thatcherclay

A space war from a long time ago and a galaxy far away? Where have I heard that before... maybe this is just a Disney marketing ploy...

Honest question - would you rather see our species capable of transforming Mars and building a colony there, or be able to create a realistic simulation environment, one that was capable of arbitrarily slowing down time.

Obligatory: https://xkcd.com/327/

My issue is that while medical technology may progress at an exponential rate, so too might the challenges of repairing our biology. So yea, we may be getting way smarter, but the problems may keep getting way harder, and then it is not as self evident as this particular article makes out that we are all going to be immortals.

Something I do not understand in this debate is the notion that creating the OS will somehow create an extra security vector - if they have the ability to create it (ie, since they have the private signing key) then isn't that already the same thing? It is not like Apple has a monopoly on good developers, presumably if someone is able to obtain their signing key it would lead to the same thing as obtaining the OS they are worried about getting out.

Sounds interesting - however, if they are successful in reconciliation, then seems logical that the next step will be addressing the transfer of wealth that happens after reconciliation (ie, clearing). I am not sure how much of a problem this truly is - ie, how much of counter party risk is attributed to how often the counter party agrees on the reconciliation but just does not have the capital to pay up at t+3.

Whatever it is, if you could instantaneously reconcile and clear, then some forms of trading could get a lot faster.

Can you post a reference for TPG (which I assume you actually mean Texas Pacific Group) being an "official" CIA proxy, as you say? That is a pretty bold assertion...

It should not be illegal to alter the car's computer, but to drive that car on a road where you are endangering other people, we have certifications and laws for that. IMHO, that is reasonable. I don't want your buggy software to deactivate your power steering and have you slam into me.

FWIW, I find the comparison to getting struck by lighting to be somewhat facetious. Sure, goes to show just how unlikely it is to be killed in a terrorist attack, but I wager that the purpose of a terrorist attack is often more than just the immediate deaths.

Evidon is an interesting company, they sort of screw with both of their customers: * Users - the main thrust of this article, potentially misleading tracking * Advertisers - holding them hostage via being one of the sole providers of certification to the legislation they support (I have witnessed instances where a company turned down their services, were reportedly threatened by an evidon sales rep, and then were shortly thereafter reported as not serving the ad choices banner. Not a smoking gun, but suspicious).

Just as an experiment to see how addicted I was, I stopped drinking coffee for the past week - was a huge awakening on how dependent I was (both physically and socially). Some reflections:

- I probably could not do this if I had to go into an office

- Headaches went away after day 4

- After day 4 I felt generally sharper/more motivated

- I feel much better after workouts

I highly recommend the exercise to others, good way to explore your limitations.

I think you guys are reading different parts of the article.

They present both results, supervised and unsupervised (where unsupervised uses the SFC dataset to train). They achieved 95.92% accuracy LFW with unsupervised (section 5.3) - so they can train on SFC and then classify a single image in a different domain with 95.92% accuracy.

They achieved the 97.25% accuracy level was achieved as you say, when they let the pairs into the training set. But they overfit with LFW alone, and has to add an additional 100k identities with more samples (30) per identity. A very impressive measure, but not quite as good as being able to generalize with 97% accuracy from a single training photo.

Yea, I agree with you. I am trying to make a different point though - there is reserve demand in both dollars and bitcoin, but if the bulk of the pricing set for bitcoin is by transactions founded in speculation (ie, not tied to a physical good or service), then it would seem there is a very limited ability to increase value in the system (aside from new users coming into the market). A currency like USD is different - you can easily and reliably exchange USD for goods and services which increase in value independent of the currency instrument itself. That, and the fact that you can amplify demand by introducing credit, creates a different dynamic, which is why I think the ratio of liquidity created by "speculative" investments versus transactions tied to something external is an interesting indicator.

Anyone know the ratio of speculative investing versus real goods/services transactions on bitcoin?

I ask because it would not surprise me if 90% of all transactions go to people buying/selling against other currencies, which just makes it a zero-sum speculative tool, and with immature market dynamics at that. Zero-sum because it does not pin itself to anything of external value - ie, consumers thinking the economy is going to do well, so they take a significant loan out (increasing demand for USD) to buy a house. Without a fundamental externality like that, I think bitcoin will always just be a novel gambling mechanism, and if we cannot figure out how to make it a legit market then a bad one at that.

It is really impressive when you think about it; the level of genius that goes into tricking our brains to like this synthetic mush is astonishing. But only the same way that the porn industry is astonishing in some ways - both appeal to a desire and then over-stimulate, and both are on tenuous morale standing.

The value prop of catastrophic health insurance never fully made sense to me - if I need it, then I am in bad shape, and whether I would have to go bankrupt to pay for it or pay through some insurance is not super important. Seems likely that if all I have is catastrophic coverage, I am going to go bankrupt after the fact anyway because of the long term effects of whatever landed me in that place (those plans do not cover much).

We had a caved in office in our last space, and the expected downsides from now having enough casual conversations bit us. Engineering did not know what client services was doing, client services did not talk to sales, and leadership was hard/nonexistent. Our new space is night and day - we have a big open floor plan and enough conference rooms so that you can always find a quiet place to brainstorm. We still have some of the old cultural problems that the old space created, but we are trending upward.

Second that - for people who make the ethical/moral reason for limiting or abstaining from eating/using animal products, this is quite troubling. Seems like food companies could strike a reasonable middle ground by labeling foods as "containing animal products" whenever they do without giving away any trade secrets.

TruEffect (http://www.trueffect.com/careers/) - NYC - Full time

TruEffect is an ad tech company that provides ad serving and advanced analytics to some of the most sophisticated digital advertisers in the country. We are looking to hire data scientists, big data engineers, and data visualization experts.

I would love to hear the perspective of a native chinese person here - a little hard to take the New York Times as an unbiased source given the tension there lately.

I think the most valuable thing this is pointing out is that the nature of the problem is shifting to an integration challenge - how to hook together the 5 or 6 services that I will need to get everything up and running. Most web driven businesses are going to need some combination of different technologies at some point - the people who can orchestrate that well are in a good position.

This is great - wish the government would go the next step and open up the traffic optimization to the general tech/data community to help. Think of Netflix recommendation challenge, but helping millions of people cut down on their commute.

Could not agree with parts of this more - in particular, companies that think about bringing on data scientists to extract value from their "big data" would often be better off investigating whether they have the right architecture in place to do that well. It is a buzz word with a goto reaction; when it truth there is from what I have seen the scariest thing about big data is just unwieldy ETL processes grown from years of not thinking clearly about problems.

There is a pretty simple economic argument that people need to consider before jumping into the cookies are bad camp. The market for display ad inventory alone in the past year was to the tune of $40b in the US. There are 250m active internet users. That $40b dollars goes into the pockets of content producers on the internet that make it the tool it is.

Net, if you wanted to drop all of that and get rid of the advertisers, that is a balance of $160 per person per year (and growing) that would fall to some other solution (like increased service charges).

As others have made the case, cookies are important for advertisers. If you want to fix the problem, need to consider another viable solution that either does not sweep their legs out or does but provides an alternative revenue stream to keep the internet alive.

Creative deal, I like it. Anything that gets people off their couch and building something is a winner in my book.

This is something that I half completed, but stopped because I lacked the marketing chops: http://www.mailreceipts.com/