arguments I've heard - never built one to scale rebuttal - it doesn't matter - build 1000 tiny ones instead if big ones don't work.
- continuous reprocessing has never been tested and may be impossible rebuttal - you don't know unless you try, and it seems feasible.
- they still spit out the same long half-life, long decay elements as conventional reactors rebuttal - most of these can be reused or salvaged for medical devices, and it burns 97% of its fuel instead of 3% or less. Also you will find almost as much naturally occurring "waste" where the Thorium came in the first place. Here is a breakdown from http://energyfromthorium.com/lftradsrisks.html:
Additionally, because LFTR burns all of its nuclear fuel, the majority of the waste products (83%) are safe within 10 years, and the remaining waste products (17%) need to be stored in geological isolation for only about 300 years (compared to 10,000 years or more for LWR waste). Additionally, the LFTR can be used to "burn down" waste from an LWR (nearly the entirety of the United States' nuclear waste stockpile) into the standard waste products of an LFTR, so long-term storage of nuclear waste would no longer be needed.
read that again - can be used to "burn down" waste from an LWR - so in addition, we can get rid of a lot of the waste from the inefficient reactors we have.
- they are really Uranium reactors and they require a seed reaction rebuttal - true reactors like this are Uranium - they convert Thorium to Uranium and then split, however the base fuel is still Thorium and the seed can be reused. It is also possible to continuously feed them if the equipment can filter out impurities. No physical research has been done here.
- Thorium is uneconomic, and costs far more than Uranium rebuttal - Thorium is much more plentiful than Uranium, easier to mine and therefore if a market emerged, would likely drop from current ~$5000/kg to potentially $10/kg or less. That is compared to enriched Uranium, which is over $1600/kg after an expensive processing and/or reprocessing. Total cost of operations is also much less - estimated at 30-50% of a LWR.
- Thorium is bad for selling weapons grade elements to the government and charging massive reprocessing fees and kickbacks that line the back pockets of reactor owners. um, exactly.
That is what you call "statistics," or more specifically, bullshit numbers pulled out of their ass to make them look worse than they are. The worst piracy number I have ever heard of was a little over 90%, and that was from an indie game with no protection. In the United States or most of western Europe there is no way they can back up that number, and that will be where 99% of their profit will be made anyway. Skip translating to Chinese, Armenian, and other high piracy languages and problem solved (if they play a pirated version, no big deal - no investment was made to cater to them).
This easily falls under the CFAA in the United States, but so does practically anything, like, say lying about your weight on a dating site (seriously - there was an article about it on the Register yesterday as of this writing). I'm sure hacking programs are also covered in an over broad way on that law.
And of course United States laws apply to everyone... but I can see Romanian authorities bowing to the whims of the United States - if the US has a friend in Europe, it is Romania. When I was there about the only anti-US thing I saw was a little kid with a CCCP T-shirt. Aside from that, I saw a lot of EU and US help, and well needed at that - the Soviet era road system was in a pretty bad state, and it was easy to tell the old trains from the new. Not sure how much has changed since then.
Not to mention that Sony has already announced their next PSP (current? - I didn't really care and my magic 8 ball says "reply hazy, try again") will be as powerful as a PS3. PS3 runs on a modified and upgraded nVidia 7800 architecture, which dates it to about 2005 tech-wise, and a 6 year old GPU is ancient in GPU years. This is before nVidia had unified shaders, so there are 8 vertex and 24 pixel shaders (or 32 total shaders) at 550MHz - I think that's entirely eclipse-able in mobile in a couple of years.
Anyhow, we'll have to see how Mali holds up to Kal El (nVidia's Tegra 3), which is already reportedly outperforming Mali on tablets (but still behind PowerVR, but I don't know the configuration tested, I just remember seeing it was slower than iPad2) and should be in phones soon. I'm personally more interested in Wayne... OpenGL 4.X and DX11 in mobile. Still a bit on the short side graphics-core wise (max of 64), but hardware tessellation is fun.
Corruption depends a lot on perception - if a senator accepts 2 million dollars of Super PAC money for their Presidential election campaign if they vote for big pharm, is that corruption, or working the system? Big pharm doesn't think so, and it is perfectly legal in the US. I don't expect such practices will ever die in the US under the current system, because what rich person wouldn't vote for a policy that either gives them more money or pays them for more media face time?
Personally I find PACs as a form of legalized bribery, but not everyone thinks so - especially rich politicians who get to vote on such things.
yeah, and in my filler classes the profs at a "research institution" usually didn't even bother to show up and had their TAs teach the class, even though they were supposed to only be assisting the professor. Apparently (and I found this out years later) TAs teaching a class needed to be listed in the class bulletin, and a couple of years after I graduated, a scandal broke about the profs not teaching. Too bad they had fired two of my favorite teachers the year before the scandal broke because those teachers didn't do enough research and didn't publish enough papers. At least the department head got demoted (he encouraged the practice in the department, but it was found to be against school rules, and he also was a rat-bastard IMO - he always talked down to students and whenever I left a chat with him I had a barely controllable urge to punch him in the face), but unfortunately, the worst teacher I had (he did show up, but I didn't have him until year 3) became department head and kept the research first students second policy. At least my user interface and design professors survived the cuts (both published a lot of papers and were great teachers, which is a rare combo) - too bad my computer graphics and C/C++ professors did not (both were awesome teachers).
