pacala

59 ms·

541.
by pacala 9y ago
I see your point. For k8s production, you'd have to use a solution provided by the cloud operator, GoogleKE / AzureKS / AmazonKS / etc. Which leaves the on-prem cluster and/or baremetal hosted cluster uncovered. Not
542.
by pacala 9y ago
GKE has been around for years.
543.
by pacala 9y ago
$ brew cask install minikube $ minikube start
544.
by pacala 9y ago
Fixed, thanks!
545.
by pacala 9y ago
OCEAN stands for Openness, Conscientiousness, Extroversion, Agreeableness and Neuroticism. Also known as https://en.wikipedia.org/wiki/Big_Five_personality_traits .
546.
by pacala 9y ago
The NYTimes article you link to doesn't contain the word "threat". This piece appears to contain a new bit of information about the ethically challenged behavior of FB and CA. It deserves a separate conversation.
547.
by pacala 9y ago
> [Spammers] don't take it seriously because people don't actually care. Some do. The vast majority don't. They might say so in a survey, but at the end of the day, Facebook (and companies like it) will continue to survive
548.
by pacala 9y ago
> Next up, browser support? Please! There is a working TypeScript client implementation [0] of gRPC-Web [1], which relies on a custom proxy for converting gRPC to gRPC-Web [2]. Would be nice to bring that proxy functionality into Nginx.
549.
by pacala 9y ago
One can do semi-supervised training with LSTMs, using http://allennlp.org/elmo or similar. That may use huge amounts of data, for which MaxEnt or CRF are not sufficient.
550.
by pacala 9y ago
Pinch of salt. The data for electrical engineering is conspicuously missing. Furthermore, the article mentions two universities, Carnegie Mellon University and Johns Hopkins University. The CMU CS department was founded in 1988 [0] and JHU
551.
by pacala 9y ago
In the current climate, I'd suggest approaching such data with a pinch of salt. Historically, computer work originated in Electrical Engineering field. Dedicated Computer Science departments appeared later on. It is entirely possible t
552.
by pacala 9y ago
Just buy an Alexa / GoogleHome / Nest / SamsungSmartTV and wave your privacy goodbye. It's a free-for-all out there for invading your privacy, and the big players are in the lead. No need for obscure knockoffs.
553.
by pacala 9y ago
Have somebody done it for the more general foundations based on category theory? If I understand correctly, Metamath is a set of tools enabling such a formulation, but not the formulation itself. See ML vs. theorem prover implemented in M
554.
by pacala 9y ago
Dilettante here: What's interesting about *TT is that people have built mechanical theorem checkers [Coq, Agda, Lean, ...], and people are working on building automated theorem provers based on those programatic foundations.
555.
by pacala 9y ago
Therein lies the problem. A given piece of code looks the same for any possible monad, but it only provides useful functionality for a specific monad. The reader of the code is left wondering _for what benefit_ the code was written, and may
556.
by pacala 9y ago
String AI will have a soul.
557.
by pacala 9y ago
> as a group women have decided that the best way to make themselves more welcome in STEM I missed when the voting took place. You mean a group of activist women (and men) assumed to talk on behalf of all women in the field?
558.
by pacala 9y ago
The poll is biased. There is no evidence that performance and explainability are at odds.
559.
by pacala 9y ago
Paranoid fantasies about 42% of the population being lured into smoking and millions of lives lost to easily preventable disease.
560.
by pacala 9y ago
Thanks for taking the time to write this. I’m confused on point 13, would "RUN wget tarball_url" work?
561.
by pacala 9y ago
It's a search space. Think of people "above" as being more likely to know which direction to search next, and people "below" as being less likely to have a clue. Ignore people "above" at your own peril, fa
562.
by pacala 9y ago
From the other side of the interview table, how do you tell apart engineers from bulls__t masters, who take credit for their team's work without ever having a positive engineering contribution?
563.
by pacala 9y ago
Not sure about "will surpass". Possibly the inflection point already happened. https://www.theatlantic.com/technology/archive/2017/02/china... > The 2017 AAAI meeting—which ultimately reloca
564.
by pacala 9y ago
That is a solid solution. As long as everyone in the ecosystem is on JVM, more power to you. If, for example, one needs to interface with some cutting edge DL modules written in Python, one needs something else. The transitive closure over
565.
by pacala 9y ago
Indeed, Kubernetes as a service is the way to go. https://cloud.google.com/kubernetes-engine/ https://azure.microsoft.com/en-us/services/container-service... https://aws.amazon.com
566.
by pacala 9y ago
Containerization helps with one thing: end-to-end dependency hell management. You get the same executable artifact in prod and on every dev machine. You get to share arcane tricks required to bootstrap library X. You get to share the comple
567.
by pacala 9y ago
Maybe you have a JSON/YAML/XML config file that, for whatever reason, is not compiled within the single Go binary.
568.
by pacala 9y ago
Accepted answer: > This looks very much as if the units may have been converted to obtain the desired result. Indeed the different conversion factors but similar result in your second link increase the likelihood that there may cognitive
569.
by pacala 9y ago
While being correct is a great start, it is necessary to also explain why one is correct. PS. Deep Learning aficionados, take note.
570.
by pacala 9y ago
What if I have data-dependent computational graphs? For example, recursive NNs.

More ›