I recently bought a 2013 MBP instead of a new one. I refuse to pay 2x more for a laptop that is 2x worse. Better to prop the second hand market than these greedy fucks that work today at Apple.
Recently, I was prompted to install an OSX update. It bricked my laptop on reboot. Fortunately the installer was on a separate partition and I could boot back into the previous OS, but I have NEVER seen such shitty software releases from Apple in 14 years of usage. Ever since Steve died the company is ruder-less.
Is that the extent of your argument? You need to realise that not all actors are well intentioned and Google is unleashing a technology that will be replicated in one or two years everywhere. Then your voice and your private personal information could be used to impersonate you. You will be put in a very uncomfortable situation if people can't trust that it is actually you speaking and not a Russian bot.
What I want from dialogue agents is to assist me in research and learning, not to turn up the heat or quote from a list of facts. I think they will eventually become good intellectual partners for us.
If Palantir has wrong information on someone won't that make this person a victim of the state? How are decisions made on data collected by various shady deals, when we don't know if the data is correct?
When you get at FaceBook size, knowing so much everything about people becomes more than a symptom. At this scale it is a problem in itself. The potential for abuse is on a whole new level.
I'm worried that apps, which are like a kind of websites, can be simply removed in large quantity. Crap as they might be, access to them is important for cultural reasons.
But they can broadcast to countries outside the US without paying a license fee, right? So they should just skip US for a few launches until wisdom comes to NOAA's head.
The 700 dimensions vector (if it's true) is not something you can make sense of. It's an embedding vector that represents your characteristics in relation to all the other people. Each individual dimension doesn't have a meaning.
Let's not call any and all predictions laws. They're not laws. They are functions fit on a short stretch of data that have no predictive power in the future. Not even experts can predict the future, a "law" has no chance here.
> You asked "to name women tech leaders" and got idiot answers, but did you ask "to name men tech leaders"? No, you didn't. You presumed you wouldn't get idiot answers.
Not just that, but they should calculate the percent of CEOs that are known by name, male and female. There are many more male CEOs that are unknown by the public.
For example in ImageNET, the most famous image recognition dataset, there are a thousand categories including hundreds of breeds of dogs and cats. Humans get 5% errors with at least a day of training. Neural nets get 4% error. When you give the net enough examples, it becomes better than human. We just need to add more training data. For more complex vision tasks such as image based question answering, neural nets still can't equal humans.
They are trying to solve the big problems of vision, speech, robotics, reasoning, unsupervised learning, learning how to act intelligently (reinforcement learning) and acquisition of common sense. They are creating virtual worlds for AI agents, such as OpenAi Gym, DeepMind Lab, MuJoCo, Microsoft AirSim, Arcade Learning Environment
etc. These environments are freely accessible and useful in training AI agents and robots.
They are creating AI frameworks like TensorFlow, PyTorch and CNTK. These libraries come for free and are really useful.
I wouldn't say they aren't even trying to solve big problems.
AI is not going to be like rocket science. You won't need billions to run an experiment or to create your own. Everyone will be able to use AI for anything they desire, it won't serve only the 1%. Even today, the AI applications we have serve the general public and many SOTA models are freely accessible on github. You only need a few GPUs to do cutting edge research, you can do AI development in a garage with less money than it cost to raise the house. The datasets we have in public domain are ever growing and more complete, the datasets that Google and FB hoard are just marketing shit useful for advertising - not the kind of data we need for creating general intelligence.
> When only a fraction of the population is needed for the few jobs that are left, the owners, producers, and shareholders are not going to want to have to pay increased taxes for those who they made obsolete.
On the other hand, when most people will be jobless, who will buy their products? How will their companies make absurd quantities of money? Buying and selling only between the 1% is a downgrade for them too.
I think the solution involves not UBI but "guaranteed jobs for everyone". These jobs should be community based and related to local needs, such as community ISPs, credit cooperatives, home solar installation, auto repair, farming, construction, education and primary health care (medical clinics), etc. A community of people can do most of these jobs internally without a need for UBI, generating jobs for its members.
On the other hand you and everyone else will be able to use AI in many ways and do things that are impossible today even for the big corporations. I think objections to Bill's argument don't have enough faith in human greed - we will create new jobs because we are always in search for more, we're too damn entitled for more to sit still. We always find roles and things to do for people, people are already like the AI we strive to create and don't cost too much to function.
Such models have no common sense yet - can't tell if "the use of the umbrella causes the rain or the other way around". They can't think like us, they just copy text and try to hit all the sub-topics with naturally sounding language based on the source material. It's more similar to Google translator than a human Wikipedia editor.
My personal experience tells me girls have different preferences when compared to boys - how they play, what games they play and what they learn. Sometimes you have to accept that girls don't have the same interests as their fathers and brothers. It's natural.
I really tried teaching my girl programming, and she was good as a beginner, immediately learned the basic principles and could write programs better than most of her class. But the thing is, the moment I stopped pushing programming, she forgot it. She's much more interested in makeup, clothes and her social circle. She doesn't have that starry look in her eyes when talking about what she can make computers do. So she has the skill and mental power but not the drive or interest to do it. I just accepted this reality out of respect for her. She has a better path for herself, and I might not be able to fully understand her values as she does.
Slashdot Mirror
I recently bought a 2013 MBP instead of a new one. I refuse to pay 2x more for a laptop that is 2x worse. Better to prop the second hand market than these greedy fucks that work today at Apple.
