danblick

Something like this:

http://www.webmd.com/food-recipes/news/20130710/could-artifi...

"After a while, Swithers said, it's like the mouth keeps crying wolf, and the brain and gut stop listening. As a result, when real sugar and real calories come along, the body doesn't respond to them as strongly as it normally might. Calories don't end up making you feel as full as they should. They aren't as rewarding. So you don't get the signals that might stop you from eating when you should."

It's been a while since I learned the scientific method per se, but I remember something about "repeatable, controlled experiments". There are many, many questions in life just can't be studied in a repeatable, controlled experiment. Macroeconomic policy comes to mind. (You might reason, "okay, I'm going to make two Earths, and in the US on planet 1 I'm going to implement policy A, then in the US on planet 2 I'm going to implement policy B, then I'm going to travel back in time to the present day so we have identical initial conditions and implement the better policy.") Anything single-shot. Anything involving a counter-factual ("if Hitler had never been born, would there have been a second world war?"). Anything where confounding variables cannot be eliminated or controlled. I could continue but I think the point is probably clear. :)

I'm not suggesting we ignore evidence; I'm just saying we should be honest about what the evidence really shows and also honest about our inability to collect conclusive evidence at times.

I agree that mouse studies can show that megadoses of these sugar substitutes aren't lethal and probably won't cause cancer. But humans aren't mice and we don't live in lab cages, so animal studies don't answer the question I'm really interested in (~"will accepting artificial sweeteners into my diet improve my health?"). What would be the right experiment to answer that question? Has that experiment actually been performed, and would it be practical? Changes in human behavior might be hard to observe and quantify, and there are always practical limits to sample sizes.

Sometimes it's​ very, very hard to design a study that actually answers the question you're interested in, but neither the scientists nor anyone else really has an incentive to be humble about that. (Most scientists want to make their work seem impactful.)

Worse, it would mean violating my Microsoft boycott. ;)

Perhaps I'm just eccentric, but I avoid artificial sweeteners because I think they support behaviors that will ultimately lead to weight gain. (In particular I suspect we may subconsciously learn that sweet foods do not provide energy.)

On a related note... Science has led to many amazing things, but that doesn't mean it can realistically answer all questions. I think we should remember to be a bit humble and honest about which questions can be answered scientifically and which can't. Unfortunately, individual researchers and journalists both have incentives to make the most out of any result.

If you read the jobs page at SRI, there's a bunch of stuff about modeling microwave communications/in the atmosphere/for offensive cyber. I wouldn't be surprised if this guy were involved in that sort of mission.

Am I wrong to see this as a bit scandalous for computer vision as a field before 2012? (It kind of seems like maybe a decade of research at the Berkeley CS department will be tossed out?)

Doesn't the claim that "algorithms need to look at better (high quality, informative) examples" suggest they will need to use less storage, not more?

If you manage to offend 8 people above you in the org chart (without winning others to defend you), and you get fired solely because these people don't trust you, what's the case against that?

Probably not. Hiring is expensive and time consuming, as is training up new employees.

I wonder if maybe it's intended for use when the filesystem is intentionally wiped out, maybe so you can deny any data exists on a disk at all. (Like anti-forensics?) I don't know, I've never really wanted to do anything like that.

https://en.wikipedia.org/wiki/Anti-computer_forensics

In the curriculum for Udacity's self-driving car nanodegree program, your suggestion was made explicitly. ~"(re: why are we focusing on vision?) Humans drive with just our eyes (even one eye would work), perhaps in the future cars will need only camera data instead." (Perhaps many cameras.)

Lately I've been thinking about how funny it is that we make heroes out of the heads of tech companies.

Especially in technology, does anyone really believe that if these particular individuals hadn't contributed their talents, the world would have turned out much different?

Would there be no operating systems without Gates? No social networks without Zuckerberg? No search engines without Page? No e-commerce without Bezos? (Granted, Bezos is pretty exceptional.)

Why should particular individuals capture such massive rewards when their success had more to do with circumstance than anything else?

(There's a connection here to the "great man theory" of history and its critiques.)

Is there any hope I'll be able to find a major ISP that doesn't sell my data? (Google Fiber, I wish?)