Sometimes you do - the top two students at my first college didn't do drugs or alcohol, maintained a 4.0 GPA (and yes, that was max), and graduated with highest honors in 4 years in a program that usually takes 5 (to graduate in 4 you needed to carry a 24-28 credit load - it took me 6 1/2, but I dropped out a year for financial reasons and switched schools, so 5 1/2 years were in school). They happened to be my extremely religious roommates in a 4 person apartment, one pretty much didn't sleep (he slept like Edison - about 2 hours every night with 3-4 15 minute naps - he also worked 3 jobs/140 hours a week when not in school) and the other had an Eidetic memory and studied 15 minutes by paging through the books where the rest of us studied for hours.
My other roommate was a bum, thankfully - there is only so much squeaky-clean farm bred Christianity + conservative Republicanism I can take. My next batch of roommates drove me equally nuts - a militant vegan and two pot smoking (and LSD, shrooms, research chemicals, etc) hippies. There is only so many times I can take yellow sticky notes with "meat is murder" on my steaks (and I suppose I provoked her a few times by putting little stickies saying "RIP - poor leafies"). Fortunately for me I didn't live in that place for long before I moved (several times, in fact, before I settled down - about 6 places in 2 years).
Parsing is nice, but Eliza did natural language parsing and that was 30+ years ago. What I want is logical deduction, and that is where AIs tend to fail the Turing test for me. Something like this: Me: Siri, I was thinking about chainsawing my cat in half. Siri: That would not be healthy for your cat. Me: I'm just kidding, I don't have a cat, in fact, I'm deathly allergic to them, but I do have this phone, and I could download a picture of a cat. Siri: Please don't saw me in half.
Any time I've used this sort of logic with a supposed Turing compliant AI it has failed. Also they tend to fail when I put absurd or suicidal situations up - ever gone skydiving without a parachute? (answer no). You should - it is an absolute rush (answer OK, I'll try it). Ever gone cliff diving onto jagged rocks? (answer no). You should - it is an absolute rush (OK, I will).
Microsoft's is no better - I was laughing at this preview message on my exchange voice mail "Holy tangerines trinity four feathers office for one tell cell"
actual call was something like Hi it's Tony [last name] you're not in your office trying your cell.
yes he has a (mild) Italian accent, but that is ridiculous.
What you don't seem to grasp is that there are more than one type of fission reactor, probably because the only one that people are told about are the big, squeezed conical towers call fast breeder pressurized water reactors (PWR).
PWRs were designed with the following priorities: 1) nuclear weapons (governments hand the utilities reams of money) 2) make lots of money (re)processing uranium (utilities hand themselves reams of money and call it a "cost") 3) electrical power (and thus more money, but the industry is regulated, so it isn't hand over fist like the above two). 4) safety (costs money, so the less spent the better - just like BP cutting corners on their oil rigs - the lower the cost, the more the profit)
and yet still they cost far less lives than coal and oil.
But if you knew anything about other types of fission reactor like the molten salt reactor experiment from the 1960s or liquid fluoride reactors (LFTR mentioned elsewhere, a modern take on MSR), you'd know we could build scalable, passively safe, self regulating, raw fuel burning reactors for electrical power that are really bad at #1 and don't need #2 and thus utilities don't want to invest any money in the technology because #1 and #2 are cash cows.
Of course, there are naysayers - http://www.guardian.co.uk/environment/2011/jun/23/thorium-nuclear-uranium The author of this was obviously biased against or chose biased views 1) nobody has ever tried to build a thorium molten salt reactor on a larger scale, but even if it ended up not scaling, many small nuclear reactors can be used much closer to residential areas. They don't spew radiation when their container is breached and they don't melt down. They also can be shut off and restarted easily, as Oak Ridge used to do with theirs on the weekends. 2) they will be uneconomic - this is a chicken or egg problem - yes Thorium currently costs about $5000/kg vs $40/kg for Uranium, but it currently is a novelty metal and not mined heavily like Uranium (in fact, it is often buried as waste). It is as abundant as lead and 4x more abundant than Uranium. It also doesn't need refining like Uranium and burns much more efficiently. 3) environmentalists say - "its reactors disgorging the same toxic byproducts and fissile waste with the same millennial half-lives" - hardly - if 97% of it is burnt up, there is 94-96% less nuclear waste than traditional reactors which burn.7%-3% of their fuel. 4) U232 byproduct - I'm not a nuclear scientist, but as I understand it, U233->U232->U233 is the reaction cycle and thus is self recycling. Leftovers can be separated from any other remaining byproducts chemically (you can't separate U233 and U233, but you can other byproducts). U233 mixed with depleted uranium creates natural uranium in time.
