you probably don't know this. Modern Airbus jets are pretty much autonomous. The pilot does very little. The reason that it can is that autonomous flight has been solved for over 2 decades and has been known in the AI field for quite a while. Don't take my word for it, read the literature. We've had autonomous cruise missiles for over 2 decades. Modern military drones have been in service for over a decade and the most advanced drones are fully autonomous. Again, don't take my word and do the research.
Given flying cars don't exist yet, it can be done and engineered such that it has to be 100% autonomous. Meaning a human has zero chance of fucking things up. There's no throttle, steering wheel or mirrors. The auto pilot does it all. We already have most of the technology needed like creating virtual highway that all flying cars must obey. The part we don't have is great batteries that are powerful, light and charge quickly. Once someone figures out the battery problem, I for one would swap out my car for a flying car. Humans suck at driving and do too much stupid shit. Like how many times you see people shaving, doing makeup, texting and talking while driving?
literally look at all the neat shit people do on kinect hacks or makerspace with 3D scanners. The TrueDepth camera just needs software and a turn table to create a 3D scan of a smallish object.
FaceID is just the first obvious use case of the TrueDepth camera. Having worked with computer vision and played with various techniques, the innovative part is shrinking XBox Kinect down to a mobile phone. Essentially the sensor array at the top means iphone X can do depth sensing and binocular vision. But it's not like having 2 color camera's which is known to have major limitations. Having IR dot projector + IR camera + regular camera means it's easier do image recognition because you don't need to calculate the depth field in software. Having the depth field calculated by hardware means you can focus on the algorithm part of image recognition. You don't have to muddle with offset calculation, depth calculation and a bunch of other tedious stuff. I just hope Apple opens it up.
You know Qualcomm, Samsung and Google are all working on similar efforts and I would expect them to bring out similar hardware acceleration in 2-3yrs
How many engineers does it take to loose their mind over stupid screens?
Answer: more than necessary
Seriously how many people can tell the difference between a 400, 425, 450, 500 ppi screen 16-24 inches from your eyes? All of this bitching and moaning about screens is stupidity. We've gone far past diminishing returns for screens. It uses more battery, increases the charges needed per week and overall is a waste of energy. Go ahead and get all bent. The "future" part of X is the neural engine and dedicated ASIC in A11. Yes I own a iPhone and I also own android tablet. I don't give a crap because at the end of the day I build software for customers. What I do care about is the built-in hardware capabilities and the ease of development process.
Frankly I could give a shit about how pretty the iPhone looks. I'd gladly swap out the expensive industrial design to save money. For a simple reason, people drop phones. My kids drop their phones. A more expensive screen just means more $$ to fix it. Buying used phones makes it affordable. Anyone that buys a new Smart phone be it iPhone or Galaxy is a fucking moron with more money than brains.
I see lots of people bitching about edge-to-edge or that OLED is old. The big thing is the hardware built for neural net. They dumped the old GPU to build one that can be used for DNN, image recognition, ARKit and MLKit. I hope they open up the neural engine in future releases of iOS. Think about all the uses. If you train a neural net to recognize foods, you could simply take a picture of your meal and record your calorie intake. You could take a picture of something and quickly search for it. The potential is huge.
At least this time the studios are admitting the movies that bombed sucked. How about make new and original movies that people want to see. There's plenty of people that want to create fresh new content. Instead of block busters, make smaller better movies and mix things up a bit.
The last time I checked, the idea of a near-vacuum transport system was old. It's nice that Elon is recycling the idea, but please enough brown nosing. The concept was used in several science fiction books, movies, cartoons and tv shows. Elon should get credit for reviving an old idea and helping "some" people think outside the gas guzzling cars. Discovery channel even had a show about a trans-atlantic train system that used near-vacuum tubes years before Elon revisited the idea.
