Veedrac

14 ms·

181.
by Veedrac 2y ago
The README covers this. Its argument is persuasive. If your point is that the constant is badly tuned for theoretical 1000 core machines that don't exist, I'm not sure I care. A 100ns stall at most every 100us becoming more likely
182.
by Veedrac 2y ago
I don't understand why it's unbelievable. They walk through the required logic to perform scheduling and dispatch; it looks incredibly simple and extremely suited to hiding on an OoO. The benchmark is summing down a binary tree wh
183.
by Veedrac 2y ago
2009 https://news.ycombinator.com/item?id=855684 2017 https://news.ycombinator.com/item?id=14790251 2018 https://news.ycombinator.com/item?id=17184054 2019 https://news.ycombinat
184.
by Veedrac 2y ago
The price of renewables scale down as demand rises, and the capacity we have building them is growing at an exponential rate at least somewhat capped by demand. It follows that if you want to displace existing power sources, you're pro
185.
by Veedrac 2y ago
Because the article is in context of new demand.
186.
by Veedrac 2y ago
Hence why I said “displacement effects are likely to dominate [...] the marginal environmental cost of the renewables themselves.”...
187.
by Veedrac 2y ago
The environmental externality of new energy in the US is likely ~0, potentially even net positive, because approximately all new energy in the US is renewable, and displacement effects are likely to dominate the small-single-digit percentag
188.
by Veedrac 2y ago
People vastly underestimate god. Players aren't just trying not to blunder, they're trying to steer towards advantageous positions. Stockfish could play perfectly against itself every move 100 games in a row, in the classical sens
189.
by Veedrac 3y ago
Given Bitcoin is at an all-time high, Ethereum is close, and if one held on to Bitcoin bought in most of 2019 they'd be up 10x, it seems not that obvious to me that the crypto skeptics were more correct than the crypto believers. Certa
190.
by Veedrac 3y ago
You're late to this party. https://news.ycombinator.com/item?id=17184054 But seriously, there's nothing in this article to debate, it is completely devoid of arguments for its position.
191.
by Veedrac 3y ago
Fair point.
192.
by Veedrac 3y ago
You can't properly make a moving image sharp at a finite refresh rate (though you can approach it for sufficiently high ones), because the object can only move in a temporally juddered path, which doesn't match the eye's move
193.
by Veedrac 3y ago
You don't solve the problem of having more demand than supply by subsidizing demand. You solve it by making it legal to build housing. There is already plenty of monetary reward in building houses in places where housing is expensive;
194.
by Veedrac 3y ago
This ruling means at minimum that a company can say ‘or sign up for less at on our website’. Cf. Spotify, who were allowed to serve customers that signed in from an account with an off-app subscription but were not allowed to tell the custo
195.
by Veedrac 3y ago
The developer had the opportunity to approve when it put the app up for its list price. IMO it is easier to see why this is OK if you imagine Apple weren't tying a waiver for this fee to being unable to inform the user about alternate
196.
by Veedrac 3y ago
I would think that case is at least meaningfully in my favour. The anti-steering ruling supports my point that while Apple can charge whatever commission it wants on its payment platform, this should not entitle it to ban or tax competing m
197.
by Veedrac 3y ago
I'm fine with Apple charging $.X per install from their app store. That should be in their right as a platform. What's not cool, and what I was excited for DMA to fix, is that Apple says that by transacting with a customer _by any
198.
by Veedrac 3y ago
> It's not hard to imagine road or environmental conditions that don't apply (significantly) to rockets: ocean breeze containing salty droplets This one is very much a staple of SpaceX's rockets.
199.
by Veedrac 3y ago
Well, more from the social media process that selects for whatever interpretation of evidence best matches the narrative they want, rather than the narrative that best matches the sum of evidence. The biography did not in fact say he did th
200.
by Veedrac 3y ago
No, this is manufactured.
201.
by Veedrac 3y ago
Thunderf00t has made a brand on putting out anti-Musk videos of generally terrible accuracy; he is probably not someone you should be sourcing. He is just as harsh about SpaceX and has been for a long, long time. He's just as harsh on
202.
by Veedrac 3y ago
This argument doesn't seem to track to me. Eg. if I rebooted any time I tried to plan how to kill someone, I don't see how this would make me materially worse at general tasks. Your argument suggests that it necessarily must. Note
203.
by Veedrac 3y ago
That AGI is likely to follow its own goals according to its interests, which we don't know how to shape or really reflect any robust properties at all, is exactly why alignment is hard and interesting. The part where you go from ‘this
204.
by Veedrac 3y ago
It might be about scalping, but also people willing to pay more are also more likely to wait if they have to, and keep future sales prices high, and people looking to sell are more likely willing to buy different vehicles. If Tesla stops ot
205.
by Veedrac 3y ago
This is almost surely just because it wasn't a real Turing Test. The paper discusses some limitations.
206.
by Veedrac 3y ago
The Turing Test is not intended to be fair. It is intended to be sufficient.
207.
by Veedrac 3y ago
...what? Are you really claiming that EA is bad for... advocating for public policy around AI xrisk, rather than just writing off the government entirely? That EA is at fault for the leading AI companies having profit motivations, and that
208.
by Veedrac 3y ago
Blame the rocket equation for this one.
209.
by Veedrac 3y ago
The Apollo Lunar Module had 6.7m³ volume. Starship HLS should have around 1000m³. Apollo LM had about a 5 ton dry mass. Starship LM would be probably well over 200. Apollo LM got around with a ton of stages. Starship LM gets around with ref
210.
by Veedrac 3y ago
Cars are great. The problem is not cars, it's bad planning for walkability and alternate transport. You can have a large, sparse, American suburb, and also have a dense, walkable, bikeable, mostly-quiet city center that feels much mo

More ›