Fairvote.org works on electoral reform issues that are meant to help with issues like this. I think that "the marketplace of ideas" in the US is too much of an oligopoly. Ideas like single transferable vote seem like realistic options for improving the situation. (I'm excited to see how thing go in Maine now that they've​ adopted some of these measures.)

Near my old workplace there was an incident where somebody tried to set a fire at night and the police got involved. The criminal got away that night, but I was impressed that they actually caught him a few days later. I got to thinking about how you could do this with today's technology when you're recording things all the time... "Give me the license plates of all the cars that arrived after 11 and left the area between 12:30am and 1:00am?" To say nothing of tracking people using cell phones. (Recall that last bit was discussed in the Snowden leaks...)

You make a good point that Netflix has a pretty sophisticated system for delivering video (not just served from AWS Cloudfront). However IIUC they still use AWS for preparing video files for their CDN.

Okay but that does not solve the problem I am concerned about. That doesn't ensure a diverse mining pool. Already it doesn't make sense for me to mine Bitcoin (I don't have access to these ASICs). I think that economic forces will tend to concentrate mining power and that's a problem.

In this context "deliberate practice" means more than just "practicing on purpose". The description here looks okay:

http://expertenough.com/1423/deliberate-practice

However in Ericsson's book he puts more emphasis on guidance by an expert teacher (who can help identify specific challenges appropriate to the learner).

This article seems to support my suggestion.

https://www.cryptocoinsnews.com/declining-profitability-for-...

Quoting the conclusion of the referenced paper:

"Each future halving of the Bitcoin block reward will narrow the range of profitable entry for new Bitcoin miners considerably. The stronger relative position of incumbents compared to that of new entrants supports further consolidation of Bitcoin mining. Consequently, the distributed Bitcoin ledger, the blockchain, may become more vulnerable to attack, and its trustless nature may ultimately even be destroyed."

If you accept that (1) there are economies of scale in Bitcoin mining and (2) miners are rational and will only mine when it is profitable, then doesn't this suggest that large amounts of mining power should become concentrated in the hands of just a few major miners? Doesn't that seriously undermine Bitcoin's validity?

I only learned about "head on" in the last month or so. In case anyone missed the reference, it's an annoying TV commercial:

https://youtu.be/f_SwD7RveNE

How do you measure - measure a job search?

https://www.youtube.com/watch?v=hj7LRuusFqo

Are you assuming these automakers won't compete with Uber directly? I think some automakers are looking to enter the "transportation services" market.

http://www.wsj.com/articles/ford-says-it-will-focus-more-on-...

Also I heard that there is a legal precedent that if the government confiscates your drugs you can attack them and demand reparations and they will have to pay.

https://en.wikipedia.org/wiki/First_Opium_War

I seem to recall a story from several years back about bombs being disguised as laptops. Requiring devices to be powered on was a simple strategy to make sure they weren't completely fake.

[1] https://www.google.com/amp/s/amp.cnn.com/cnn/2016/02/11/afri...

[2] https://www.fedscoop.com/tsa-targets-explosives-disguised-tr...

I wouldn't be surprised if all electronics had to be powered on for that reason. (Snooping aside.)

I think "home price appreciation" is one of the more interesting factors here. If home prices grow at 4% and wages don't keep up, it seems like that's not likely to be a sustainable over the long term.

I have this sense that home prices all over the US are being artificially inflated. That said, if you figure 4% interest on $1 million, that works out to about $3333 per month. Q: Are there enough people in the Bay Area who are willing to spend ~$3333/month to live in a home currently priced at $1M in order to support that price level? A: Yes, I think so. (Granted, $1M would be quite low for a house near where I live in Mountain View, but maybe the numbers don't seem so crazy?)

I guess I'd worry more about rogue nations developing (cheaper) biological weapons and chemical weapons. I imagine it would be easier to build something like [1] than a robot like this.

[1] https://en.wikipedia.org/wiki/M139_bomblet

I've worked as an SDE at two of the big five tech companies. By far the most common route to getting a first job at these companies is to start with a college internship and continue from there. Basically with only a high school education you wouldn't get an interview, and I'd say pre-university work experience is not too important in the application process either. I'd recommend focusing on your studies first - take the time to learn about graphics or AI or some topic that interests you and will help you succeed in university. You can worry about getting a job afterwards. Personally I had small jobs making web pages before college but the first programming jobs on my resume were with scientific research groups at university.