72
points
chjohasbrouck
Joined 606 karma
- Haha, fair enough. :)
- Some proof or data to back up the article's claim would be great. I'm not really buying it.
If moms auto-like every post, then how is that a relevant signal? Everyone has a mom. That would mean every post is getting penalized in the same way (which effectively means no posts are getting penalized).
And if circumventing this was as simple as excluding his mom, wouldn't the effect be even greater if he excluded all non-technical friends and family?
Which pretty much just means you're posting this for the greater public, which presumably a lot of users of Facebook's API already do. Since his intention is for his content to be seen by the greater public, then... go ahead and tell the API that?
It's a great angle for an article, and it's very shareable, but he provides no data (even though he seems like someone who would have all of the data).
- That's interesting, but in that scenario, how important is the composer really?
Couldn't we just connect the variable output of the AI to some kind of Mechanical Turk-style focus group, feed those results as inputs back into the AI, and repeat that process millions of times? Might that produce a better song than an individual composer?
I've seen interviews with some modern artists, and they'll often take inspiration from decades-old hits (which are essentially de-risked melodies) and apply them to a modern template. I'm assuming the bigger labels already focus-group the output and make revisions. This seems like something that could eventually be automated.
- > While true ...
This just broke my brain.
I think it's time for me to go to bed...
- Is this possible? It doesn't seem like the necessary information is captured in a picture.
I don't think there's a person that could tell you whether food in a picture has allergens, let alone an app.
- 3 points
- Maybe this has been said already, but one major problem with Twitter is that by default, the average person will publish and share content on it that's a complete waste of time and of no value to anybody.
The average unit of content shared by a user on Instagram or Snapchat is vastly more entertaining and relevant, and I think they achieve that by being more visual. Those platforms are image- and video-first, and they have all kinds of image and video filters. If we gave those tools to a monkey, it would produce something worth sharing.
The quality of their content is enforced inherently by the tools they provide you to share with.
Facebook achieves the same in a different way, by tightly integrating with your identity and social circles.
Twitter doesn't have any form of inherent quality control. Tweets are just words with little context. The people tweeting are usually strangers.
- Risk is one of the things that debt can represent, sometimes, but there can be debt that doesn't involve any risk, and there are risks that don't involve debt.
What's risky about using a credit card to buy a kayak?
And how does investing money you already have in small cap equities represent a debt?
I think for the vast majority of individuals just trying to make good personal finance decisions, "don't take on debt" is (on average) great advice.
- One caveat to this kind of thinking is that money is fungible.
One can easily pay for a vacation or years of eating at nice restaurants every day all in cash, and also go into debt for an education.
Also, one individual taking out an auto loan and a loan from Sallie Mae isn't 1 good loan and 1 bad loan. You could've foregone the high-interest loan on the big depreciating asset and used the savings to fund the education instead of the loan from Sallie Mae. In that scenario both loans were probably a bad idea, regardless of the return the education nets you.
I think another key metric beyond interest rate and rate of return, is degree of necessity. That's harder to measure though.
- I'd be surprised if this house meets state regulations for residential construction in any US state.
That'll be one of the biggest obstacles to 3D printed housing in the US. Even if you somehow get efficiency gains through 3D printing, it's going to require a different configuration depending on the regulatory environment, which varies by nation and state and county and city and topography, and changes every year.
These regulations affect every detail of the construction of your house, from the foundation to the window panes. Even details as innocuous as sink depth are regulated.
- Completely agree.
I take this one step further and say tablets are probably handicapping your kids in the home.
Having no tablets or smartphones available during the PC revolution forced most of us here to default to exactly the correct set of tools for maximum learning and efficiency: a keyboard, a mouse, a full-size monitor, sitting at a desk, with a full computer with some kind of (more-or-less) exposed desktop operating system.
This isn't a choice any of us made, it was just the climate and the time. That lucky coincidence has given most of us a lifetime of experience using these tools, extending back all the way into our childhood.
Over the last 10 years though, a lot of people/families have started replacing their desktops/laptops in favor of tablets/smartphones.
I can't even imagine how I would've become a programmer if I had grown up in a house that just used tablets and smartphones for everything.
They say that one of Bill Gates' greatest advantages in life was early access to the PDP-10.
I think this trend toward "convenience computing" is especially bad for lower-income families (which are more likely to rely entirely on smartphones/tablets), and could even be a significant contributor to income inequality.
- I think this is an example of something machines can't quite do better than humans yet, though it's an interesting problem to solve (trying to analyze how people in general will feel about something).
Personally I'd give it to either How to Disappear Completely, or Exit Music (For a Film).
- This is why no matter how much sense it makes to use something like vim or emacs as your primary development editor, I could never justify the amount of time it would take me to actually be totally comfortable and start seeing benefits (beyond just using it for small quick edits at the command line).
Typing speed is never the bottleneck. You could double most developers' typing speed when they code and it'd probably have a negligible impact on their overall productivity.
There are benefits, but do they really outweigh the benefits of those same hours spent studying a data structure or design pattern or trying new tech?
- > Urgh. It has been several years now since I last had to deal with Magento, and nothing I've seen over the years suggests it has got any better.
One of their biggest vendors, WebShopApps, happened to write a pretty scathing open letter to Magento yesterday if you're interested:
http://webshopapps.com/blog/2017/01/what-magento-should-spen...
This is a company that just a year ago was making almost all of its revenue off of the Magento platform, and they are one of the most-respected brands in that community. Not a good sign for Magento.
- 12 points
- I had intimate knowledge of an Act of Congress and some DOJ seizures that resulted from it years ago, and that's the first time I really experienced what you just described. The basic inaccuracy and sensational bias in the reporting and the level of discourse it generated was truly eye-opening, and I'll never see the news (or people reacting to the news) the same way again.
