Back in September, Intrade had Obama's odds at 80% chance of winning. With the possibility of an October surprise, I felt that betting on Romney at 4-to1 odds sounded pretty good. If I had, I could have doubled my money when Romney reached 40% chance a couple weeks later. It turns out a lot of other guys thought Romney at 20% was low, too, and there was a discussion about why. One of the best answers was from an investor who said he was hedging his bets by betting on Obama. If Obama wins, he will likely owe more in taxes. His feeling was that Obama's odds were overrated on Intrade because of guys like him.
Another interesting thing happened on Intrade the last week before the election. Someone made it perfectly clear that they intended to manipulate the market to lower Obama's odds, and increase Romney's. They did this by making massive sales of Obama stock over short periods, like a couple of minutes, causing the price to fall by 5-10%. In forcing down the price, the manipulator sold most of his stock at around 5% below what the market was willing to pay. He did this over and over, about twice an hour, for days, selling something like $2-$4 million of Obama stock, losing something like $200,000 through irrational trading behavior. Anyone with an account had an easy chance to make money over those days, because after every massive sell-off, the price would recover 90% of it's price within 15 minutes. Buying low and selling high was never so predictable.
I always state "programmer" as my occupation, in business and social gatherings. If I want a better title, I'll hold out for God Emperor.
Re:To bad that non college education does not resp
on
MOOC Mania
·
· Score: 2
Unfortunately, with zero experience, there are hardly any employers that will hire you without a degree. Employers need to start taking responsibility for workforce training again.
I think MOOCs may offer exactly this. What good is a printed certificate of MOOC completion? Not much. However, a course designed by employers as an entrance exam could be pretty cool. Several years ago, our VP of engineering asked me to put together some hard problems to solve that are related to the work we do, and then distributed them at a local university. Several grad students were interested in the challenge, and wrote up solutions. We invited them all to our company for free pizza, and went over solutions. We made offers to two of them, and one accepted. He's still with us today, and he's fabulous. We could theoretically turn those sorts of problems into a company specific MOOC that gave applicants a good feel for the sorts of skills we need. We'd grade the solutions ourselves.
Since down below there's a bunch of PO-ed Romney supporters bashing Obama rather than talking about Tesla being sued, I find replying to Combustible Motors more on-topic!
Getting back to TFA, Tesla does not have stores selling Telsa's in those states. Instead, you buy them over the Internet. All Tesla has is showrooms where they can explain their story to people, but they can't sell cars. Tesla believes, and I agree, that this avoids running afoul of dealer franchise laws. Frankly, I don't see how these dealers can explain why it is illegal for people in New York to buy a car over the Internet, or why it is illegal for Tesla to have showrooms in malls. Laws to protect car dealers could only have been sold to state legislatures by used car salesmen. I hope these lawsuits crash and burn.
I'm surprised no one else is talking about Gallup here. They screwed up, big time. There are variations between elections which are impossible to correct for through polling science, but Gallup made unforgivable mistakes. For example, three percent of the population have no phones. Every pollster but Gallup used population data to take them into account, while Gallup computed their predictions as if these people did not exist. Every Obama voter without a phone who voted contributed to Gallup's error. This probably threw Gallup off by 1.5%. Then, Gallup under-estimated the black and Hispanic voting populations, because they changed how they ask about race in a way that increased the likelihood of a voter saying they were black or Hispanic by around 2%. This meant their samples consistently had more blacks and Hispanics than there should be according to population data, so they reduced the weight of their votes. Add up these two mistakes, and you get most of Gallup's 4% error in the national vote. The remaining error was due to Gallup "twiddling" weighting factors to reduce outliers, which they felt were the data points that consistently said minorities would vote in higher numbers than they were predicting. These three mistakes explain Gallup's crappy job at estimating Obama's approval rating as well. A 4% error! How do you interview 20,000 people and get a 4% error? By being morons. They were worse than any automated crappy poll that didn't even call cell phones.
Not only were they total morons, but evidence strongly suggests that they purposely manipulated published polling data in order to create the impression of more momentum for Romney than there was. Before the first debate, Gallup published only registered voter data, and the day after, they switched to "likely voters", which they claimed would cause a 7% swing towards Obama, rather than the 2-4% estimated by other pollsters. Gallup and Rasmussen's incredible swing caused a 1-2% error in RealClearPolitic's polling average of the popular vote. They used Sandy as an excuse to stop polling, and the day before the election they seem to have rewritten their likely voter model to make their prediction less wishful for Romney. The dropped Romney's advantage in the popular vote from 7% to 1%, mostly over the Sandy outage. The other polls showed Sandy having far less impact. This points to willful manipulation of the likely voter model by Gallup.
In contrast, Pew Research, who uses very similar methods as Gallup, but without the dumb errors, guessed Obama would win the popular vote by 3%. They were right. They never had major swings that could only be explained by either wishful thinking or deliberate poll manipulation. They were consistently within sampling error of polling averages (minus Gallup and Rasmussen). I've followed Gallup for years, and I'm sad to see them fall so low. They're practically synonymous with public polling, and it's heartbreaking to see them fall apart. Since we obviously can't trust them anymore, I'll have to get my polling data elsewhere.