Just in case people don't want to look it up or don't know it already, it is Liquid Fluoride Thorium Reactor (LFTR), which is an update of the Molten Salt Reactor (MSR) built in the 1960s and abandoned.
What's not to love about a reactor that burns 97% of its fuel rather than.7-3%, self-regulates, won't melt down, is passively safe, burns raw fuel rather than refined fuel, scales from about 100MW up, and doesn't create very much usable nuclear weapon material? Oh - I guess the bad is it doesn't generate much nuclear weapon material (at least that was the rationale in the 1960s).
Some other Gen IV designs are interesting, some burn nuclear waste, for instance (like the one Bill Gates is funding in Russia according to TED a couple of years ago). VHTR (Very High Temperature Reactor) sounded promising to me as well, also being passively safe, but the team experimenting with it hit some technical problems and South Africa pulled funding.
It would not surprise me if Microsoft has in-house versions for other platforms, as part of their requirements used to be (and maybe still are) for the API to work on multiple kernels. Microsoft had Windows built on other platforms including supposedly PowerPC back when they were supporting Alpha NT (reportedly little-endian, however, the PPC chip being bi-endian), though I haven't heard much recently. It would not surprise me if they've also built Windows 7/8 (full) on ARM.
Well, Portugal did become the first European country to fully decriminalize drugs for personal use (they are still illegal), and from what I heard that was very successful and drug use is way down, so sometimes taking a gamble on a wild card pays huge dividends. Maybe OSS is that wild card once again...
CS? No, no, no, that isn't fatal enough - use VX... actually no, what you want the unidentified gas from Return of the Living Dead that raises the zombies. You can't just be SWAT on Halloween, you need to be psycho SWAT.
I remember several features of languages like Lisp being hard (if not impossible) to implement without a runtime interpreter. I believe garbage collection (which was also invented by Lisp inventor John McCarthy), was one of them, though I'm a bit foggy on the details of why. I know gc exists for C/C++, but not for all data types (notably unions) and it tends to be unreliable. Bjarne Stroustrup doesn't think garbage collection is necessary for the most part, though he acknowledges there are some that could use it, and said in an interview that C++ is the least garbage generating language (because the programmer cleans it up?).
Despite my dislike of Lisp as a language, which I call parenthesis hell (In my AI class I had a function with ~85 close parenthesis on a 40 column display due to a large nested function for God's sake...) I admit there are many features I like, but I got swept up in the OOP bandwagon when I learned Smalltalk and C++ at about the same time (and C++ got just about everything wrong, but it was much faster and that is what the world wanted at the time).
yes, exactly - I can do voice search by holding down the ? on my android phone, and have been able to do that since I got it a year ago. I can also tell the voice search to use a certain app (i.e. "navigate to the White House"). I can't, however, tell it I'm drunk and get cab services. I can, however, just call a cab, so IMO the feature is more amusement than value add.
Siri is definitely much more sassy than my android phone. I can't ask it "Do you love me" and get a response "I'm not allowed to, Dickhead"
Not sure about other carriers, but in the US on Verizon you can't even tether one device without paying more than your whole plan ($30 per device, I believe, which is ~22 Euro right now).
Add in $20 for texting, $10-30 for internet (depending on use, unlimited is no longer available from the top 3 carriers), and a minimum of $30 for basic phone service and you can see the ass fscking we get in America.
Yes, there are better plans if you don't ever leave a city center, but I travel to areas that are not covered by anyone but Verizon (and yes, I tested both Sprint and AT&T, and neither got reception at my grandparent's house, and AT&T barely got reception along the main road near there, which is a major road, but not a freeway). Verizon got 5 bars...
IMO, Thunderbolt is mainly Intel attempting to replace PCI express - it has bi-directional 10Gbps throughput with a potential of 100Gbps, which is faster than current PCI express 2.0 16x slots (but it looks like it will be slower than the upcoming PCIe 3.0). Intel is a member of the PCI-SIG, but they don't own the technology - it is owned by a 900+ member non-profit, I believe.
As for USB 3.0, no _single_ drive can saturate it, but a striped RAID array can easily saturate it, which is why in demos of thunderbolt Intel saturated a striped RAID array while showing full video to show that the video didn't slow down and the files were transferred. Intel also says not to expect more than about 3.2Gbps throughput for USB 3.0, and that is about what fast SSDs can handle (I've seen SSDs with read/write over 500MB/sec, but the fastest platter HDDs were about 250MB/sec last time I checked [at least a year ago, but I don't think the tech has advanced much] and they would need to be 400MB/sec to saturate 3.2Gbps - note for the less technical that I use b for bits and B for bytes, as is proper, where 8 bits=1 byte).
By Thunderbolt committee, do you mean Intel? They developed and own the technology and AFAIK, nobody else had any input. It was brought to market in a collaboration with Apple, but that was mainly for Apple's mini DisplayPort technology, which Intel thinks will help broaden adoption because it is well suited to laptops.
