HNHacker News
TopNewBestAskShowJobs

anabranch

278 karma · joined July 3, 2014

Graduate student at UC Berkeley studying machine learning, user experience. Interested in entrepreneurship and technology.
submissionscomments
anabranch··on Opus 4.7 to 4.6 Inflation is ~45%
I wanted to better understand the potential impact for the tokenizer change from 4.6 and 4.7.

I'm surprised that it's 45%. Might go down (?) with longer context answers but still surprising. It can be more than 2x for small prompts.

anabranch··on Python 3 makes it easier to develop high-quality software
Probably the same reason (or at least somewhat related to) that Google App Engine only supports python 2.7. (https://cloud.google.com/appengine/docs/python/)
anabranch··on Analyzing Flight Data: A Gentle Introduction to Spark's GraphX
Any feedback anyone has would be greatly appreciated! Just getting this site started to teach people about the basics of Apache Spark.
anabranch··on Brunch: Replace gulp, grunt and increase your dev speed
Welcome to hell.

The relentless creation of new javascript frameworks never ceases to amaze. They're all over the place and every one of them claims to be a magic bullet.

My recommendation would be to choose one that is extremely stable that underlies a lot of the core principles of other systems. Backbone comes to mind because it's fairly unopinionated about how the system should be built - this makes you understand how data can be tied together, how systems can be structured (because you have to make your own choices) and which ones seems to make the most sense to you. Couple that with something like react and I, personally, think that you've covered a lot of the front-end principles.

http://yeoman.io/ is a popular flow that's much more plug and play than writing your own gruntfiles/gulpfiles.

anabranch··on Changes to Linkedin API Developer Program
You will still be able to login with linkedin: https://developer.linkedin.com/docs/signin-with-linkedin

They just rehashed the website @ developer.linkedin.com

anabranch··on Changes to Linkedin API Developer Program
Is this legal to scrape linkedin? What's the legality of obtaining + selling this information?
anabranch··on Ask HN: How can I remedy scatter brain and information overload?
http://selfcontrolapp.com/

Works well. It cannot be turned off and you can set it for a specific amount of time. You choose the websites you would like to block.

Your phone is obviously a weak point, but it's a start.

anabranch··on Show HN: Quiver – The Programmer's Notebook
I made a quick and dirty script to convert markdown to quiver already but unfortunately have a ton on my plate right now.

https://gist.github.com/anabranch/09c421e3ff13e1245316

This basically just splits it on ```code``` blocks and only supports python at the moment, but could be a resource for others.

anabranch··on Show HN: Quiver – The Programmer's Notebook
A way to convert IPython Notebooks to Quiver notes would certainly be useful (as would the other way around). Searching ipython notebooks for code is a nightmare.

edit: good work btw, I'm not completely sold on this solution but I do think there's a problem here. I'm pretty invested in evernote but think this has potential.

anabranch··on The Right to Privacy (1890)
You're touching on this but the importance of tangible vs intangible property cannot be understated. In Katz v United States: "Because the Fourth Amendment protects people, rather than places, its reach cannot turn on the presence or absence of a physical intrusion into any given enclosure. " https://supreme.justia.com/cases/federal/us/389/347/

That sets a good precedent but gets destroyed in Smith v Maryland

"Petitioner in all probability entertained no actual expectation of privacy in the phone numbers he dialed, and even if he did, his expectation was not "legitimate." First, it is doubtful that telephone users in general have any expectation of privacy regarding the numbers they dial, since they typically know that they must convey phone numbers to the telephone company and that the company has facilities for recording this information and does, in fact, record it for various legitimate business purposes. And petitioner did not demonstrate an expectation of privacy merely by using his home phone, rather than some other phone, since his conduct, although perhaps calculated to keep the contents of his conversation private, was not calculated to preserve the privacy of the number he dialed. Second, even if petitioner did harbor some subjective expectation of privacy, this expectation was not one that society is prepared to recognize as "reasonable." When petitioner voluntarily conveyed numerical information to the phone company and "exposed" that information to its equipment in the normal course of business, he assumed the risk that the company would reveal the information" https://supreme.justia.com/cases/federal/us/442/735/case.htm...

That's where things get fucked up.

Simply because you use a service means that you assume the risk that the company would reveal that information. Privacy is not guaranteed as soon as you use a service - setting the precedence for NSA mass surveillance. Unfortunately, these laws were passed when the fact that everyone you know, interact with, and communicate with is in your pocket all the time.

Different laws for different times...

anabranch··on The Right to Privacy (1890)
Recent inventions and business methods call attention to the next step which must be taken for the protection of the person, and for securing to the individual what Judge Cooley calls the right "to be let alone."

Instantaneous photographs and newspaper enterprise have invaded the sacred precincts of private and domestic life; and numerous mechanical devices threaten to make good the prediction that "what is whispered in the closet shall be proclaimed from the house-tops."

124 years later, still relevant.