Let's here it for Nate Silver. I doubted him when he said Gallup was smoking dope, but he was right. I'll certainly take him more seriously going forward.
I turn 49 in three weeks, and I still love programming. It remains my work, hobby, and passion. I think my ability to crank out awesome code leveled off when I was about 30, and since then I've had to settle for enjoying mentoring the next generation rather than soaking up knowledge like a sponge. At one point, I looked around and realized there wasn't anyone left to learn from, at least not anyone who I was capable of emulating, and that many people were looking at me to help them. I started a company back in 2000, and continue to work in the position I created for myself, and I am still having a great time.
However, I agree... If I had to go find a new job as a programmer, my age would be an issue. I intend to stick with my company as long as they need me, but after that, I'll probably start another one. I haven't become a stronger programmer with time, but the experience I've gained working in startups has made me a better entrepreneur.
I've found some interesting statistics. Start with Nate Silver's article today. Note that if Romney wins even 0.55% more of the popular vote than Obama, Nate gives Romney a 50-50 chance of winning. Now go to Intrade, which currently is being massively manipulated, but ignore the popular Obama and Romney markets, and look at the Romney/Obama wins by more than 0.5% of the vote market. It's currently showing 42% for Romney and 57% for Obama, meaning they can't predict within 1% where the vote will land with more than 15% probability! Even the pollsters are confused.
So, I'd rather be Obama than Romney at this point, but with such crummy ability to guess how people are going to vote, Romney's odds aren't bad. That would help explain why Intrade is giving around 66% chance of Obama win, rather than anything like the probabilities Nate is giving him. I think if Nate took this into account properly, he'd be more in line with Intrade, even if Intrade is being manipulated.
There are really two very interesting races going on here. I'll watch the Romney/Obama results tomorrow eagerly, but there's also the Gallup/Nate Silver race. If Gallup is right about the popular vote (Romney by 5%), Romney wins by a landslide, Nate loses, and Gallup will become the only pollster of any worth. If all the other pollsters are right, Obama wins, Nate's status grows, and Gallop is surely run by worthless fools. It's really Gallup sticking their neck out, and I'm going to enjoy seeing if it gets chopped off or not.
In 1981, my high school counselor, for reasons unclear to me, found me a job programming in Fortran at Lockheed Martin Missiles and Space, for $4/hour. I was in Heaven and knew that I'd do it for the rest of my life, even at minimum wage. Today, on Sunday, I wrote code for two or three hours on an audio stack for screen readers for the blind. I still love it. I'm just lucky that industry decided to pay me more than minimum wage.
I don't think the US is worried about traditional a-hole "hacker" who is deluded into thinking cyber-vandalism is a good thing. I think now that Israel and the US teamed up to cyber-attack Iranian nuclear enrichment facilities, the entire world is racing to bone up their cyber-warfare divisions. Real war stuff where people die, not just kids who misunderstood what the term hacker meant.
Has anyone heard what the extra transistors TSMC was adding to every Xilinx FPGA were for? I doubt they're a good thing. Now that everything from a phone to a car has a computer in it, there's serious potential for government-sponsored hanky-panky. RMS seemed concerned about that in e-mails I read a few years back. He was wondering if it is possible or practical to verify that ASIC chips are built as specified, and unfortunately the short answer is no, as the customer typically doesn't have the cash, time, or interest to reverse-engineer their own chips. We don't even know if all our tech toys have built-in remote controls or not. A pissed off Iran launching cyber attacks is a bit scary, but a pissed off cyber-China...
Great post. Up until 1995, hackers who believed in sharing ruled in the OSS movement. Then, money came. Redhat figured out how to package OSS to make money selling support to corporations, while Debian figured out how to deliver binary packages with complex inter-dependencies. With the rise of binary distribution of packages in popular distros, we saw the fall of hackers sharing with each other. Instead of sharing directly, hackers have to lobby to have their projects included in popular distros. The process is several times harder than publishing an app in the Apple App Store, where hackers are much freer to share their work. It's heartbreaking.
Not a bad idea. He could even make it clear he had everyone's password and user name by encrypting each user name with their password after doing about 5 seconds worth of serial SHA-1 hashes on each.
It all comes down from the guy in charge. I've set the coding style at two of the last three companies I've worked for, and while new guys may not like it, it does promote effective teamwork. It's a lot more than just indentation. We use the same data structure styles, iterator styles, and everything from start to finish. It can be very difficult to determine who wrote any given piece of code, and a git-blame shows that it's a patchwork quilt with everyone contributing all over the place.
The guy who I will say is now the lead developer of our main software suite writes code that looks exactly like mine. We've even on occasion wrote entire functions, in the same place, which were identical down to the last character, even in the obligatory comment. No git conflict in that case. At the other company, there literally were no coding style guidelines, and the result is no one was able or willing to easily edit anyone else's code. That makes it very difficult to move people around to projects that need them most.