If too many people die or get in serious accidents because the technology was deployed before it is ready, there will be a backlash and slow progress. Google has been at autonomous cars longer than any other public company. They didn't rush it out and took a methodical approach. Elon created a problem for himself by having annual hype cycle. He setup the expectation that every year he had to give a big announcement. Elon is his own worst enemy. If he had the discipline to not fall for the hype cycle, and taken a more conservative approach, a person wouldn't have died in a Tesla. If they required the driver to be in the driver's seat and use face recognition to make sure they are paying attention to traffic, the accident could have been avoided. Google's steady approach is the right way to go.
you clearly haven't seen how many open source projects work. People fork projects for even dumber reasons, so why should node.js work any different? Plus, the whole point of open source and github style development is that everyone can fork. Were you asleep the last 3 years?
Did someone hold a gun to his head and force him to write that idiotic opinion piece? Why anyone feels a need to defend his mistake is beyond me. If Damore shot his foot with a shotgun in google office, would you defend him? What he did was inappropriate and he was too egotistic and stupid to realize it. How he did it is the problem. If he was more constructive and eloquent, he "could" have made the discussion fruitful. Instead, he chose the wrong words and then decided to publish it internally. Let it go already, he made his bed and now he has to suck it up and deal with the consequences.
As any consultant will testify, most functional specs are either non-existent or complete garbage. How many times have you discovered the business use case was wrong, the user story was wrong and the functional specification wrong. The problem with software development isn't technical, it's a human problem. I seriously doubt within the next 30 years, AI will make progress on the problem of "humans don't know what the hell they want and change their minds".
We have all these different development methodologies to try and fail ideas fast and avoid wasting time. Often what a business person says isn't what they want. If someone invents an AI that can read minds, then AI writing software will be practical. Until then, garbage in != usable software.
Anyone that has seen corporate networks knows just how crappy security is. It's been that way for over 2 decades. If you haven't noticed how bad IT security is, you either didn't care, didn't pay attention or don't really understand security.
I actually worked in Boston financial sector building compliance systems. I've talked to banks like Wachovia, Bears Stearn and others. It's not a secret how fraud occurs. The crooks at these institutions "break" the compliance system to let things slip through. How do you do that? One way is to edit a rule and blame it on "human error". Another is to get the compliance officer to override a violation. Yet another is to setup complex transactions. The compliance engine can only detect fraud if it has enough data. Some system purposely require a minimum number of fields entered so that transactions can go through. Data is also falsified to by pass compliance. Watson can't catch this stuff, because all they will is break watson. It's pretty easy. Until you put bankers in maximum security jails, they will keep breaking laws.
because human lives matter and there's thousands of vehicle accidents every year. It costs in so many ways and it also wastes fuel. Driving sanely and efficiently can save a lot of fuel and reduce pollution.
if you prefer to live in pollution and consume oil at a faster rate, then don't change. Pretty simple. That or figure out how to make a finite resource infinite. If you can do that, no need to change anything at all, keep burning gas like it's an infinite resource.
Here's some news for you, drivers changing lanes without looking happens thousands of times a day. A friend of mine had that happen with a 18 wheeler. His car was wrecked and they tested the driver for alcohol. He wasn't drunk, maybe he was tired. Either way, he changed lanes without looking and my friend is lucky he survived.
Any autonomous vehicle needs to handle this, otherwise people will die as they do now when careless drivers make mistakes. It's going to take a decade or more before all vehicles on the freeway are autonomous, which it should be. AI is useful, but we need to tackle the whole problem holistically. Cars have to change, policy has to change, culture has to change and people have to change.
there's additional panels on the garage and other structures, so the calculation is missing a couple thousand sq meters. I haven't done the math, so I don't actually know what the projected output is.
Anyone find that to be super creepy, wrong and just gross. BigData makes this possible, but don't be a dick or creepy ass fuckwad. Some people just have no sense of personal space or being a decent human being.
No where do they list the actual questions they used, so the only thing you can "take away" from it is surveys can't be trusted without full disclosure of the questions used.
The evidence does suggest more than half of US citizens don't know dick about security or encryption, which also suggests the education system is totally fucked. It's not that hard to understand "the basics" of security and encryption. Being an expert, which I am not, takes decades. You don't need to be an expert to understand what the FBI is asking is a backdoor. With respect, I say fuck any government that demands a backdoor on smart phones.
