You’re making a lot of assumptions about what was happening behind the scenes with a payment processor agreement you weren’t privy to. Kickstarter was always and devoutly uninterested in medical fundraising, so missing out on GoFundMe’s growth was a feature, not a bug.
I’m referring to the period between 2010 and 2013. I suspect we were both privy to a lot of conversations we can’t share here. I am familiar with those processor conversations. GoFundMe didn’t pivot to primarily medical fundraising until later.
But as this is all ancient history why did Kickstarter walk back their anti-tech / anti-hardware position? At the time it felt like they really wanted to double down on artistic creators.
It's kind of wild that these tools just transfer a copy of these models every time they're spun up (whether it's to a Google Colab notebook or a local machine.)
This must mean Hugging Face's bandwidth bill must be crazy, or am I missing something (maybe they have a peering agreement? heavily caching things?)
Their Python module caches the downloads, which is checked before downloading them again...but you're probably not wrong on the crazy bandwidth bill. Looks like they have crazy VC money though, considering the current climate.
Unmetered 10+ gigabit connections were on the order of $1/mbit/mo wholesale over a decade ago when I priced out a custom CDN so for the cost of 100 TB of data transfer out of AWS you could get a 24/7 sustained 10gbit/s (>3 PB per month at 100% utilization).
Not all connections are created equal. Even some big providers clearly have iffy peering agreements upstream that’ll manifest as terrible performance if you have a widely-geographically-distributed bandwidth-heavy load.
That’s pretty expensive. Sonic offers 1-10gbps (depending on where you live) unmetered symmetric connections for $60/mo to the Bay Area… they’re also the only ISP that petitioned the FCC in favor of net neutrality.
For work I end up transferring 50-150 gigs often, sometimes daily. Never heard a word from them that this has been a problem.
If you host copies of your data with a few big providers could you do something smart like detect and redirect requests from AWS to an S3 bucket and not pay for bandwidth leaving the provider?
1. AWS is far behind Azure and GCP in AI, so they gotta partner up to gain credibility.
2. Huggingface probably does face insane bills compared to github. But AWS can probably develop some optimizations to save bandwidth costs. There's 100% some sort of generalized differential storage method being developed for AI models.
Is hugging face just a model repository like GitHub is a code repository? Seems you can rent compute both cpu & gpu, but you are right that most models seem to be run elsewhere.
I haven’t used windows in a while but I thought it supported some form of cross-volume symlink? Or at least mounting an image stored on another volume to an arbitrary path.
So not-well-known that several tools that really should know better don't check for junctions with occasionally disastrous results in a fs walk. (Using junctions sounded really clever to me until this had me up all night figuring out why the backup system crashed.)
Thanks – I'm the one that picked The Nix for the list (I work for YC on their admissions team).
One reason I picked it is that is indeed a great book that I have been recommending plenty to friends.
Another is that I do think it's incredibly important to be able to enjoy and indulge and make time for fiction, especially when you work in tech. It's too easy to "read for facts / information / theory" and forget the joy that good writing can bring.
It took me some reflection to realize I heavily bias towards non-fiction unless I make an effort. But when I do get the chance to read a great novel, it tends to affect me more than reading non-fiction.
The world needs more literature, poetry, and art, so I also like supporting authors who ship that product too :)
(FWIW, the book probably isn't for everyone, but I think reading great literature is – give it a shot if you haven't in a while!)
Artists have depended on public support for generations (if not longer) and that support does not artificially prop up its value. In some cases artists and their work can languish for decades until publicly supported exhibitions choose to highlight them. Only then does "the market" value their work, and only sometimes does it do it properly.
And that's not to mention the value art plays in civic life and its contribution to society: markets can frequently fail to properly support artists when society needs them the most (during political strife or civil unrest.)
Author of Emoji Dick here. The $200 price tag is that high because its a one-off print-on-demand of an 800 page color laser hardcover book. I would LOVE to find a way to produce these cheaper (even do another edition) but haven't had the time to get a publisher / deal / etc. together.
I never really seriously considered it. Made the book and the page before this was really all that common online. Maybe I'll A/B test it to see if it generates more sales :)
Since emojis often bear graphic resemblances to our real faces, the understanding has often been that there would be no problems in interpreting them, and that the sender and the recipient would agree on such interpretation.
As someone fairly immersed in the emoji community, this is a strawman argument (i.e., no one really tries to argue this).
People love and use emoji not in spite of their ambiguity but rather because of it.
Even Unicode encourages emoji to have multiple meanings:
A: No. Because emoji characters are treated as pictographs, they are encoded in Unicode based primarily on their general appearance, not on an intended semantic.
Many people want to think there are some folks out there like myself who are seriously arguing Emoji are a language, but this isn't really true. And I say that as the author of a book called "How to Speak Emoji". The thing is, it's a humor book designed to be sold in Urban Outfitters. It's not a real language guide.
If you're curious about more nuanced takes on how emoji are actually being used, here are some good resources:
The tl;dr: journalists / bloggers would love to get someone to argue that emoji are a language so they can "Well, actually" them, but the truth is this isn't really happening much.
However, some of us are deeply curious about whether our usage of emoji are evolving language-like characteristics and grammars. See this recent research on whether emoji have their own syntax:
Hi there! I was Kickstarter employee #2 and my unofficial (which briefly became my official title) was "R&D", so I think it was just a matter of being in the right place at the right time and for me, realizing that the most effective way I could be valuable for our small team was to dig into the data and do the research to answer all the questions popping up day to day. Sometimes that was writing queries for our Year in Review posts, other times it was figuring out the best ways to mitigate payments risk.
At first we were just using Google Analytics and SQL. A couple years into it I deployed Mixpanel, and eventually we built our entire analytics infrastructure from scratch and used Looker to expose it to the rest of the company.
I was at the company for over 6 years and by the time I left, my data team had people working on data science, machine learning, traditional BI, and of course infrastructure.