There are a ton of developers out there who are loners. It's a field that attracts geeks who are somewhat social-phobic. I've never seen programmers voluntarily adopt a common coding style. It's more fun to fiddle around on your own. So, instead of working together as a team to solve real problems rapidly, individual programmers go off in different directions, learning more about different ways to write code than actually writing code.
I'm not sure that was a real quote from Cisco. It looks to me like they simply didn't want the business. In such cases, business file what are called f-you quotes, which are outrageously priced to take into account that the bidder may not currently have the capability to fill the contract, or that it would be defocusing. Priced high enough, they could sub-contract to HP, for example, and still make a lot of money.
That said, I went to our local office the other day and poked my head into the networking closet. I see the same cheap crummy wifi routers I put there before our little company got bought. Right next to them is a Cisco router worth maybe $10-20K. It's worth more than all the computers and related hardware in the office combined.
Microsoft and Amazon pay an average of $30K/year for experienced programmers in India. If you want that same talent through an off-shore contracting company, it'll cost you double, or $60K/year. Idiots who think they can get $100K worth of development for $10K paid to an offshore contracting company deserve the failure they pay for. Even if you pay for good off-shore talent, your source code will be stolen, no matter what the contracting company tells you, and you'll risk competing with an offshore company that got it for free.
Off-shoring software development only makes sense in one case: open-source software development. In this case, your business model already takes into account the possibility of others using your source code to compete with you. If you're OK with that, just hire programmers directly through online sites designed for the purpose. $10K/year is a reasonable budget in this model, where if you're careful, you can get a great deal, while helping improve the life of some software developer in a place where $10K/year is a fantastic wage.
Bottom line: if keeping your source code private is important to you, hire locally. If you don't care, go open source and hire remotely.
I agree that he should get a 4-year CS degree. I'm old enough to remember when co-workers didn't have CS degrees because they weren't offered when they went to college, but now days I rarely see a coder in a decent job without a 4-year degree of some kind. I've hired mathematicians and physicists, because frankly they're often super smart, and coding is easy to teach. However, if you're not a genius, you can still be a valuable contributor, but I rarely see one without a CS or engineering college degree.
I'm not so sure about the death of the high-school drop out genius coder. While not a dropout, I admire kids like the 18-year-old behind the XBMC flavor of the Raspberry Pi OS. Someone with business experience should offer to cooperate on some startup with him. Even Steve Jobs had a mentor to guide him while building Apple. An attempt at building a startup right out of high school could be considered a postponement of college in order to gain real world experience. In an economy which is not generating enough jobs for new graduates, starting a company on your own is a great way to create one for yourself.
I probably read about Freicoin on slashdot. They've got a proposal for a fix to BitCoin that defeats hoarding. I'm pessimistic. The whole premiss of BitCoin is people don't trust each other. To "mine" bitcoins, we burn real resources to power all those machines computing SHA-1 hashes. Someone said that mining bitcoins was like burning real money and getting a fake bitcoins to replace it. No, "real" money is just stupid paper. It's amazing that we still use dead tree money. Burning it doesn't hurt the economy, other than reducing inflation. Wasting electricity and all those machines computing bitcoin hashes... that's real harm.
I know Ripple has not caught on, but I like the basic premiss. In Ripple, I create money by giving people I trust coupons with dollar values in proportion to how much I trust them. This gives us a way to generate cash out of nothing but trust, while eliminating incentives for hoarding.
The existing phone system is a dinosaur. We should switch to a modern digital P2P system where everyone has an online identity. The first time someone wants you to receive and e-mail from them, charge them $0.01. The first time they want you to answer their call, charge them $0.05. We need an electronic currency that enables fast micro-transactions, and we need to stop acting like the world is still plastered with individual analog phone lines rather than being all digital. Simply put, we need to take advantage of he capabilities of the hardware we already built.
Yep. You have to figure the author of the system wanted it to be deflationary, and he likely has at least a 100K himself if not a million. Apparently most coins are hoarded, not traded. At even $1/coin, the original guys have to have huge balls not to sell all their imaginary money for real money. At 22 coins, I figure it's just entertainment money. If someone hacks my computer and steal them, I'll be only a little pissed off. If they become worthless, oh well. Imagine what it's like to be sitting on 100K of them, when you got them for free? My guess is these guys must have a lot to do with regulating time price on Mt Gox.
I see the biggest threat to the price of a BTC as a new, better system. The world does need an electronic currency which is easily traded, but bitcoins are simply to hard to get hold of.
It's not the complexity that's increased exponentially, but the total amount of compute power applied world-wide to mine bitcoins. I mined about 20 coins about a year ago with a new graphics card, and then converted the machine to a new gaming PC for my son. Even then, it only made sense if:
1) You wanted a gaming machine with a powerful graphics card anyway. 2) You counted your effort to build the machine and do the mining as entertainment.