Anyone that has driven in New York or Boston will tell you just how screwed up driving is. Driving on the freeway is much easier than driving city streets by atleast 2 orders of magnitude. I think it's great they're working on it and making great progress, but all of this hype is a bit silly.
Having cars handle freeway driving would definitely improve things and help reduce congestion. If we're talking freeway only autonomous driving, that feels feasible by 2030. The other 10% is going to take much longer than anyone thinks.
Slashdot Mirror
Given flying cars don't exist yet, it can be done and engineered such that it has to be 100% autonomous. Meaning a human has zero chance of fucking things up. There's no throttle, steering wheel or mirrors. The auto pilot does it all. We already have most of the technology needed like creating virtual highway that all flying cars must obey. The part we don't have is great batteries that are powerful, light and charge quickly. Once someone figures out the battery problem, I for one would swap out my car for a flying car. Humans suck at driving and do too much stupid shit. Like how many times you see people shaving, doing makeup, texting and talking while driving?
literally look at all the neat shit people do on kinect hacks or makerspace with 3D scanners. The TrueDepth camera just needs software and a turn table to create a 3D scan of a smallish object.
You know Qualcomm, Samsung and Google are all working on similar efforts and I would expect them to bring out similar hardware acceleration in 2-3yrs
Answer: more than necessary
Seriously how many people can tell the difference between a 400, 425, 450, 500 ppi screen 16-24 inches from your eyes? All of this bitching and moaning about screens is stupidity. We've gone far past diminishing returns for screens. It uses more battery, increases the charges needed per week and overall is a waste of energy. Go ahead and get all bent. The "future" part of X is the neural engine and dedicated ASIC in A11. Yes I own a iPhone and I also own android tablet. I don't give a crap because at the end of the day I build software for customers. What I do care about is the built-in hardware capabilities and the ease of development process.
Frankly I could give a shit about how pretty the iPhone looks. I'd gladly swap out the expensive industrial design to save money. For a simple reason, people drop phones. My kids drop their phones. A more expensive screen just means more $$ to fix it. Buying used phones makes it affordable. Anyone that buys a new Smart phone be it iPhone or Galaxy is a fucking moron with more money than brains.
I see lots of people bitching about edge-to-edge or that OLED is old. The big thing is the hardware built for neural net. They dumped the old GPU to build one that can be used for DNN, image recognition, ARKit and MLKit. I hope they open up the neural engine in future releases of iOS. Think about all the uses. If you train a neural net to recognize foods, you could simply take a picture of your meal and record your calorie intake. You could take a picture of something and quickly search for it. The potential is huge.
the article indicates they signed the contract and now want to muscle their way out of certain parts of the agreement. how predictable of comcast.
At least this time the studios are admitting the movies that bombed sucked. How about make new and original movies that people want to see. There's plenty of people that want to create fresh new content. Instead of block busters, make smaller better movies and mix things up a bit.
The last time I checked, the idea of a near-vacuum transport system was old. It's nice that Elon is recycling the idea, but please enough brown nosing. The concept was used in several science fiction books, movies, cartoons and tv shows. Elon should get credit for reviving an old idea and helping "some" people think outside the gas guzzling cars. Discovery channel even had a show about a trans-atlantic train system that used near-vacuum tubes years before Elon revisited the idea.
If too many people die or get in serious accidents because the technology was deployed before it is ready, there will be a backlash and slow progress. Google has been at autonomous cars longer than any other public company. They didn't rush it out and took a methodical approach. Elon created a problem for himself by having annual hype cycle. He setup the expectation that every year he had to give a big announcement. Elon is his own worst enemy. If he had the discipline to not fall for the hype cycle, and taken a more conservative approach, a person wouldn't have died in a Tesla. If they required the driver to be in the driver's seat and use face recognition to make sure they are paying attention to traffic, the accident could have been avoided. Google's steady approach is the right way to go.
you clearly haven't seen how many open source projects work. People fork projects for even dumber reasons, so why should node.js work any different? Plus, the whole point of open source and github style development is that everyone can fork. Were you asleep the last 3 years?