Recently, I was prompted to install an OSX update. It bricked my laptop on reboot. Fortunately the installer was on a separate partition and I could boot back into the previous OS, but I have NEVER seen such shitty software releases from Apple in 14 years of usage. Ever since Steve died the company is ruder-less.
> Retarded fucking pleb.
Is that the extent of your argument? You need to realise that not all actors are well intentioned and Google is unleashing a technology that will be replicated in one or two years everywhere. Then your voice and your private personal information could be used to impersonate you. You will be put in a very uncomfortable situation if people can't trust that it is actually you speaking and not a Russian bot.
Is IRC still a thing? https://xkcd.com/1782/
AI can also train in a simulator, it's not necessary to use datasets. A simulator is like a dynamic dataset. That's how AlphaGo beat humans.
What I want from dialogue agents is to assist me in research and learning, not to turn up the heat or quote from a list of facts. I think they will eventually become good intellectual partners for us.
If Palantir has wrong information on someone won't that make this person a victim of the state? How are decisions made on data collected by various shady deals, when we don't know if the data is correct?
> It's a loss all around
Only if you never need to drive outside your neighbourhood.
When you get at FaceBook size, knowing so much everything about people becomes more than a symptom. At this scale it is a problem in itself. The potential for abuse is on a whole new level.
Quantum computing isn't even able to find its way out of a closet. No, really, quantum computing? Is this bullshit? Fake news?
I'm worried that apps, which are like a kind of websites, can be simply removed in large quantity. Crap as they might be, access to them is important for cultural reasons.
But they can broadcast to countries outside the US without paying a license fee, right? So they should just skip US for a few launches until wisdom comes to NOAA's head.
The 700 dimensions vector (if it's true) is not something you can make sense of. It's an embedding vector that represents your characteristics in relation to all the other people. Each individual dimension doesn't have a meaning.
Let's not call any and all predictions laws. They're not laws. They are functions fit on a short stretch of data that have no predictive power in the future. Not even experts can predict the future, a "law" has no chance here.
They are just like Republicans, but reversed. It's a religion.
> You asked "to name women tech leaders" and got idiot answers, but did you ask "to name men tech leaders"? No, you didn't. You presumed you wouldn't get idiot answers.
Not just that, but they should calculate the percent of CEOs that are known by name, male and female. There are many more male CEOs that are unknown by the public.
For example in ImageNET, the most famous image recognition dataset, there are a thousand categories including hundreds of breeds of dogs and cats. Humans get 5% errors with at least a day of training. Neural nets get 4% error. When you give the net enough examples, it becomes better than human. We just need to add more training data. For more complex vision tasks such as image based question answering, neural nets still can't equal humans.
They are trying to solve the big problems of vision, speech, robotics, reasoning, unsupervised learning, learning how to act intelligently (reinforcement learning) and acquisition of common sense. They are creating virtual worlds for AI agents, such as OpenAi Gym, DeepMind Lab, MuJoCo, Microsoft AirSim, Arcade Learning Environment etc. These environments are freely accessible and useful in training AI agents and robots. They are creating AI frameworks like TensorFlow, PyTorch and CNTK. These libraries come for free and are really useful.
I wouldn't say they aren't even trying to solve big problems.
AI is not going to be like rocket science. You won't need billions to run an experiment or to create your own. Everyone will be able to use AI for anything they desire, it won't serve only the 1%. Even today, the AI applications we have serve the general public and many SOTA models are freely accessible on github. You only need a few GPUs to do cutting edge research, you can do AI development in a garage with less money than it cost to raise the house. The datasets we have in public domain are ever growing and more complete, the datasets that Google and FB hoard are just marketing shit useful for advertising - not the kind of data we need for creating general intelligence.
> When only a fraction of the population is needed for the few jobs that are left, the owners, producers, and shareholders are not going to want to have to pay increased taxes for those who they made obsolete.
On the other hand, when most people will be jobless, who will buy their products? How will their companies make absurd quantities of money? Buying and selling only between the 1% is a downgrade for them too.
I think the solution involves not UBI but "guaranteed jobs for everyone". These jobs should be community based and related to local needs, such as community ISPs, credit cooperatives, home solar installation, auto repair, farming, construction, education and primary health care (medical clinics), etc. A community of people can do most of these jobs internally without a need for UBI, generating jobs for its members.
On the other hand you and everyone else will be able to use AI in many ways and do things that are impossible today even for the big corporations. I think objections to Bill's argument don't have enough faith in human greed - we will create new jobs because we are always in search for more, we're too damn entitled for more to sit still. We always find roles and things to do for people, people are already like the AI we strive to create and don't cost too much to function.
Such models have no common sense yet - can't tell if "the use of the umbrella causes the rain or the other way around". They can't think like us, they just copy text and try to hit all the sub-topics with naturally sounding language based on the source material. It's more similar to Google translator than a human Wikipedia editor.
My personal experience tells me girls have different preferences when compared to boys - how they play, what games they play and what they learn. Sometimes you have to accept that girls don't have the same interests as their fathers and brothers. It's natural.
I really tried teaching my girl programming, and she was good as a beginner, immediately learned the basic principles and could write programs better than most of her class. But the thing is, the moment I stopped pushing programming, she forgot it. She's much more interested in makeup, clothes and her social circle. She doesn't have that starry look in her eyes when talking about what she can make computers do. So she has the skill and mental power but not the drive or interest to do it. I just accepted this reality out of respect for her. She has a better path for herself, and I might not be able to fully understand her values as she does.
> other than obvious differences in reproductive role These differences can influence career choice. It's not all about the equipment in the pants.
What speeds can we expect from this sat network?