a_bonobo

49 ms·

541.
by a_bonobo 5y ago
Sounds a bit like Australia's illegal Robodebt scheme: https://en.wikipedia.org/wiki/Robodebt_scheme The Australian system was simpler than the Dutch system, comparing social benefits paid with averaged tax-income
542.
by a_bonobo 5y ago
There could be an opportunity costs, researchers working on this could be working on a different problem. Unlikely, but not impossible.
543.
by a_bonobo 5y ago
Interestingly, hands are also something humans struggle to draw. They're a very complex anatomical form, many small tendons and muscles. Many artists struggle to depict hands. They're not made out of a few straight lines like a to
544.
by a_bonobo 5y ago
> The interesting thing is that people who multitask often feel very productive Identifying these 'feelings' is what turned my productivity around. There's a lot you can do at work that makes you feel productive - fiddli
545.
by a_bonobo 5y ago
Nature Futures, the short SF-heavy fiction run by the Nature journal, also has a word-count target of 850 to 950 https://www.nature.com/nature/articles?type=futures Rules are here: https://www.nature.com&#x2
546.
by a_bonobo 5y ago
>That hasn't been the case for at least a few decades. Authors don't really keep up with all the science or technology, for whatever reason As a biologist, I'll be a bit snipe - it's because the famous SF authors usua
547.
by a_bonobo 5y ago
Sometimes! Not for the GPU part, but to make the summary data for the ML data. For example, genome sequencing data and intermediate results are easily in the TB area of space, but the resulting table of genomic variants (k-mers like n-grams
548.
by a_bonobo 5y ago
For me, the second-biggest obstacle would be data size and correlated, data transfer speed. I regularly analyse GBs to TBs of data on supercomputers which have PBs of free space, regular consumer PCs don't have the harddrive space for
549.
by a_bonobo 5y ago
As a researcher applying ML in my work the OP reads like me looking from the outside at web-frameworks: there's a LOT of noise but in practice, people run just the same 3 things. For example, when I first got into ML the ADAM optimizer
550.
by a_bonobo 5y ago
(2010) I was wondering what Google Buzz was :)
551.
by a_bonobo 5y ago
It's quite similar to the Australian system. The most prestigious, hard -to-get award with ~10% success rate, which took me two months to write last time (I was rejected) awards on average AUSD 400k over 3 years. That's a Discover
552.
by a_bonobo 5y ago
Overheard this from HPC people: 'it's easy to detect cryptominers on the system, it's the only software that uses the nodes efficiently'
553.
by a_bonobo 5y ago
I guess the field is constantly making a decision between 'do we want to outperform humans, but lose interpretability' and 'we should always be interpretable' Human-generated artificial data will always contain the human
554.
by a_bonobo 5y ago
There's this fun Japanese crime mystery show called Galileo where a physics prof solves similar 'classic' crime story tropes. My favorite 'modern' solution to the locked room was, spoilers!, that the door with a lar
555.
by a_bonobo 5y ago
This is a great approach. How do you deal with outdated info that's still useful to keep around? I have a few internal wiki-pages which are 80% 'stuff I tried that doesn't work but I'd like to keep around because the tri
556.
by a_bonobo 5y ago
That's a lot of words for 'whose bread I eat, their song I sing'. Author didn't like ads for good reasons, now likes ads because he's making money of them, followed by mental gymnastics as to why they're good.
557.
by a_bonobo 5y ago
>which bits of the madness turn out accurate. But none of the bits in your example turned out to be accurate? Afghanistan was a disaster for a long time, and child sex traffiking rings involving the powerful have been around for a long t
558.
by a_bonobo 5y ago
I read Ann Vannetta's The Vaccinators last year, on the first days of Western-style medicine in Japan, and one of my main takeaways was how unusual these doctors were for their time and their culture, often risking political persecutio
559.
by a_bonobo 5y ago
This text glances over it but the beginnings of Western-style anatomy might be of interest to hackers: All the early Japanese surgeons had was a Dutch-language version of an anatomy book, but nobody could read medical Dutch well. The corpse
560.
by a_bonobo 5y ago
Cureus is a bit of a weird journal, they have very fast peer review time instead relying officially on post-publication peer review Here's a weird case on a different paper in the same journal: http://retractionwatch.com&#x2
561.
by a_bonobo 5y ago
I need todo lists - paper is usually enough, and easier, more in-my-face than any app can be. A single sheet on my desk. I have about 20 to 30 things to do at work on any given day, a to-do list saves me from falling down the rabbit-hole
562.
by a_bonobo 5y ago
How is this different from Zettelkasten? https://en.wikipedia.org/wiki/Zettelkasten
563.
by a_bonobo 5y ago
One other explanation is that many of these monuments didn't stand around undisturbed for 2,000 years, a lot of work of hundreds of generations of humans went into these monuments. The Porta Nigra is a great example - it's an anci
564.
by a_bonobo 5y ago
Is there any source as to why it was removed? The OP says it was because of a 'massive Twitter campaign' but I can't find no source on that. I can only find some lukewarm Reddit discussion, it seems to just be a bad essay? h
565.
by a_bonobo 5y ago
If you can, give Tim Maughan’s Infinite Detail a try. It's SF on what would happen if the Internet disappears, and goes quite deeply into collapsing supply chains.
566.
by a_bonobo 5y ago
There's actually a good incentive for moderators to be in on this, as they aren't paid by reddit for their (in sometimes massive) unpaid labour. If a marketing company comes along and says, 'hey, you've done this so long
567.
by a_bonobo 5y ago
Interestingly, Mussolini getting the trains to run on time is a myth as well. https://www.snopes.com/fact-check/loco-motive/ It's similar to how the Nazis claimed the Autobahn as their success, even though th
568.
by a_bonobo 5y ago
I've benefitted an enormous amount from the USDA soybean research. They run soybase.org, a repository of all things data around soybean cultivars. In my experience, no other crop has a database as complete as soybase. Perhaps it's
569.
by a_bonobo 5y ago
As an ECR with English as a second language, the paid version of Grammarly has clarified my writing quite a bit. I think there's more unexplored value in this space.
570.
by a_bonobo 5y ago
His son Goro Miyazaki's movies haven't been as well-received as his own movies, I guess he can't let his baby go knowing it won't do as well without him. I found this exchange between a then-retired Hayao Miyazaki and a

More ›