I really hope Oracle hands off Java to Eclipse or Apache. That would make me happy.
damn those vegans! Joking aside, they should setup a second table with meat for carnivores, but I'm a meat eater :)
Not all rich people are evil assholes, but this jerk off definitely is an asshole to the power 1000.
Did someone hold a gun to his head and force him to write that idiotic opinion piece? Why anyone feels a need to defend his mistake is beyond me. If Damore shot his foot with a shotgun in google office, would you defend him? What he did was inappropriate and he was too egotistic and stupid to realize it. How he did it is the problem. If he was more constructive and eloquent, he "could" have made the discussion fruitful. Instead, he chose the wrong words and then decided to publish it internally. Let it go already, he made his bed and now he has to suck it up and deal with the consequences.
We have all these different development methodologies to try and fail ideas fast and avoid wasting time. Often what a business person says isn't what they want. If someone invents an AI that can read minds, then AI writing software will be practical. Until then, garbage in != usable software.
Anyone that has seen corporate networks knows just how crappy security is. It's been that way for over 2 decades. If you haven't noticed how bad IT security is, you either didn't care, didn't pay attention or don't really understand security.
I actually worked in Boston financial sector building compliance systems. I've talked to banks like Wachovia, Bears Stearn and others. It's not a secret how fraud occurs. The crooks at these institutions "break" the compliance system to let things slip through. How do you do that? One way is to edit a rule and blame it on "human error". Another is to get the compliance officer to override a violation. Yet another is to setup complex transactions. The compliance engine can only detect fraud if it has enough data. Some system purposely require a minimum number of fields entered so that transactions can go through. Data is also falsified to by pass compliance. Watson can't catch this stuff, because all they will is break watson. It's pretty easy. Until you put bankers in maximum security jails, they will keep breaking laws.
if you prefer to live in pollution and consume oil at a faster rate, then don't change. Pretty simple. That or figure out how to make a finite resource infinite. If you can do that, no need to change anything at all, keep burning gas like it's an infinite resource.
Here's some news for you, drivers changing lanes without looking happens thousands of times a day. A friend of mine had that happen with a 18 wheeler. His car was wrecked and they tested the driver for alcohol. He wasn't drunk, maybe he was tired. Either way, he changed lanes without looking and my friend is lucky he survived. Any autonomous vehicle needs to handle this, otherwise people will die as they do now when careless drivers make mistakes. It's going to take a decade or more before all vehicles on the freeway are autonomous, which it should be. AI is useful, but we need to tackle the whole problem holistically. Cars have to change, policy has to change, culture has to change and people have to change.
there's additional panels on the garage and other structures, so the calculation is missing a couple thousand sq meters. I haven't done the math, so I don't actually know what the projected output is.
Anyone find that to be super creepy, wrong and just gross. BigData makes this possible, but don't be a dick or creepy ass fuckwad. Some people just have no sense of personal space or being a decent human being.
No where do they list the actual questions they used, so the only thing you can "take away" from it is surveys can't be trusted without full disclosure of the questions used.
The evidence does suggest more than half of US citizens don't know dick about security or encryption, which also suggests the education system is totally fucked. It's not that hard to understand "the basics" of security and encryption. Being an expert, which I am not, takes decades. You don't need to be an expert to understand what the FBI is asking is a backdoor. With respect, I say fuck any government that demands a backdoor on smart phones.
it seems like the stupid ads on slashdot from flash ads cause firefox to crash atleast 2 times a day. So yeah, screw flash.
Having cars handle freeway driving would definitely improve things and help reduce congestion. If we're talking freeway only autonomous driving, that feels feasible by 2030. The other 10% is going to take much longer than anyone thinks.
Didn't MIT already show off their Femtosecond camera in a TED talk a few years back? That's 1 trillion per second, so this new one is slower!