10 billion USB connectors wouldn't surprise me in the least - my PC has 8, my wife's PC 6, my old PC, my 10 year old mac and my laptop 4 each, the add-on card for my old PC 4 (USB 2.0 vs 1.0 because the 1.0 controller flaked out on the Mobo, but everything else worked), my cell phone 1, my wife's cell phone 1, each of my last 3 cell phones 1 each, each of my four cameras 1 each, my satellite TV 1... I probably missed many more - USB is basically being shoved into everything.
I don't know if Sony used a different plug or not, but iLink is Sony not wanting to pay to use the brand name Firewire, which was owned by Apple - both are IEEE1394.
We don't mind you protesting Jobs' funeral if you don't mind us making a ceremony of pouring piss and vinegar on your grave when you die... I'm sure your loved ones would be insulted if you had any.
You don't even need to get to the Bible Belt for that.
One of the reasons I dislike organized religion is because I went to my cousin's Catholic wedding in South Dakota and after the 3 hour ceremony including a full mass (or as I would call it, church service) and lots of nasty smelling incense, I was pulled aside into a room by the priest because he saw that I didn't take communion (several other heathens like me were there, but I think I was maybe seated in the most visible spot, or maybe he saw me as an impressionable teen at the time). He asked me why I hadn't taken communion. I said that I felt it was wrong to take communion if I wasn't Catholic. He then asked me several questions on why I was not Catholic and started to go down the Catholicism is the only way blah-blah-blah rhetoric, but at one point he started attacking my beliefs and that is where I drew the line - he insisted that my God was false and that I needed to repent and convert today or that I was damned to hell, that I would burn in eternal fire, and Satan would consume my soul (actually, it was even more illustrious and long winded, but you get the gist). In response I said something I probably shouldn't have like "Jesus is so stuffed up your ass he couldn't see Satan if he tried" and he got really mad - I pretty much got dragged and tossed out the door. My parents found me in the car a couple of hours later and I told them what happened but I think they though I provoked him rather than he provoked me.
I remember him saying never to come back, but in hindsight I don't think he really said that (I think he just slammed the door after tossing me out) and I probably embellished it in when telling the story later (it makes for a rather nice story). Incidentally, I did come back a few years later for my uncle's funeral and I learned that priest had died of a heart attack a few years before. I didn't see the new priest attacking anyone's beliefs, so he may be more tolerant, or maybe a funeral is less appropriate for recruiting than a wedding.
For the grandparent, Wozniak is and always will be the inventor of the Apple computer innards, but Jobs gave it the appliance like form factor. Jobs wanted it completely sealed, but Wozniak insisted that the cover be removable for expansions and tinkering, something he got until Jobs got his way with the Mac (or maybe Lisa - I never got to play with one of those). Gates and Microsoft mainly created the most widely used BASIC at the time, even selling it to Apple (aka Applesoft BASIC, whereas Woz's BASIC was thenceforward Integer BASIC). As far as Microsoft's OS, they didn't really get into that market until IBM paid them to get into it, and even then, they bought the software rather than make it in-house. Jobs also gave us NeXT, which is the platform where I used my first graphical web browser. And Jobs bought the Graphics Group from Lucasfilm and founded Pixar with them, and thus creating all three of my nephew's favorite film at least for part of their young lives, Cars.
As far as the parent's suggestions, Commodore came out with their PET computer about a year after the Apple ][, with an equally frustrating and horrifically failure prone tape drive compared to Apple's (1978's Disk ][ had me sold on Apple's for years 0 Commodore's 1541 that came out a couple of years later was much cheaper, but also 10x slower). Atari didn't release a computer until 1979 with the Atari 400, which had this awful membrane keyboard (but I wanted one so bad at the time, being a huge 2600 fan). Commodore's mainstream computer, the Vic 20, didn't come out until 1980, but it was a step into the stone ages if you'd used an Apple ][ (back to tape drive and 20+ minute loads that sometimes took 3-4 tries to load until they came out with the 1541 late that year, tiny memory, underpowered graphics, etc), but it was also about 10x cheaper. The C64 is where Commodore really started to shine, but the 1541's load times still drove me nuts. IBM was last to bat, and actually failed in the market, lasting only 5 years (sure their design won in the market, but they started making computers in late 1981 and stopped manufacturing PCs by about 1986-7), but their floppy drive speeds were decent. Many years later I had a similar experience with 1x optical drives loading a CD ROM (20 minutes for Fallout to load a level was typical - when I upgraded to a 4x, it loaded in about 2 minutes, so it wasn't all a drive speed issue).
Slashdot Mirror
arguments I've heard
- never built one to scale
rebuttal - it doesn't matter - build 1000 tiny ones instead if big ones don't work.
- continuous reprocessing has never been tested and may be impossible
rebuttal - you don't know unless you try, and it seems feasible.
