn’t protected under the first amendment either.
99.
▲
by
datsci_est_2015
2mo ago
One step further, our current generative AI is not discrete. “AGI checking other AGI” doesn’t really make technical sense. Currently, every interaction between two generative AI units is treated as agglomerative. We talk about “Claude”, not
100.
▲
by
datsci_est_2015
2mo ago
The fact that this is getting such little traction in comparison to a very related post yesterday is funny: - The US tried to stop cartel money-laundering; devastated mom-and-pop businesses: https://www.theguardian.com/us
101.
▲
by
datsci_est_2015
2mo ago
“Will no one rid me of this turbulent priest?” — “I didn’t directly tell them to make white nationalist content, I just paid them for being an engaging content creator.” — Hope that we get to a point as a society that we acknowledge that co
102.
▲
by
datsci_est_2015
2mo ago
The fact that antisocial (e.g. white nationalist) content creators are the most “popular” (e.g. drive the most engagement) is not immaterial to this topic. It’s a very flimsy defense. “It’s not my fault that the cartels pay the most for hea
103.
▲
by
datsci_est_2015
2mo ago
Not to get weighed down in pedantry, but the Wikipedia article is “American imperialism”, not “the American empire”. Taking imperialist actions is a low bar for determining what is or what is not an empire. I wouldn’t call Yugoslavia an emp
104.
▲
by
datsci_est_2015
2mo ago
Did the US? I would argue it’s been going on for 80 years, in tandem with the rise of the military industrial complex. Go back much further than that, though, and the argument starts leaking into whether the US was ever great to begin with
105.
▲
by
datsci_est_2015
2mo ago
I agree, much of the Balkan suffering from the latter half of the 20th century could have been curtailed by properly stamping out ultranationalist groups as a punishment for their ethnic cleansing. The Ustaše, for example.
106.
▲
by
datsci_est_2015
2mo ago
It’s funny you bring up Union vs Confederacy, because Reconstruction should be a blueprint of what not to do. We’re still paying today for the failures of our forefathers to not have the bravery to stamp out the worst cultural elements of
107.
▲
by
datsci_est_2015
2mo ago
I had to think carefully, but I think they’re implying that the worst thing the US has done is invent Twitter, and comparing it to the worst thing that Europe has done (implying that that’s the holocaust), and that Europeans specifically(?)
108.
▲
by
datsci_est_2015
2mo ago
I think the whole “empire collapsing” is a bit of a dramatized phrasing, though historians can correct me. I think it’s much more death by a thousand cuts, and the UK is a great analogue in that regard. Look at Brexit, for example. Or Black
109.
▲
by
datsci_est_2015
2mo ago
> Take, for example, Elon Musk in SpaceX. Elon knew enough of rocket science to assemble the early team and participate as a self-assigned chief designer. In many cases such actions would qualify a person as a smart one. There’s so much
110.
▲
by
datsci_est_2015
2mo ago
What does it mean to be “smart”? I could list a ton of synonyms just to highlight not only the ambiguity of such a claim, but also its relative worthlessness: - Conniving - Autistic - Empathetic - Talented - Quick-witted - Wel
111.
▲
by
datsci_est_2015
2mo ago
Couldn’t disagree more. AI is a button that you press and magically get results. The ability to judge the quality of those results is “taste”. Anyone can push the button, but not everyone can discern whether the result is a good one or not.
112.
▲
by
datsci_est_2015
2mo ago
I’ve started requiring that unless absolutely necessary, all comments and docstrings be one-liners in our code 10k+ codebase(s). LLM-generated code has a tendency to “over-justify”, which is reasonable when first reviewing the code as a hum
113.
▲
by
datsci_est_2015
2mo ago
> Also the game was on a timer, and maybe there are some very abusive workplaces where you feel that kind of pressure, but I think most of us actually take the time to understand what a being asked before approving it. Are there people o
114.
▲
by
datsci_est_2015
2mo ago
> But in a very different sense of the word "Conservative", aerial bombing is often extremely Conservative. That sense is minimizing political commitment to a conflict. A leader (with effective air superiority) can drop some bo
115.
▲
by
datsci_est_2015
2mo ago
I agree mostly with the other commenter that it’s like nutritionist vs health influencer. But I’d also say that there’s an inherent bias to the word “eugenics” fundamentally: the implication that there are objectively good genetics and bad
116.
▲
by
datsci_est_2015
2mo ago
I'd personally label the factions within the military that view aerial bombing as an effective strategy as "conservative", especially the ones that follow in the footsteps of Henry Kissinger and Curtis LeMay. It's hard n
117.
▲
by
datsci_est_2015
2mo ago
> I mean, if we wanted every human on Earth to be over 6' tall, we could just cull the short ones (or stop them from reproducing), and this is scientifically valid, but, of course, evil. What's scientifically valid about that?
118.
▲
by
datsci_est_2015
2mo ago
> Eugenics can be understood as a subset of the ideas of plant and animal breeding that have been practiced for centuries. That's called "artificial selection", and people who believe that certain groups of humans are some
119.
▲
by
datsci_est_2015
2mo ago
> You can't win a war purely through air superiority and bombs. That's not a new lesson but it's one the administration failed to learn through history and is apparently now failing to learn through practice. This is a les
120.
▲
by
datsci_est_2015
2mo ago
Sounds like OP’s friend was looking for a website, and your girlfriend was looking for a hobby.
More ›