You can do it in Windows or Linux. It's not a way to make money, IMO, unless you've got access to free electricity, and hopefully cheap labor. The great halving will occur in December, at which point a lot of miners will pack up and go home.
The currency is deflationary, or at least it will be once new bitcoins can no longer be mined. If you lose your wallet, the bitcoins are gone forever. Over time, there will be fewer in circulation, yet more people in theory will be using them. That's OK, because a bitcoin can be broken into fractions of 1 in 100 million. So, if deflation causes the currency to increase in value, it wont be a problem until each bitcoin is worth about 1 cent times 100 million, or $1M each. There will eventually be 21 million bitcoins, at which point it's maximum total value before there's not enough fractions of coins to make penny sized transactions would be $2.1 quadrillion, which is more than 100X the US GDP. If the coins were distributed equally among 10 billion people, each would own 2,100 micro-bitcoins, each of which can be broken into 100 pieces, called a "satoshis" after the author of the system.
I figured I'm mine a few for fun and hope one day they'll be worth a lot. On a planet with 7 billion people, I'm guaranteed to have far more than average at 22 BTC. The average person will have probably around 0.002 BTC.
Back in the days of sharing source code on lists, before the dominance of the pre-compiled package, there was no binary incompatibility. With the advent of pre-compiled packages, we also both the rise of binary incompatibility and the rise of the package sponsors, the people who decide what's good enough for the rest of us to share. It used to be a bazaar, where I could lay out my blanket and sell the crud I'd created. Now it's a temple, where we all download from the True Source.
I've got a second-hand story to share about the good old days. At Berkeley, a friend of mine who was a Ph. D. student actually drove to Palo Alto now and then to attend the Home Built Computer Club meetings. There was a guy there with his latest hack that he wanted to share. He was selling it literally on a blanket. He had an un-stuffed circuit board and a copy of what parts to buy to stuff it, and claimed that if you did it, you'd have a working computer. His big idea at the time seemed to be selling empty circuit boards on a blanket. My friend said, "What a loser," and never thought of him again until that guy, Woz, co-founded Apple and changed the world.
The days of open collaboration between Linux developers has been hampered by binary incompatibility, and high hurdles to share software on popular software platforms like Debian and Fedora, and Gnome/GTK. We've seen hard feelings and fractures between groups like Ubuntu and Gnome, and lot's of unhappy users.
Are the days of freely sharing software on lists essentially in the past, or is there some way to once again pump life into that creative engine? Can we work smarter?
After Urban Terror, I had to let her play the usual kid games. She played Pokemon on her DS for years, and found Wizard 101, Roblox and Mindcraft, just like all the other kids. I wound up installing two Windows machines so the kids could play the games their friends were into. My son has gone off with Mindcraft, and built some insane worlds. My daughter is now a vegetarian into horses and animals. Both kids loved playing Star Craft against the computer with me, though I can't play it anymore due to vision issues. Partly because I suck so badly now days, both kids love playing Halo with me. They call it "kill dad," as if that were the name of the game, rather than Halo.
Few of you will agree with this one, but I let my daughter play it from 2 to 5 years old. She loved it more than any other game. It was hysterical. If you know this game, it's an awesome first-person shooter with guns, grenades, and blood, played on various urban maps. Last I played, it was mostly very skilled diehards with years of experience. She was slow compared to the other players, so she compensated by using some insane machine gun used for close range (IMI Negev). Her favorite thing was killing guys on the other team, but that was really hard because of their skill and her age. She developed unique strategies. For example, on one map she would smoke grenade her own team's flag, and sit just outside the smoke waiting for the other team to appear with the flag, and then lay into them with the negev. She had a few other strategies, but they were always ambushes to compensate for her low speed and accuracy. She loved grenades and that big inaccurate machine gun.
I never had the heart to IM the other players that a they'd just been killed by a four year old girl. I also never worried that her goal in the game was rarely protecting the flag or trying to get the other team's flag. Sometimes it was just wondering around, looking for a good place to ambush a player. Sometimes she just wanted to see what all the different guns do, or simply explore the map. Even at her best, her kill ratio was awful, maybe 1:4, but she didn't mind the dieing, though if she was trying to do something it frustrated her to have to start over. Most of her ambushes had to be set up very close to home base, or if she wanted just to play around, she'd find a mostly useless part of the map rarely visited by the other players.
I had to end her time playing Urban Terror when she began to read. I never told her what those IM messages scrolling past said, but they were pretty horrible. She wasn't exactly a good team player, and some of these guys take the game waaaaay to seriously. My son is 2 years younger, and I only let him play it about a year, because my daughter would always watch and coach him, so her starting to read IM messages became a problem. She would ask why she couldn't play anymore, but at that age such questions are easily deflected, just like whether or not Santa Clause is real.
The process you're describing is only acceptable to people who get paid while going through that process. Tell me of one example of someone who just wanted to do the right thing by informing the world, who actually went through that process. Someone currently alive.
