back
user profile

moultano

11,023karma·1,963submissions·May 20, 2009
about
I work at Google trying to make your search results not suck. rmoulton@google.com

I write articles here: http://moultano.wordpress.com/

recent activity (1,963 total)
comment
I did not take Jonathan Blow's tweet to be saying it's a bad idea, or won't work, but rather that it's an indictment of the whole web stack that it's necessary.
5y ago·view thread
comment
It entirely depends on how much CO2 and methane we emit, but these sorts of predictions are by convention pegged to 2100. Most of the emissions pathways, unless we stabilize things soon, keep getting …
5y ago·view thread
comment
A rough rule of thumb is that for every degree global temperatures increase temperatures on land will go up twice that amount. So with 3.5°C, which is on the low end of what we're headed for, ima…
5y ago·view thread
comment
The fact that over parametrized models can generalize better than under parameterized models, and that both are better than models that can just barely interpolate is a genuinely new insight that was …
5y ago·view thread
comment
If the price will bear it that means people want to live there more than elsewhere. Don't impose your idea of dystopia on other people.
5y ago·view thread
comment
NYC has one of the lowest rates of housing starts per capita in the US.
5y ago·view thread
comment
Tip the artists you love. If you want you can send them money by paying for prints, or buying originals, but ultimately just send them money. There's no need for it to be transactional. If there …
5y ago·view thread
comment
HeLa would be a better title. https://en.wikipedia.org/wiki/HeLa Copying the remains of a human around with ambiguous ethics, largely because they're "standard" an…
5y ago·view thread
comment
Seems like if you're going to write about something a lot of people find a lot of value in, you at least owe it to your readership to explain why they find it valuable.
5y ago·view thread
comment
> The unceremonious dumping of A Prairie Home Companion, which wasn't exactly my schtick but provided a little variety at least, made their coastal preference clear. WHaaa? Garrison Keillor h…
5y ago·view thread
comment
The guy is a troll. The last nuclear power plant built in the US took 40 years to come fully online and has never made money. Meanwhile solar is going through a moore's law like cost reduction an…
5y ago·view thread
comment
A monorepo does not let you deduce the revenue consequences of experiments, but regardless, even if you could figure it out, it isn't part of the incentive structure.
5y ago·view thread
comment
> If a search change demonstrably has a negative effect on company earnings (even if users like it) it will almost certainly be reversed. This is completely false. I've personally been involv…
5y ago·view thread
comment
This is why there's an aggressive firewall between the ads and search organizations. The ranking team isn't even aware of revenue numbers by ACL.
5y ago·view thread
comment
https://www.techdirt.com/articles/20200531/23325444617/hello... …
5y ago·view thread
comment
If I'm understanding correctly, it doesn't just set the kernel after seeing the data, but also after training the entire model, because the path kernel can't be defined without the opti…
5y ago·view thread
comment
That the difficulty of finding the video in YouTube search was a deliberate change made at the request of the CIA.
5y ago·view thread
comment
That sounds more like doing challenge trials for an ebola vaccine, for which I'd expect you'd get very few volunteers. Still though, if there are people willing to risk sacrificing themselve…
5y ago·view thread
comment
A traditional trial requires you to wait for the pandemic to spread in order to get enough positive people in your trial for significance. It's effectively like "challenging" random peo…
5y ago·view thread
comment
People join the military not knowing whether or not they will ever go to war, yet they still join. By doing so they risk death, dismemberment, and all manner of long term disability under the belief t…
5y ago·view thread
comment
Even once safety is established, our current protocol requires waiting for wave of infections to sweep so thoroughly through the population that some of the study volunteers are infected by chance.
5y ago·view thread
comment
The point that challenge trials don't determine safety is accurate, but most of the article is again arguing about the ethics rather than the efficacy. Imagine if volunteers were not permitted by…
5y ago·view thread