News for hackers. Stuff that matters.

altairiumblue

There have been many attempts for interventions that increase IQ with little to no evidence that it's at all possible.
Re:Human Contact Is Now a Luxury Good on Sunday March 24, 2019 @08:45AM (#19476024)
You really don't. That doesn't mean that it doesn't exist or that it should be ignored
Re:Petition: Revoke Article 50 and remain in the EU on Thursday March 21, 2019 @03:21AM (#19449808)
In general, you can't take a strictly advisory referendum with a win margin of a couple % to dictate your international policy for decades.

More specifically, when the winning side has committed campaign violations and the UK's own National Crime Agency has found illegal foreign interference, maybe you should investigate further instead of driving head on. If you ask me, this was the "assault on democracy", not the "do-over" approach as you wrote.

On do-overs - the Prime Minister is pushing to have a third vote on the exact same deal with the EU, after her proposal was rejected twice by the members in the course of a couple of weeks. So it's especially concerning when do-overs in parliament can happen until the government gets the desired result, while at the same time the people aren't given a chance to vote again and the outcome of the referendum is considered final.

Re:Reasons Not to Use Apple (2018) on Saturday March 16, 2019 @08:34AM (#19407513)
Their incentives with regards to data are aligned differently - Apple make most of their revenue from selling hardware to their users, Google make most of their revenue from selling ads to someone different from their users.
What do I see differently - just about anything.

What would I tell myself? I wouldn't bother at all. Any advice that I can tell to my younger self was probably also available at the time. And I was really bad at listening to it. Sometimes you have to earn your own experience - other people's advice only goes so far.

Personal machine - Apple Mail - it's simple and it looks good, I don't need anything else.

At work - Outlook - also used by everyone else at the company for email + sharing calendars and scheduling meetings.

I never use the web UIs.

Re:The head of luxury at Facebook and Instagram on Sunday March 10, 2019 @08:43AM (#19351706)
Personally, I don't care for luxury items or brands.

But your argument here is a mess - you despise people with passion because of their consumer choices... and you're mixing up consumerism, environment, politics and income inequality into something completely incoherent.

Re:Ask HN: What podcasts do you listen to regularly? on Saturday March 09, 2019 @03:01PM (#19348140)
Different opinion - I have some interest in the topics he covers but I cannot listen to someone speak with such indifference in their voice for long periods of time. I've tried to listen to his podcast a few times and have never been able to get past the monotone delivery.
Rebranding according to what's in demand is all fine, but if you choose data science please make sure that you actually have a solid understanding of statistics and software development. Because ultimately that's what data science is.

"Hugely credible" and actual experience in data analysis/statistics are completely different things. And if you don't have the statistics/math background, you'll find out that now instead of just programming, you need to learn programming and statistics. Which might be too much for someone who doesn't know where to start.

Re:Ask HN: How `private` will the Oculus Quest be? on Saturday March 09, 2019 @11:03AM (#19346739)
Why would it be private?

At this point this isn't just a cynical position, it has to be your default if you've watched the privacy problems that every single facebook product has had in the past and is still having at the moment. Why would a company that is still breaking a number of ethical and legal norms suddenly decide to abide by any reasonable principles of decency? If information about your device and actions on a social platform is worth billions, what's the worth of full access to people's sensory information and behaviour?

Math is huge so how you approach this will depend a lot on your starting point and your goal.

Which branch of math are you interested in?

Do you need it for something specific to your career or is it just a hobby?

What's you current level - are you comfortable with high-school level algebra, trigonometry, calculus?

In any case, I second the recommendation for Khan Academy as a good starting point - not only for the range of content and Sal's silky voice, but also for their approach to learning: you get to practice topics until you "master" them instead of just watching/reading, maybe taking a test and then moving on to the next topic.

Re:A Privacy-Focused Vision for Social Networking on Wednesday March 06, 2019 @02:35PM (#19321916)
When I saw the headline, at first I thought it was an old article that you can read in a different light and have a laugh.

Writing it now is just obnoxious. A lot of text with no meaningful content, mentioning a bunch of problems without offering solutions and without taking any responsibility.

Don't talk about "people prefer", "people expect", "working towards" and "we as a society" - you created this. With all of its problems, and all of the ethical and legal violations. You did things this way because it was extremely profitable and you are still doing it.

How does anyone take him seriously at this point?

Re:Ask HN: What software products do you find expensive? on Tuesday March 05, 2019 @06:35AM (#19309839)
Depending on your use case - numpy, pandas, scipy, matplotlib, statsmodels. Also check out R - you may or may not prefer it to Python, especially if you're coming from Matlab.
Re:Ask HN: What software products do you find expensive? on Tuesday March 05, 2019 @04:38AM (#19309357)
Alteryx - limited functionality, doesn't scale, can't debug properly. But they have a great sales team I suppose.

Also SAS.

It amazes me how much money some companies are willing to spend on products that are worse than their open source alternatives.

All of my books are in pdf and receive zero maintenance - many of them are still incredibly useful.
Any discussion on blockchain AI has to also include a serverless big data solution as a quantum cloud microservice. Otherwise you're just leaving money on the table.
Re:Listening to music 'significantly impairs' creativity on Sunday March 03, 2019 @02:42PM (#19296358)
I think this is because you expect creativity to mean something resembling artistic ability (e.g. painting), while the paper's definition is closer to problem solving (performance on CRATs).
To be more accurate, swap "global warming" with "climate change" and see if your original argument about the double standard holds up.
Re:Xkcd-Style Plots in Matplotlib (2012) on Sunday March 03, 2019 @04:26AM (#19293584)
Jake is a great teacher - I would definitely recommend his book to people new to Python for data science, as well as any talk by him that you can find on youtube.

Also, it's kind of a shame that matplotlib is still so deeply ingrained in the Python data ecosystem.

Yes, this always struck me as strange and at the time my naive thought was that Apple was overvalued.

https://www.youtube.com/watch?v=lTd7Iax-I60

What he says around 00:27 about Snickers applies very much to Apple as a brand. This is the only Buffet-ish explanation of why Berkshire would invest in Apple that I could think of.

And apart from this, the decision to buy a lot of Apple stock a while back was made by other asset managers, not Warren himself, and I'm sure was the result of thorough analysis, not based on some offhand comment in an interview.

I've been wondering about this - why would you pass data from one language to the other? If you already have existing R code, you can easily manage it through R itself.
- I really don't think that Python needs an additional layer of abstraction, especially when using scikit-learn which is high level enough as it is.

- The workflow in the readme is missing the part when you actually use the model. This will often need to be connected to your original preprocessing in some way - for example, if the dataset that you're predicting on has a categorical variable with a unique value which wasn't present in the training dataset, this effectively introduces a new feature in your dataset, which makes it impossible to do model.predict(). The need to manage things like this changes that workflow chart quite a bit.

The network effects of everyone being on Facebook/Twitter/Instagram are huge. You can't expect your contacts to move to your network of choice - even if they set up an account, their behaviour and the content they post will be very different. Most likely you will end up connected to a handful of inactive accounts.

What I personally do - use signal/text/email (I wouldn't bother with hosting my own email server) for 1-to-1 communication; no accounts on Facebook/Twitter/Instagram/Snapchat - they're just not worth the time and attention. Neither are Mastodon or Diaspora. I have an up-to-date LinkedIn account that I don't really use - again, worth having an account but the feed is a giant waste of time.

Re:Amazon Personalize: Real-Time Recommendation on Wednesday February 27, 2019 @02:44AM (#19261645)
They have all the data. They should be able to do better than this.

Older »