Slashdot Mirror
Back in September, Intrade had Obama's odds at 80% chance of winning. With the possibility of an October surprise, I felt that betting on Romney at 4-to1 odds sounded pretty good. If I had, I could have doubled my money when Romney reached 40% chance a couple weeks later. It turns out a lot of other guys thought Romney at 20% was low, too, and there was a discussion about why. One of the best answers was from an investor who said he was hedging his bets by betting on Obama. If Obama wins, he will likely owe more in taxes. His feeling was that Obama's odds were overrated on Intrade because of guys like him.
Another interesting thing happened on Intrade the last week before the election. Someone made it perfectly clear that they intended to manipulate the market to lower Obama's odds, and increase Romney's. They did this by making massive sales of Obama stock over short periods, like a couple of minutes, causing the price to fall by 5-10%. In forcing down the price, the manipulator sold most of his stock at around 5% below what the market was willing to pay. He did this over and over, about twice an hour, for days, selling something like $2-$4 million of Obama stock, losing something like $200,000 through irrational trading behavior. Anyone with an account had an easy chance to make money over those days, because after every massive sell-off, the price would recover 90% of it's price within 15 minutes. Buying low and selling high was never so predictable.
I always state "programmer" as my occupation, in business and social gatherings. If I want a better title, I'll hold out for God Emperor.
I think MOOCs may offer exactly this. What good is a printed certificate of MOOC completion? Not much. However, a course designed by employers as an entrance exam could be pretty cool. Several years ago, our VP of engineering asked me to put together some hard problems to solve that are related to the work we do, and then distributed them at a local university. Several grad students were interested in the challenge, and wrote up solutions. We invited them all to our company for free pizza, and went over solutions. We made offers to two of them, and one accepted. He's still with us today, and he's fabulous. We could theoretically turn those sorts of problems into a company specific MOOC that gave applicants a good feel for the sorts of skills we need. We'd grade the solutions ourselves.
Since down below there's a bunch of PO-ed Romney supporters bashing Obama rather than talking about Tesla being sued, I find replying to Combustible Motors more on-topic!
Getting back to TFA, Tesla does not have stores selling Telsa's in those states. Instead, you buy them over the Internet. All Tesla has is showrooms where they can explain their story to people, but they can't sell cars. Tesla believes, and I agree, that this avoids running afoul of dealer franchise laws. Frankly, I don't see how these dealers can explain why it is illegal for people in New York to buy a car over the Internet, or why it is illegal for Tesla to have showrooms in malls. Laws to protect car dealers could only have been sold to state legislatures by used car salesmen. I hope these lawsuits crash and burn.
I'm surprised no one else is talking about Gallup here. They screwed up, big time. There are variations between elections which are impossible to correct for through polling science, but Gallup made unforgivable mistakes. For example, three percent of the population have no phones. Every pollster but Gallup used population data to take them into account, while Gallup computed their predictions as if these people did not exist. Every Obama voter without a phone who voted contributed to Gallup's error. This probably threw Gallup off by 1.5%. Then, Gallup under-estimated the black and Hispanic voting populations, because they changed how they ask about race in a way that increased the likelihood of a voter saying they were black or Hispanic by around 2%. This meant their samples consistently had more blacks and Hispanics than there should be according to population data, so they reduced the weight of their votes. Add up these two mistakes, and you get most of Gallup's 4% error in the national vote. The remaining error was due to Gallup "twiddling" weighting factors to reduce outliers, which they felt were the data points that consistently said minorities would vote in higher numbers than they were predicting. These three mistakes explain Gallup's crappy job at estimating Obama's approval rating as well. A 4% error! How do you interview 20,000 people and get a 4% error? By being morons. They were worse than any automated crappy poll that didn't even call cell phones.
Not only were they total morons, but evidence strongly suggests that they purposely manipulated published polling data in order to create the impression of more momentum for Romney than there was. Before the first debate, Gallup published only registered voter data, and the day after, they switched to "likely voters", which they claimed would cause a 7% swing towards Obama, rather than the 2-4% estimated by other pollsters. Gallup and Rasmussen's incredible swing caused a 1-2% error in RealClearPolitic's polling average of the popular vote. They used Sandy as an excuse to stop polling, and the day before the election they seem to have rewritten their likely voter model to make their prediction less wishful for Romney. The dropped Romney's advantage in the popular vote from 7% to 1%, mostly over the Sandy outage. The other polls showed Sandy having far less impact. This points to willful manipulation of the likely voter model by Gallup.
In contrast, Pew Research, who uses very similar methods as Gallup, but without the dumb errors, guessed Obama would win the popular vote by 3%. They were right. They never had major swings that could only be explained by either wishful thinking or deliberate poll manipulation. They were consistently within sampling error of polling averages (minus Gallup and Rasmussen). I've followed Gallup for years, and I'm sad to see them fall so low. They're practically synonymous with public polling, and it's heartbreaking to see them fall apart. Since we obviously can't trust them anymore, I'll have to get my polling data elsewhere.