- they still spit out the same long half-life, long decay elements as conventional reactors :
rebuttal - most of these can be reused or salvaged for medical devices, and it burns 97% of its fuel instead of 3% or less. Also you will find almost as much naturally occurring "waste" where the Thorium came in the first place. Here is a breakdown from http://energyfromthorium.com/lftradsrisks.html
read that again - can be used to "burn down" waste from an LWR - so in addition, we can get rid of a lot of the waste from the inefficient reactors we have.
- they are really Uranium reactors and they require a seed reaction
rebuttal - true reactors like this are Uranium - they convert Thorium to Uranium and then split, however the base fuel is still Thorium and the seed can be reused. It is also possible to continuously feed them if the equipment can filter out impurities. No physical research has been done here.
- Thorium is uneconomic, and costs far more than Uranium
rebuttal - Thorium is much more plentiful than Uranium, easier to mine and therefore if a market emerged, would likely drop from current ~$5000/kg to potentially $10/kg or less. That is compared to enriched Uranium, which is over $1600/kg after an expensive processing and/or reprocessing. Total cost of operations is also much less - estimated at 30-50% of a LWR.
- Thorium is bad for selling weapons grade elements to the government and charging massive reprocessing fees and kickbacks that line the back pockets of reactor owners.
um, exactly.
That is what you call "statistics," or more specifically, bullshit numbers pulled out of their ass to make them look worse than they are. The worst piracy number I have ever heard of was a little over 90%, and that was from an indie game with no protection. In the United States or most of western Europe there is no way they can back up that number, and that will be where 99% of their profit will be made anyway. Skip translating to Chinese, Armenian, and other high piracy languages and problem solved (if they play a pirated version, no big deal - no investment was made to cater to them).
This easily falls under the CFAA in the United States, but so does practically anything, like, say lying about your weight on a dating site (seriously - there was an article about it on the Register yesterday as of this writing). I'm sure hacking programs are also covered in an over broad way on that law.
And of course United States laws apply to everyone... but I can see Romanian authorities bowing to the whims of the United States - if the US has a friend in Europe, it is Romania. When I was there about the only anti-US thing I saw was a little kid with a CCCP T-shirt. Aside from that, I saw a lot of EU and US help, and well needed at that - the Soviet era road system was in a pretty bad state, and it was easy to tell the old trains from the new. Not sure how much has changed since then.
Not to mention that Sony has already announced their next PSP (current? - I didn't really care and my magic 8 ball says "reply hazy, try again") will be as powerful as a PS3. PS3 runs on a modified and upgraded nVidia 7800 architecture, which dates it to about 2005 tech-wise, and a 6 year old GPU is ancient in GPU years. This is before nVidia had unified shaders, so there are 8 vertex and 24 pixel shaders (or 32 total shaders) at 550MHz - I think that's entirely eclipse-able in mobile in a couple of years.
Anyhow, we'll have to see how Mali holds up to Kal El (nVidia's Tegra 3), which is already reportedly outperforming Mali on tablets (but still behind PowerVR, but I don't know the configuration tested, I just remember seeing it was slower than iPad2) and should be in phones soon. I'm personally more interested in Wayne... OpenGL 4.X and DX11 in mobile. Still a bit on the short side graphics-core wise (max of 64), but hardware tessellation is fun.
Corruption depends a lot on perception - if a senator accepts 2 million dollars of Super PAC money for their Presidential election campaign if they vote for big pharm, is that corruption, or working the system? Big pharm doesn't think so, and it is perfectly legal in the US. I don't expect such practices will ever die in the US under the current system, because what rich person wouldn't vote for a policy that either gives them more money or pays them for more media face time?
Personally I find PACs as a form of legalized bribery, but not everyone thinks so - especially rich politicians who get to vote on such things.
yeah, and in my filler classes the profs at a "research institution" usually didn't even bother to show up and had their TAs teach the class, even though they were supposed to only be assisting the professor. Apparently (and I found this out years later) TAs teaching a class needed to be listed in the class bulletin, and a couple of years after I graduated, a scandal broke about the profs not teaching. Too bad they had fired two of my favorite teachers the year before the scandal broke because those teachers didn't do enough research and didn't publish enough papers. At least the department head got demoted (he encouraged the practice in the department, but it was found to be against school rules, and he also was a rat-bastard IMO - he always talked down to students and whenever I left a chat with him I had a barely controllable urge to punch him in the face), but unfortunately, the worst teacher I had (he did show up, but I didn't have him until year 3) became department head and kept the research first students second policy. At least my user interface and design professors survived the cuts (both published a lot of papers and were great teachers, which is a rare combo) - too bad my computer graphics and C/C++ professors did not (both were awesome teachers).
Sometimes you do - the top two students at my first college didn't do drugs or alcohol, maintained a 4.0 GPA (and yes, that was max), and graduated with highest honors in 4 years in a program that usually takes 5 (to graduate in 4 you needed to carry a 24-28 credit load - it took me 6 1/2, but I dropped out a year for financial reasons and switched schools, so 5 1/2 years were in school). They happened to be my extremely religious roommates in a 4 person apartment, one pretty much didn't sleep (he slept like Edison - about 2 hours every night with 3-4 15 minute naps - he also worked 3 jobs/140 hours a week when not in school) and the other had an Eidetic memory and studied 15 minutes by paging through the books where the rest of us studied for hours.
