He does not deny that there are people on the tail of the normal curve. But his point would be that there are more people that started at 12 and programmed a couple of hours on the weekend v. 4 hours a day. And there are still more than that that started at 15. This creates a normal curve, not a bimodal.
Sure, the revenue structure based on the assumption "10 guys, most of them hedge fund managers, will make over a billion dollars, guaranteed" seems sustainable.
Most of such income is off one-time events http://www.forbes.com/sites/robertwood/2013/12/20/mark-zucke... You can hope they're repeatable, and that works during the boom years, but during the bust years it becomes a "California double whammy" - not only your projections are over-blown, but the high-income tax events (which tend to be IPOs or big hedge fund wins, judging by the caliber of people stuck with multi-billion dollar tax bills) have also been eradicated by a stagnating/falling economy.
My biggest problems with C++ (one of my favorite languages)
1. ABI issues.
2. Template Error messages.
3. Templates need be defined in header files (generally).
4. Object Slicing.
5. Cross platform Unicode/ASCII string encoding.
I understand the reasons behind these, none the less they are really annoying.
Why is it so hard for twitter to implement some optional filter for this stuff?
You have a complete history of a users tweets and a record of how long the person has been active, number of followers, textual analysis of the tweets, block/mute status with respect to other people, whether you follow ect. It seems like it should be really easy for twitter to put messages into a spam/harassment folder, never to be seen by the target.
Edit: Not only would this alter a core and fundamental aspect of how Twitter currently operates but it would certainly be non-trivial to implement in an unobtrusive manner.
Who wants what blocked? Is there a generic 'spam/abuse' filter? Is it on by default? Are those tweets put into a spam/abuse section or entirely unviewable. What happens when someone sends a tweet that is blocked by the filter, are they notified?
Try to clearly define what metrics/identifiers you would use to determine a tweet is offensive and should be blocked and then call this implementation trivial.
Federated twitter blocking where people can choose to outsource the blocking to people they trust would be pretty cool. Kind of like AdBlockPlus lists.
He does not deny that there are people on the tail of the normal curve. But his point would be that there are more people that started at 12 and programmed a couple of hours on the weekend v. 4 hours a day. And there are still more than that that started at 15. This creates a normal curve, not a bimodal.