Let's here it for Nate Silver. I doubted him when he said Gallup was smoking dope, but he was right. I'll certainly take him more seriously going forward.
Nate was amazingly accurate. Gallup... crash and burn in flames.
I turn 49 in three weeks, and I still love programming. It remains my work, hobby, and passion. I think my ability to crank out awesome code leveled off when I was about 30, and since then I've had to settle for enjoying mentoring the next generation rather than soaking up knowledge like a sponge. At one point, I looked around and realized there wasn't anyone left to learn from, at least not anyone who I was capable of emulating, and that many people were looking at me to help them. I started a company back in 2000, and continue to work in the position I created for myself, and I am still having a great time.
However, I agree... If I had to go find a new job as a programmer, my age would be an issue. I intend to stick with my company as long as they need me, but after that, I'll probably start another one. I haven't become a stronger programmer with time, but the experience I've gained working in startups has made me a better entrepreneur.
I've found some interesting statistics. Start with Nate Silver's article today. Note that if Romney wins even 0.55% more of the popular vote than Obama, Nate gives Romney a 50-50 chance of winning. Now go to Intrade, which currently is being massively manipulated, but ignore the popular Obama and Romney markets, and look at the Romney/Obama wins by more than 0.5% of the vote market. It's currently showing 42% for Romney and 57% for Obama, meaning they can't predict within 1% where the vote will land with more than 15% probability! Even the pollsters are confused.
So, I'd rather be Obama than Romney at this point, but with such crummy ability to guess how people are going to vote, Romney's odds aren't bad. That would help explain why Intrade is giving around 66% chance of Obama win, rather than anything like the probabilities Nate is giving him. I think if Nate took this into account properly, he'd be more in line with Intrade, even if Intrade is being manipulated.
There are really two very interesting races going on here. I'll watch the Romney/Obama results tomorrow eagerly, but there's also the Gallup/Nate Silver race. If Gallup is right about the popular vote (Romney by 5%), Romney wins by a landslide, Nate loses, and Gallup will become the only pollster of any worth. If all the other pollsters are right, Obama wins, Nate's status grows, and Gallop is surely run by worthless fools. It's really Gallup sticking their neck out, and I'm going to enjoy seeing if it gets chopped off or not.
In 1981, my high school counselor, for reasons unclear to me, found me a job programming in Fortran at Lockheed Martin Missiles and Space, for $4/hour. I was in Heaven and knew that I'd do it for the rest of my life, even at minimum wage. Today, on Sunday, I wrote code for two or three hours on an audio stack for screen readers for the blind. I still love it. I'm just lucky that industry decided to pay me more than minimum wage.
I don't think the US is worried about traditional a-hole "hacker" who is deluded into thinking cyber-vandalism is a good thing. I think now that Israel and the US teamed up to cyber-attack Iranian nuclear enrichment facilities, the entire world is racing to bone up their cyber-warfare divisions. Real war stuff where people die, not just kids who misunderstood what the term hacker meant.
Has anyone heard what the extra transistors TSMC was adding to every Xilinx FPGA were for? I doubt they're a good thing. Now that everything from a phone to a car has a computer in it, there's serious potential for government-sponsored hanky-panky. RMS seemed concerned about that in e-mails I read a few years back. He was wondering if it is possible or practical to verify that ASIC chips are built as specified, and unfortunately the short answer is no, as the customer typically doesn't have the cash, time, or interest to reverse-engineer their own chips. We don't even know if all our tech toys have built-in remote controls or not. A pissed off Iran launching cyber attacks is a bit scary, but a pissed off cyber-China...
Great post. Up until 1995, hackers who believed in sharing ruled in the OSS movement. Then, money came. Redhat figured out how to package OSS to make money selling support to corporations, while Debian figured out how to deliver binary packages with complex inter-dependencies. With the rise of binary distribution of packages in popular distros, we saw the fall of hackers sharing with each other. Instead of sharing directly, hackers have to lobby to have their projects included in popular distros. The process is several times harder than publishing an app in the Apple App Store, where hackers are much freer to share their work. It's heartbreaking.
Not a bad idea. He could even make it clear he had everyone's password and user name by encrypting each user name with their password after doing about 5 seconds worth of serial SHA-1 hashes on each.
It all comes down from the guy in charge. I've set the coding style at two of the last three companies I've worked for, and while new guys may not like it, it does promote effective teamwork. It's a lot more than just indentation. We use the same data structure styles, iterator styles, and everything from start to finish. It can be very difficult to determine who wrote any given piece of code, and a git-blame shows that it's a patchwork quilt with everyone contributing all over the place.
The guy who I will say is now the lead developer of our main software suite writes code that looks exactly like mine. We've even on occasion wrote entire functions, in the same place, which were identical down to the last character, even in the obligatory comment. No git conflict in that case. At the other company, there literally were no coding style guidelines, and the result is no one was able or willing to easily edit anyone else's code. That makes it very difficult to move people around to projects that need them most.