My other roommate was a bum, thankfully - there is only so much squeaky-clean farm bred Christianity + conservative Republicanism I can take. My next batch of roommates drove me equally nuts - a militant vegan and two pot smoking (and LSD, shrooms, research chemicals, etc) hippies. There is only so many times I can take yellow sticky notes with "meat is murder" on my steaks (and I suppose I provoked her a few times by putting little stickies saying "RIP - poor leafies"). Fortunately for me I didn't live in that place for long before I moved (several times, in fact, before I settled down - about 6 places in 2 years).
Parsing is nice, but Eliza did natural language parsing and that was 30+ years ago.
What I want is logical deduction, and that is where AIs tend to fail the Turing test for me. Something like this:
Me: Siri, I was thinking about chainsawing my cat in half.
Siri: That would not be healthy for your cat.
Me: I'm just kidding, I don't have a cat, in fact, I'm deathly allergic to them, but I do have this phone, and I could download a picture of a cat.
Siri: Please don't saw me in half.
Any time I've used this sort of logic with a supposed Turing compliant AI it has failed. Also they tend to fail when I put absurd or suicidal situations up - ever gone skydiving without a parachute? (answer no). You should - it is an absolute rush (answer OK, I'll try it). Ever gone cliff diving onto jagged rocks? (answer no). You should - it is an absolute rush (OK, I will).
Microsoft's is no better - I was laughing at this preview message on my exchange voice mail
"Holy tangerines trinity four feathers office for one tell cell"
actual call was something like
Hi it's Tony [last name] you're not in your office trying your cell.
yes he has a (mild) Italian accent, but that is ridiculous.
What you don't seem to grasp is that there are more than one type of fission reactor, probably because the only one that people are told about are the big, squeezed conical towers call fast breeder pressurized water reactors (PWR).
PWRs were designed with the following priorities:
1) nuclear weapons (governments hand the utilities reams of money)
2) make lots of money (re)processing uranium (utilities hand themselves reams of money and call it a "cost")
3) electrical power (and thus more money, but the industry is regulated, so it isn't hand over fist like the above two).
4) safety (costs money, so the less spent the better - just like BP cutting corners on their oil rigs - the lower the cost, the more the profit)
and yet still they cost far less lives than coal and oil.
But if you knew anything about other types of fission reactor like the molten salt reactor experiment from the 1960s or liquid fluoride reactors (LFTR mentioned elsewhere, a modern take on MSR), you'd know we could build scalable, passively safe, self regulating, raw fuel burning reactors for electrical power that are really bad at #1 and don't need #2 and thus utilities don't want to invest any money in the technology because #1 and #2 are cash cows.
Unfortunately, this is not a thorium molten salt reactor like the proposed LFTR and it is more like traditional reactors, but China is working on them, and there are test reactors being built in Japan and Europe (see here for China - http://www.telegraph.co.uk/finance/comment/ambroseevans_pritchard/8393984/Safe-nuclear-does-exist-and-China-is-leading-the-way-with-thorium.html ).
Of course, there are naysayers - http://www.guardian.co.uk/environment/2011/jun/23/thorium-nuclear-uranium .7%-3% of their fuel.
The author of this was obviously biased against or chose biased views
1) nobody has ever tried to build a thorium molten salt reactor on a larger scale, but even if it ended up not scaling, many small nuclear reactors can be used much closer to residential areas. They don't spew radiation when their container is breached and they don't melt down. They also can be shut off and restarted easily, as Oak Ridge used to do with theirs on the weekends.
2) they will be uneconomic - this is a chicken or egg problem - yes Thorium currently costs about $5000/kg vs $40/kg for Uranium, but it currently is a novelty metal and not mined heavily like Uranium (in fact, it is often buried as waste). It is as abundant as lead and 4x more abundant than Uranium. It also doesn't need refining like Uranium and burns much more efficiently.
3) environmentalists say - "its reactors disgorging the same toxic byproducts and fissile waste with the same millennial half-lives" - hardly - if 97% of it is burnt up, there is 94-96% less nuclear waste than traditional reactors which burn
4) U232 byproduct - I'm not a nuclear scientist, but as I understand it, U233->U232->U233 is the reaction cycle and thus is self recycling. Leftovers can be separated from any other remaining byproducts chemically (you can't separate U233 and U233, but you can other byproducts). U233 mixed with depleted uranium creates natural uranium in time.
Just in case people don't want to look it up or don't know it already, it is Liquid Fluoride Thorium Reactor (LFTR), which is an update of the Molten Salt Reactor (MSR) built in the 1960s and abandoned.
What's not to love about a reactor that burns 97% of its fuel rather than .7-3%, self-regulates, won't melt down, is passively safe, burns raw fuel rather than refined fuel, scales from about 100MW up, and doesn't create very much usable nuclear weapon material? Oh - I guess the bad is it doesn't generate much nuclear weapon material (at least that was the rationale in the 1960s).
