revelio

56 ms·

571.
by revelio 4y ago
Satellite data isn't that good unfortunately. Tide gauges say 1.5mm/year of rise, satellites say 3mm/yr, and they estimate satellite error to be +/- 0.5mm year, so the error bars are a significant fraction of the size of
572.
by revelio 4y ago
Heh. Also important: anyone can object to the presence of information that mentions them or they created being known to the AI at any time, and if they object within writing you have 3 days to re-train the AI to remove whatever they objecte
573.
by revelio 4y ago
"Scientists" and "journalists" are the very people being accused of propagating pseudo-science so who cares what they think? To resolve disputes about whether certain claims are scientific you can't just ask one sid
574.
by revelio 4y ago
But your experience isn't universal or even common. Most of the engineers I've encountered can and do critically evaluate work in social science and the humanities, correctly conclude it's built on a very flimsy foundation
575.
by revelio 4y ago
You're not rebutting Walter's argument successfully. Everything you've said there is extremely subjective, difficult to pin down and mostly an emotional argument to begin with. Even your starting "obvious" premise t
576.
by revelio 4y ago
Academics don't scrutinize left wing ideology so "academic scrutiny" is pretty meaningless for claiming it's accurate.
577.
by revelio 4y ago
What? The word "speech" in the legal/rights/democratic context doesn't literally mean vocalizations that create vibrations in the air, only someone engaged in motivated reasoning would interpret it that way. The wor
578.
by revelio 4y ago
It's not actually. If you parody a corporation this way they're going to come after you and nobody will be surprised. If you want to parody Google or Apple or whomever then you need to make a company name that's different but
579.
by revelio 4y ago
>> a unhealthy overdose of Der Stürmer and lashing out against some scapegoat minorities lead Europe into WWII Except that: 1. Hitler rose to power mostly on the back of widespread civil unrest caused by the conditions placed on Ger
580.
by revelio 4y ago
Does it take a lot of effort? The article claims this particular tribe only had "first contact" with western civilization in the 50s or 60s or so.
581.
by revelio 4y ago
You're right. I thought tove was a man's name for some reason.
582.
by revelio 4y ago
OK, let's get one thing straight. Everyone claims to be pro-science. There is literally no political party on earth that stands up and says "we hate science". When you describe people as pushing back against "science&q
583.
by revelio 4y ago
No it's not, because the same effect is seen everywhere but only the US has the Republican party. Which would argue the opposite anyway - it does in fact believe in the scientific method, which is why they criticize academic research s
584.
by revelio 4y ago
On academic takeover, the statistics are quite clear that 60 years ago there were far more Republicans in US academia than there are today. Economics departments may have somewhat less skewed numbers but you would certainly expect to see a
585.
by revelio 4y ago
Are you arguing that they were once a lot more sophisticated and then regressed back to the stone age? It seems somehow unlikely given the description. Even if they fled contact with outsiders, they wouldn't have forgotten everything a
586.
by revelio 4y ago
The focus on the Yamomamö appears to be simply to ground the article in something concrete, making it more interesting. He does claim that this once heretical view is now widely accepted in anthropology and it's a description of many t
587.
by revelio 4y ago
>> observations from one tribe The article claims otherwise: As time went by, evidence for Chagnon's claims became too overwhelming to ignore for most anthropologists. Numerous other anthropologists came to the same conclusion
588.
by revelio 4y ago
Midjourney has stopped offering free generation. I made all of three or four images and now it is saying they don't have capacity for free trials anymore.
589.
by revelio 4y ago
Right, that's totally it. I came away thinking the interviewer seemed way sharper than the interviewee which is pretty rare. The sheer throughput and speed of interesting questions was incredible. Too bad that many of the answers were
590.
by revelio 4y ago
I emailed Twitter's official press email address, press@twitter.com, for comment. Twitter automatically replies to all emails at that account with the message "<poop emoji>" Jesus Christ. It's a literal shit-show
591.
by revelio 4y ago
The interviewer and his questions is in some way more impressive than the answers, which is weird.
592.
by revelio 4y ago
It's not just about hardware, it's about the software infrastructure to go with it. Other than OpenAI most researchers weren't interested in "merely" scaling things up because that was/is seen as simple enginee
593.
by revelio 4y ago
I'm a recent convert, been experimenting with converting one PL to another. GPT-3.5 will get the gist of what the code is doing, and then provide what looks like a direct translation but differs in numerous details whilst having a bunc
594.
by revelio 4y ago
I was sitting in a Starbucks yesterday sitting opposite two girls doing some language study and one of them was joking that the other should ask ChatGPT about some grammar point. This idea that only tech people know about it seems way off.
595.
by revelio 4y ago
Yep. The best man's speech at my brother's wedding was a disaster zone in every possible way, it was a sad blemish on an otherwise great day. It'd have been far better if he'd have some AI help along the way.
596.
by revelio 4y ago
But in practice (in the USA) broadcast licenses are just a way to manage spectrum and don't get revoked for political reasons. Funding can go up or down depending in order to condition behavior.
597.
by revelio 4y ago
That last one is golden.
598.
by revelio 4y ago
Broadcast licensing isn't the same thing as direct subsidization, though.
599.
by revelio 4y ago
I think that touches on one of the things that could be easiest to improve. HN has upvotes and downvotes without any guidance on what people are meant to be voting about (agreement? post quality?). So people just downvote stuff they disagr
600.
by revelio 4y ago
That's not really addressing the point though, is it. That tradition at Google evolved when they had technology that was well ahead of their competitors, which they currently are not and therefore this paper is more like a marketing do

More ›