There are a ton of developers out there who are loners. It's a field that attracts geeks who are somewhat social-phobic. I've never seen programmers voluntarily adopt a common coding style. It's more fun to fiddle around on your own. So, instead of working together as a team to solve real problems rapidly, individual programmers go off in different directions, learning more about different ways to write code than actually writing code.
I'm not sure that was a real quote from Cisco. It looks to me like they simply didn't want the business. In such cases, business file what are called f-you quotes, which are outrageously priced to take into account that the bidder may not currently have the capability to fill the contract, or that it would be defocusing. Priced high enough, they could sub-contract to HP, for example, and still make a lot of money.
That said, I went to our local office the other day and poked my head into the networking closet. I see the same cheap crummy wifi routers I put there before our little company got bought. Right next to them is a Cisco router worth maybe $10-20K. It's worth more than all the computers and related hardware in the office combined.
Microsoft and Amazon pay an average of $30K/year for experienced programmers in India. If you want that same talent through an off-shore contracting company, it'll cost you double, or $60K/year. Idiots who think they can get $100K worth of development for $10K paid to an offshore contracting company deserve the failure they pay for. Even if you pay for good off-shore talent, your source code will be stolen, no matter what the contracting company tells you, and you'll risk competing with an offshore company that got it for free.
Off-shoring software development only makes sense in one case: open-source software development. In this case, your business model already takes into account the possibility of others using your source code to compete with you. If you're OK with that, just hire programmers directly through online sites designed for the purpose. $10K/year is a reasonable budget in this model, where if you're careful, you can get a great deal, while helping improve the life of some software developer in a place where $10K/year is a fantastic wage.
Bottom line: if keeping your source code private is important to you, hire locally. If you don't care, go open source and hire remotely.
I agree that he should get a 4-year CS degree. I'm old enough to remember when co-workers didn't have CS degrees because they weren't offered when they went to college, but now days I rarely see a coder in a decent job without a 4-year degree of some kind. I've hired mathematicians and physicists, because frankly they're often super smart, and coding is easy to teach. However, if you're not a genius, you can still be a valuable contributor, but I rarely see one without a CS or engineering college degree.
I'm not so sure about the death of the high-school drop out genius coder. While not a dropout, I admire kids like the 18-year-old behind the XBMC flavor of the Raspberry Pi OS. Someone with business experience should offer to cooperate on some startup with him. Even Steve Jobs had a mentor to guide him while building Apple. An attempt at building a startup right out of high school could be considered a postponement of college in order to gain real world experience. In an economy which is not generating enough jobs for new graduates, starting a company on your own is a great way to create one for yourself.
I probably read about Freicoin on slashdot. They've got a proposal for a fix to BitCoin that defeats hoarding. I'm pessimistic. The whole premiss of BitCoin is people don't trust each other. To "mine" bitcoins, we burn real resources to power all those machines computing SHA-1 hashes. Someone said that mining bitcoins was like burning real money and getting a fake bitcoins to replace it. No, "real" money is just stupid paper. It's amazing that we still use dead tree money. Burning it doesn't hurt the economy, other than reducing inflation. Wasting electricity and all those machines computing bitcoin hashes... that's real harm.
I know Ripple has not caught on, but I like the basic premiss. In Ripple, I create money by giving people I trust coupons with dollar values in proportion to how much I trust them. This gives us a way to generate cash out of nothing but trust, while eliminating incentives for hoarding.
The existing phone system is a dinosaur. We should switch to a modern digital P2P system where everyone has an online identity. The first time someone wants you to receive and e-mail from them, charge them $0.01. The first time they want you to answer their call, charge them $0.05. We need an electronic currency that enables fast micro-transactions, and we need to stop acting like the world is still plastered with individual analog phone lines rather than being all digital. Simply put, we need to take advantage of he capabilities of the hardware we already built.
Yep. You have to figure the author of the system wanted it to be deflationary, and he likely has at least a 100K himself if not a million. Apparently most coins are hoarded, not traded. At even $1/coin, the original guys have to have huge balls not to sell all their imaginary money for real money. At 22 coins, I figure it's just entertainment money. If someone hacks my computer and steal them, I'll be only a little pissed off. If they become worthless, oh well. Imagine what it's like to be sitting on 100K of them, when you got them for free? My guess is these guys must have a lot to do with regulating time price on Mt Gox.
I see the biggest threat to the price of a BTC as a new, better system. The world does need an electronic currency which is easily traded, but bitcoins are simply to hard to get hold of.
It's not the complexity that's increased exponentially, but the total amount of compute power applied world-wide to mine bitcoins. I mined about 20 coins about a year ago with a new graphics card, and then converted the machine to a new gaming PC for my son. Even then, it only made sense if:
1) You wanted a gaming machine with a powerful graphics card anyway.
2) You counted your effort to build the machine and do the mining as entertainment.