Some other Gen IV designs are interesting, some burn nuclear waste, for instance (like the one Bill Gates is funding in Russia according to TED a couple of years ago). VHTR (Very High Temperature Reactor) sounded promising to me as well, also being passively safe, but the team experimenting with it hit some technical problems and South Africa pulled funding.
also near Seattle - just read an article about arena net (Guild Wars 2) recording audio there ( http://www.arena.net/blog/video-audio-team-field-recording-trip )
It would not surprise me if Microsoft has in-house versions for other platforms, as part of their requirements used to be (and maybe still are) for the API to work on multiple kernels. Microsoft had Windows built on other platforms including supposedly PowerPC back when they were supporting Alpha NT (reportedly little-endian, however, the PPC chip being bi-endian), though I haven't heard much recently. It would not surprise me if they've also built Windows 7/8 (full) on ARM.
Well, Portugal did become the first European country to fully decriminalize drugs for personal use (they are still illegal), and from what I heard that was very successful and drug use is way down, so sometimes taking a gamble on a wild card pays huge dividends. Maybe OSS is that wild card once again...
CS? No, no, no, that isn't fatal enough - use VX... actually no, what you want the unidentified gas from Return of the Living Dead that raises the zombies. You can't just be SWAT on Halloween, you need to be psycho SWAT.
I remember several features of languages like Lisp being hard (if not impossible) to implement without a runtime interpreter. I believe garbage collection (which was also invented by Lisp inventor John McCarthy), was one of them, though I'm a bit foggy on the details of why. I know gc exists for C/C++, but not for all data types (notably unions) and it tends to be unreliable. Bjarne Stroustrup doesn't think garbage collection is necessary for the most part, though he acknowledges there are some that could use it, and said in an interview that C++ is the least garbage generating language (because the programmer cleans it up?).
Despite my dislike of Lisp as a language, which I call parenthesis hell (In my AI class I had a function with ~85 close parenthesis on a 40 column display due to a large nested function for God's sake...) I admit there are many features I like, but I got swept up in the OOP bandwagon when I learned Smalltalk and C++ at about the same time (and C++ got just about everything wrong, but it was much faster and that is what the world wanted at the time).
yes, exactly - I can do voice search by holding down the ? on my android phone, and have been able to do that since I got it a year ago. I can also tell the voice search to use a certain app (i.e. "navigate to the White House"). I can't, however, tell it I'm drunk and get cab services. I can, however, just call a cab, so IMO the feature is more amusement than value add.
Siri is definitely much more sassy than my android phone. I can't ask it "Do you love me" and get a response "I'm not allowed to, Dickhead"
Not sure about other carriers, but in the US on Verizon you can't even tether one device without paying more than your whole plan ($30 per device, I believe, which is ~22 Euro right now).
Add in $20 for texting, $10-30 for internet (depending on use, unlimited is no longer available from the top 3 carriers), and a minimum of $30 for basic phone service and you can see the ass fscking we get in America.
Yes, there are better plans if you don't ever leave a city center, but I travel to areas that are not covered by anyone but Verizon (and yes, I tested both Sprint and AT&T, and neither got reception at my grandparent's house, and AT&T barely got reception along the main road near there, which is a major road, but not a freeway). Verizon got 5 bars...
Actually, the remake is back on I hear, but so far only this guy is known to be cast: http://en.wikipedia.org/wiki/Ryan_Gosling and the director is now http://en.wikipedia.org/wiki/Nicolas_Winding_Refn - too bad, I had hope when Brian Singer was attached to direct, not so sure about Winding Refn.
There still is hope for Selina and Justin as I don't believe they've started casting (and I know it's what you secretly want ;)
IMO, Thunderbolt is mainly Intel attempting to replace PCI express - it has bi-directional 10Gbps throughput with a potential of 100Gbps, which is faster than current PCI express 2.0 16x slots (but it looks like it will be slower than the upcoming PCIe 3.0). Intel is a member of the PCI-SIG, but they don't own the technology - it is owned by a 900+ member non-profit, I believe.
As for USB 3.0, no _single_ drive can saturate it, but a striped RAID array can easily saturate it, which is why in demos of thunderbolt Intel saturated a striped RAID array while showing full video to show that the video didn't slow down and the files were transferred. Intel also says not to expect more than about 3.2Gbps throughput for USB 3.0, and that is about what fast SSDs can handle (I've seen SSDs with read/write over 500MB/sec, but the fastest platter HDDs were about 250MB/sec last time I checked [at least a year ago, but I don't think the tech has advanced much] and they would need to be 400MB/sec to saturate 3.2Gbps - note for the less technical that I use b for bits and B for bytes, as is proper, where 8 bits=1 byte).
By Thunderbolt committee, do you mean Intel? They developed and own the technology and AFAIK, nobody else had any input. It was brought to market in a collaboration with Apple, but that was mainly for Apple's mini DisplayPort technology, which Intel thinks will help broaden adoption because it is well suited to laptops.
