davidbnewquist

Actually, I think 'arguably' could be removed in that context without being disingenuous. Thank you for the feedback.

So by removing that word, I'd be adding 'truthiness'?

Misspellings are deliberate and enhance security. Note that you still knew to pick the misspelled "striped plan" and passed. Another label for that picture might be "stryped plain".

Also, you correctly identified hydrant because "moon lander" fortunately was not one of the choices:)

Its true that a 2 panel peoplesign is slightly larger than most text-based CAPTCHAs. However, there will soon be an option to use only 1 panel, which will hopefully make it small enough even for space conscious sites.

Thanks JJ. You are right about English literacy being useful to pass. Soon, I will expose an option that uses picture labels instead of natural language phrases.

But don't throw the proverbial baby out with the metaphorical bath water! The literacy requirement may be an effective defense against the armies of human sweatshop CAPTCHA solvers that we hear about in the news.

Regarding attacks, would be attackers have more important targets for now. peoplesign isn't yet protecting any major sites.

peoplesign images have been recolored using a "secret sauce" algorithm. Recoloring one of two similar pictures reduces effectiveness of color histogram matching. It even reduces effectiveness of more sophisticated color profile matching techniques. In short, the recoloration is defense against a particular type of attack.

Good call, in fact I'm working on this. I force numPanels >=2 because 1 panel currently causes layout issues.

You answered your own question. 30-40% > 3%.

But more deserves to be said. A site like Hotmail is a big, focused target for spammers. They face pain if a bot can beat their CAPTCHA even 10% of the time. Smaller sites get spam too, but they generally get it from dumber bots looking for unprotected sites. A 2 panel peoplesign CAPTCHA is probably all they need.

With 2 panels at 6 labels each, a random guess has a 1/36 chance to pass. Fortunately, a large number of incorrect guesses is behavior that can be detected and blacklisted.

Futhermore, peoplesign can be customized to offer more security. Number of panels (and soon number of menu labels) and other characteristics can be modified. Check out a customization demo at http://peoplesign.com/main/pickTheLabelDemo.html

Well, if the genes regulating sleep patterns are half as tedious as the gc params for the 1.4.2 JVM, I wouldn't want to consider every possibility either:P

In software, optimizing this type of garbage collection for low pauses usually results in reduced application throughput. Evolution in general clearly does not favor "low pause" sleep patterns.

However, there are anecdotes of rare people that subsist on "low pause" sleeping. Allegedly, Thomas Edison had a 24 word day and frequently cat-napped. Contrast that with Einstein who allegedly slept 10 hrs a night.

For modern garbage collection systems (e.g. the JVM), there is no single setting that is best (or even acceptable) for all applications. The same is probably true for sleep patterns.

time-ster! J/k. Maybe time2time?

The hypothetical scenario you describe would be tough for Hazardous software; however, they had the opportunity to do a patent search for related pending applications prior to investing any resources creating Achron.

The Patent subsystem of our govt can serve a useful purpose for inventors and society in general.

But as with nearly any system of this type, there are "corner-cases" for undesirable usage.

A big problem is that someone can submit a patent and "sit on it" for years and years. A fallow patent provides no value to taxpayers. At best, the issue of the patent had a marginal cost for taxpayers, but at worst, there is a huge opportunity cost if commercial development of the invention yields value.

Our patent system should demand inventors make timely, reasonable effort towards commercial development of their inventions.

Kudos to the Achron inventor (Christopher Hazard?) for doing just that.

Good points, good links. True, this is a time traveler's Starcraft/Warcraft/Command and Conquer, etc. On the Achron website they state:

"The company's innovation will create new sub-genres of games, starting with the game Achron." Achron is in the MTS sub-genre of RTS, to be pedantic:) However, the scope of differences in here make the term "sub-genre" seem unfair.

The new time related gameplay will demand a huge new strategic skill-set from the players. I don't think this will be true for Starcraft II. And you won't be able to learn this skill-set by playing any of the "old-school" RTS games.

I may not have posed my question clearly.

You say: "Nearly every task either adds ... to my stress or relieves it." My question is: why does this seems to be true for everyone?

The answer is not necessarily related to energy or time expenditure. For me, playing consecutive games of intense full-court basketball is fun/relaxing/play.

The complexity of their game engine must be staggering...

The time related gameplay elements shown in the videos struck me as extensive, entirely unprecedented and exciting. Wow!

Your "work/chore/business", "fun/relaxing/pleasure" dichotomy seems to be universal. I can look at any task on my to-do list and instantly put it in one of those buckets. Probably anyone with a to-do list can.

Why is this?! Assuming this ability is universal (and not just something chime and I are good at:), there must be an innate psychological reason. Possibly even a "simple" reason.

I will venture a guess: real or imagined self-threat is always associated with failure to complete a "work" task. This is never true for an individual "play" task.

Thoughts?

This has to be the mother of all OCR test cases:)

Sorry: I was actually joking. I should have capped with an emoticon. But it's cool to know Akamai actually has a triangulation DB: I didn't know it was feasible. So I gave you a karma point:)

The method: dynamic programming algorithm, triangulation and tracert.

Keyword based searching definitely has limitations. If a search keyword does not appear in the context of any "good" documents you're searching for, the search engine will not return any "good" results. In the example above, a search for "bull terrier" and imgcolor=red only returned 1/20 "good" result. And "bull terrier" has high semantic relevance. This is because the authors of the "good" documents I was looking for (e.g. the Target dog) rarely use "bull terrier" in the document context.

Plus keyword based search engines can be easily hacked to associate garbage with certain keywords.

I guess a color parameter would be considered semantic search? Here's a cool link (new today) about google integrating more semantic search features. http://www.pcworld.com/businesscenter/article/161890/semanti...

Good idea, but the results remind me of a Target commercial:P

If you switch the query to "dog bullseye" with imgcolor=red, 10 out of the first 20 results show the Target dog. If you remove the color restriction, only 5/10 feature the dog. So this feature potentially aids search accuracy.

It would be nice to submit an image as the search query and have results ordered by color histogram euclidean distance. Result pics would be of similar color and would probably often contain similar objects.

From the article:

"online advertising share of all advertising could reach as high as 23% if a scalable advertising solution for social media could be found."

Adpinion.com (y-combinator startup) may have an answer for that. Among other things, they're trying to optimize ad-campaigns based on in-ad user feedback.

Great idea, but is it possible that some "effective" ads might result in a knee-jerk reaction to click a down-mod button?

E.G. the picture of the frightened pointing guy on a certain colorful candy homepage:)

Do you feel that Wolfram can rightly claim the thesis offered and explored in the book? Regardless, how well do you think he supported the thesis?

The thesis of A New Kind of Science is something like "systems comprised of a small number of simple rules can perform arbitrarily complex computations."

The book proceeds to support the thesis. The content is comprised of descriptions of such systems, corresponding Mathematica execution trace diagrams, and analysis. These analyses are related to a ambitiously large scope of natural phenomenon and scientific knowledge.