You can do it in Windows or Linux. It's not a way to make money, IMO, unless you've got access to free electricity, and hopefully cheap labor. The great halving will occur in December, at which point a lot of miners will pack up and go home.
The currency is deflationary, or at least it will be once new bitcoins can no longer be mined. If you lose your wallet, the bitcoins are gone forever. Over time, there will be fewer in circulation, yet more people in theory will be using them. That's OK, because a bitcoin can be broken into fractions of 1 in 100 million. So, if deflation causes the currency to increase in value, it wont be a problem until each bitcoin is worth about 1 cent times 100 million, or $1M each. There will eventually be 21 million bitcoins, at which point it's maximum total value before there's not enough fractions of coins to make penny sized transactions would be $2.1 quadrillion, which is more than 100X the US GDP. If the coins were distributed equally among 10 billion people, each would own 2,100 micro-bitcoins, each of which can be broken into 100 pieces, called a "satoshis" after the author of the system.
I figured I'm mine a few for fun and hope one day they'll be worth a lot. On a planet with 7 billion people, I'm guaranteed to have far more than average at 22 BTC. The average person will have probably around 0.002 BTC.
Back in the days of sharing source code on lists, before the dominance of the pre-compiled package, there was no binary incompatibility. With the advent of pre-compiled packages, we also both the rise of binary incompatibility and the rise of the package sponsors, the people who decide what's good enough for the rest of us to share. It used to be a bazaar, where I could lay out my blanket and sell the crud I'd created. Now it's a temple, where we all download from the True Source.
I've got a second-hand story to share about the good old days. At Berkeley, a friend of mine who was a Ph. D. student actually drove to Palo Alto now and then to attend the Home Built Computer Club meetings. There was a guy there with his latest hack that he wanted to share. He was selling it literally on a blanket. He had an un-stuffed circuit board and a copy of what parts to buy to stuff it, and claimed that if you did it, you'd have a working computer. His big idea at the time seemed to be selling empty circuit boards on a blanket. My friend said, "What a loser," and never thought of him again until that guy, Woz, co-founded Apple and changed the world.
The days of open collaboration between Linux developers has been hampered by binary incompatibility, and high hurdles to share software on popular software platforms like Debian and Fedora, and Gnome/GTK. We've seen hard feelings and fractures between groups like Ubuntu and Gnome, and lot's of unhappy users.
Are the days of freely sharing software on lists essentially in the past, or is there some way to once again pump life into that creative engine? Can we work smarter?
After Urban Terror, I had to let her play the usual kid games. She played Pokemon on her DS for years, and found Wizard 101, Roblox and Mindcraft, just like all the other kids. I wound up installing two Windows machines so the kids could play the games their friends were into. My son has gone off with Mindcraft, and built some insane worlds. My daughter is now a vegetarian into horses and animals. Both kids loved playing Star Craft against the computer with me, though I can't play it anymore due to vision issues. Partly because I suck so badly now days, both kids love playing Halo with me. They call it "kill dad," as if that were the name of the game, rather than Halo.
Few of you will agree with this one, but I let my daughter play it from 2 to 5 years old. She loved it more than any other game. It was hysterical. If you know this game, it's an awesome first-person shooter with guns, grenades, and blood, played on various urban maps. Last I played, it was mostly very skilled diehards with years of experience. She was slow compared to the other players, so she compensated by using some insane machine gun used for close range (IMI Negev). Her favorite thing was killing guys on the other team, but that was really hard because of their skill and her age. She developed unique strategies. For example, on one map she would smoke grenade her own team's flag, and sit just outside the smoke waiting for the other team to appear with the flag, and then lay into them with the negev. She had a few other strategies, but they were always ambushes to compensate for her low speed and accuracy. She loved grenades and that big inaccurate machine gun.
I never had the heart to IM the other players that a they'd just been killed by a four year old girl. I also never worried that her goal in the game was rarely protecting the flag or trying to get the other team's flag. Sometimes it was just wondering around, looking for a good place to ambush a player. Sometimes she just wanted to see what all the different guns do, or simply explore the map. Even at her best, her kill ratio was awful, maybe 1:4, but she didn't mind the dieing, though if she was trying to do something it frustrated her to have to start over. Most of her ambushes had to be set up very close to home base, or if she wanted just to play around, she'd find a mostly useless part of the map rarely visited by the other players.
I had to end her time playing Urban Terror when she began to read. I never told her what those IM messages scrolling past said, but they were pretty horrible. She wasn't exactly a good team player, and some of these guys take the game waaaaay to seriously. My son is 2 years younger, and I only let him play it about a year, because my daughter would always watch and coach him, so her starting to read IM messages became a problem. She would ask why she couldn't play anymore, but at that age such questions are easily deflected, just like whether or not Santa Clause is real.
The process you're describing is only acceptable to people who get paid while going through that process. Tell me of one example of someone who just wanted to do the right thing by informing the world, who actually went through that process. Someone currently alive.