Sony, OTOH, appears to be implementing the Intel prototype from a couple of years ago, which is essentially a modified USB 3.0 (or very early 4.0?) with an added optical channel for Thunderbolt. See http://www.engadget.com/2009/09/24/video-intels-light-peak-running-an-hd-display-while-transferri/
10 billion USB connectors wouldn't surprise me in the least - my PC has 8, my wife's PC 6, my old PC, my 10 year old mac and my laptop 4 each, the add-on card for my old PC 4 (USB 2.0 vs 1.0 because the 1.0 controller flaked out on the Mobo, but everything else worked), my cell phone 1, my wife's cell phone 1, each of my last 3 cell phones 1 each, each of my four cameras 1 each, my satellite TV 1... I probably missed many more - USB is basically being shoved into everything.
I don't know if Sony used a different plug or not, but iLink is Sony not wanting to pay to use the brand name Firewire, which was owned by Apple - both are IEEE1394.
No need for violence -
We don't mind you protesting Jobs' funeral if you don't mind us making a ceremony of pouring piss and vinegar on your grave when you die... I'm sure your loved ones would be insulted if you had any.
You don't even need to get to the Bible Belt for that.
One of the reasons I dislike organized religion is because I went to my cousin's Catholic wedding in South Dakota and after the 3 hour ceremony including a full mass (or as I would call it, church service) and lots of nasty smelling incense, I was pulled aside into a room by the priest because he saw that I didn't take communion (several other heathens like me were there, but I think I was maybe seated in the most visible spot, or maybe he saw me as an impressionable teen at the time). He asked me why I hadn't taken communion. I said that I felt it was wrong to take communion if I wasn't Catholic. He then asked me several questions on why I was not Catholic and started to go down the Catholicism is the only way blah-blah-blah rhetoric, but at one point he started attacking my beliefs and that is where I drew the line - he insisted that my God was false and that I needed to repent and convert today or that I was damned to hell, that I would burn in eternal fire, and Satan would consume my soul (actually, it was even more illustrious and long winded, but you get the gist). In response I said something I probably shouldn't have like "Jesus is so stuffed up your ass he couldn't see Satan if he tried" and he got really mad - I pretty much got dragged and tossed out the door. My parents found me in the car a couple of hours later and I told them what happened but I think they though I provoked him rather than he provoked me.
I remember him saying never to come back, but in hindsight I don't think he really said that (I think he just slammed the door after tossing me out) and I probably embellished it in when telling the story later (it makes for a rather nice story). Incidentally, I did come back a few years later for my uncle's funeral and I learned that priest had died of a heart attack a few years before. I didn't see the new priest attacking anyone's beliefs, so he may be more tolerant, or maybe a funeral is less appropriate for recruiting than a wedding.
For the grandparent, Wozniak is and always will be the inventor of the Apple computer innards, but Jobs gave it the appliance like form factor. Jobs wanted it completely sealed, but Wozniak insisted that the cover be removable for expansions and tinkering, something he got until Jobs got his way with the Mac (or maybe Lisa - I never got to play with one of those). Gates and Microsoft mainly created the most widely used BASIC at the time, even selling it to Apple (aka Applesoft BASIC, whereas Woz's BASIC was thenceforward Integer BASIC). As far as Microsoft's OS, they didn't really get into that market until IBM paid them to get into it, and even then, they bought the software rather than make it in-house. Jobs also gave us NeXT, which is the platform where I used my first graphical web browser. And Jobs bought the Graphics Group from Lucasfilm and founded Pixar with them, and thus creating all three of my nephew's favorite film at least for part of their young lives, Cars.
As far as the parent's suggestions, Commodore came out with their PET computer about a year after the Apple ][, with an equally frustrating and horrifically failure prone tape drive compared to Apple's (1978's Disk ][ had me sold on Apple's for years 0 Commodore's 1541 that came out a couple of years later was much cheaper, but also 10x slower). Atari didn't release a computer until 1979 with the Atari 400, which had this awful membrane keyboard (but I wanted one so bad at the time, being a huge 2600 fan). Commodore's mainstream computer, the Vic 20, didn't come out until 1980, but it was a step into the stone ages if you'd used an Apple ][ (back to tape drive and 20+ minute loads that sometimes took 3-4 tries to load until they came out with the 1541 late that year, tiny memory, underpowered graphics, etc), but it was also about 10x cheaper. The C64 is where Commodore really started to shine, but the 1541's load times still drove me nuts. IBM was last to bat, and actually failed in the market, lasting only 5 years (sure their design won in the market, but they started making computers in late 1981 and stopped manufacturing PCs by about 1986-7), but their floppy drive speeds were decent. Many years later I had a similar experience with 1x optical drives loading a CD ROM (20 minutes for Fallout to load a level was typical - when I upgraded to a 4x, it loaded in about 2 minutes, so it wasn't all a drive speed issue).