Every time they cover a subject I know well, it's been the same story, even among major reputable news outlets. No facts are safe, not even basic high-level details.
- > it's a different system from the server, lots of tricks are needed to make it work invisibly (either with a VM or a lighter equivalent)
I used to use Linux for this reason, but it's no longer a factor in my opinion.
Ideally your development environment should mirror the production environment, and just saying "the server is running Ubuntu Server 14.04LTS and I'm running Ubuntu 14.04LTS" is a pretty lazy and incomplete way of half-achieving that goal. What about your nginx config? What about your permissions? What about your PostgreSQL version?
Whether you run Mac or Linux, you're still forced to solve that problem, and the best tools to do it are all available and work great on a Mac. Vagrant is relatively painless and works really well.
- I think the market's expectation is that a DNS provider is prepared for a DDoS of any size, but not necessarily any level of complexity, so that's a lot of incentive to talk up the complexity of the attack.
What's described in this incident report is totally within the capabilities of a single individual with public knowledge, though. If they could have proven otherwise, they probably would have (unless that somehow conflicted with their criminal investigation).
- He didn't get a 250k salary, that was total compensation, largely RSUs on a standard vesting schedule. He also donates 1/3rd of his pre-tax income to charity.
He wasn't just a graduate of a hacker bootcamp, he was hired on by App Academy after he finished and was quickly promoted up to Director of Product. He spent a lot of time teaching people how to pass whiteboard interviews. He was probably better prepared for an interview than your average Stanford CS grad, and already had prior experience as Director of Product and a portfolio of side projects.
He didn't bluff any company. He went through Triple Byte, the trendiest new tech recruitment startup. He got an offer from Triple Byte themselves even though he didn't even apply for them, got an offer from Google, and when word about the Google offer got out every other company started bidding for him.
He's also just a really smart guy in general. If you spend 15 minutes talking to him, he'll leave an impression. It's not a bluff or a con, he's just a really smart, really analytical guy.
- Another interesting way to look at this is to look at what level of income equality people are advocating for, on a macro scale.
Most people making arguments against income inequality in the US seem to want income to be more equal down to and including themselves, but never very far below themselves. They want it to be equal within the US, but I have a hard time believing they'd be as excited to extend that same income equality worldwide. Doing so would see even the poorest Americans taking a massive cut in lifestyle and living at a level of poverty most people in this country have never seen or experienced.
- I've always thought that maybe they optimize their designs specifically for A/B testability.
If everything is just kind of flat and grey, there's less noise, and it's easier to write A/B tests that produce meaningful results. That also means every product is flat and grey, though.
It could also just be a side-effect of having to appeal to every person on the planet, though. Sort of the design equivalent of a good politician, appealing to the lowest common denominator in all of us. Nothing to object to, but also nothing to love. Just flat and grey, with clearly displayed content and A/B optimized calls to action.
I find it incredibly boring and I'm getting tired of them rolling out new products that all look and feel exactly the same, but I suspect they have numbers that show any big design change would just be leaving money on the table.
- Those are just two product categories well-known for their manipulation of the nutrition facts label, and there are definitely millions of consumers that think they confer health benefits as a result of how the nutrition facts label was manipulated.
Most people think VitaminWater is healthy, and feel like they're doing something good for their body when they drink it. They'd probably be more discerning and come to a better conclusion about that product if you completely removed its nutrition facts label.
- One problem with products like Soylent is that the nutrition label is used for marketing, and as a result they consciously engineer the product to optimize for that.
Nutrition facts labels and recommended daily intakes vary by country, and they're backed up by a lot of bad science and guessing games.
Consider that the USDA recently lifted their recommended maximum intake of cholesterol from 300mg/day (1.5 eggs/day) to no upper limit at all (after 50 years), or that study after study has failed to find any significant benefit from taking daily multivitamins (and to the contrary has found that taking too many, or the wrong kind, can cause liver damage), or that every sugary breakfast cereal and energy drink would seem quite healthy if all you looked at was the nutrition facts label.
I don't think nutrition facts labels are particularly meaningful, so a product sold almost entirely on the basis of how strictly it conforms to those government recommendations isn't of much interest to me.
If you're Soylent, you also have to lean heavily on government recommendations as a form of insurance against lawsuits. There's a lot of risk in encouraging people to eat only your food for every meal of the day, and being able to tell a judge that your product closely adhered to government recommendations is undoubtedly helpful in various litigation scenarios, but I don't think that necessarily forms a good foundation for what people should be eating.
Most people just don't like doing the shopping, cooking, and cleaning involved with daily meal preparation. Most people also don't like cleaning in general, and hate their commutes. Thinking of these as therapeutic tasks is as much a luxury of white collar life as unlimited access to maids and food delivery services is.
And the idea that you'd save time or money with proper knowledge of how to source and prep your own fruits and vegetables, or that time spent cleaning disappears because you consider it exercise, all seems like magical thinking to me.
As soon as you place a realistic dollar figure on the value of your time, for most of us, all of those calculations break down dramatically in favor of outsourcing most tasks related to food, cleaning, and transportation, not only as a matter of comfort, but also as a matter of economics.
It makes sense that it would work out this way, too. A single Uber driver with a single car can satisfy the transportation needs of a dozen people. An Instacart shopper can gather everything 3 families need in a single pass through the store. The driver can get those groceries to 3 homes in a single round trip instead of 3 separate round trips.
Outsourcing can seem wasteful and opulent, but in doing so you're often participating in a sharing economy or benefiting from economies of scale and division of labor that actually serve to reduce waste and inefficiency.