Matthew J Price
9,312 followers -
Father, Transhumanist, Futurist, Philosopher, Economist, Master Floss Roller
Father, Transhumanist, Futurist, Philosopher, Economist, Master Floss Roller
9,312 followers
Matthew J's interests
Matthew J's posts
Post has shared content
I once paid a guy $1 to make a comic for me. He was thrilled, it was the first time someone paid him to make a comic. Here is that comic.
Post has shared content
A propellentless electric thruster would be a game changer. I have no opinion on whether this is going to work or not, but I wouldn't be mad if it did.
Post has attachment
//“A plan of business tax cuts that has no offsets, to use some very esoteric language, is not a thing,” Callas said. “It’s not a real thing. And people can come up with whatever plans they want. Not only can that not pass Congress, it cannot even begin to move through Congress day one. And there are political reasons for that. No. 1, members wouldn’t vote for it. But there are also procedural, statutory procedural, legal reasons why that can’t happen.”
Post has attachment
//What does this mean for the Trucking industry?
“We believe an autonomous, electric truck can be a game changer for trucking carriers and by significantly lowering operating costs, improving productivity and even driving industry consolidation,” the analyst wrote.
They estimated that an autonomous, electric truck could be 60 to 70 percent less expensive to operate compared with a human-driven diesel big rig. The cost savings come from drivers, fuel, maintenance and insurance.
The analysts said they have talked to motor carriers that are open to using such trucks made by new, non-incumbent manufacturers “as long as the performance, service and operating costs are superior.”
“In fact, we would not be surprised if Tesla revealed large carrier and shipper partners during its truck reveal in September,” they wrote.//
“We believe an autonomous, electric truck can be a game changer for trucking carriers and by significantly lowering operating costs, improving productivity and even driving industry consolidation,” the analyst wrote.
They estimated that an autonomous, electric truck could be 60 to 70 percent less expensive to operate compared with a human-driven diesel big rig. The cost savings come from drivers, fuel, maintenance and insurance.
The analysts said they have talked to motor carriers that are open to using such trucks made by new, non-incumbent manufacturers “as long as the performance, service and operating costs are superior.”
“In fact, we would not be surprised if Tesla revealed large carrier and shipper partners during its truck reveal in September,” they wrote.//
Post has shared content
Like the open Internet? It's time to stand up for it.
Post has shared content
I imagine much more complicated games in the future where you direct complex operations or military efforts by talking in natural language to your agents or lieutenants who execute your wishes and simultaneously learn and improve themselves. All in AR of course.
AI learns to play video game from instructions in plain English
An AI has learned to tackle one of the toughest Atari video games by taking instructions in plain English. The system, developed by a team at Stanford University in California, learned to play the game Montezuma’s Revenge, in which players scour an Aztec temple for treasure. The game is challenging for AI to learn because it offers sparse rewards, requiring players to make several moves before earning any points. Most video-game-playing AIs use reinforcement learning to develop a strategy, relying on feedback like game points to tell them when they are playing well. To help their AI pick up game tactics quicker, the Stanford team gave their reinforcement learning system a helping hand in the form of natural language instructions, for example advising it to “climb up the ladder” or “get the key”. “Imagine teaching a kid to play tennis by handing them a racket and leaving them in front of a ball machine for 10 years. That’s basically how we teach AI right now,” says team member Russell Kaplan. “It turns out kids learn a lot faster with a coach.” Teaching an AI in this way could have far-reaching applications, because using natural language means anyone could advise the AI, not just computer programmers.
An AI has learned to tackle one of the toughest Atari video games by taking instructions in plain English. The system, developed by a team at Stanford University in California, learned to play the game Montezuma’s Revenge, in which players scour an Aztec temple for treasure. The game is challenging for AI to learn because it offers sparse rewards, requiring players to make several moves before earning any points. Most video-game-playing AIs use reinforcement learning to develop a strategy, relying on feedback like game points to tell them when they are playing well. To help their AI pick up game tactics quicker, the Stanford team gave their reinforcement learning system a helping hand in the form of natural language instructions, for example advising it to “climb up the ladder” or “get the key”. “Imagine teaching a kid to play tennis by handing them a racket and leaving them in front of a ball machine for 10 years. That’s basically how we teach AI right now,” says team member Russell Kaplan. “It turns out kids learn a lot faster with a coach.” Teaching an AI in this way could have far-reaching applications, because using natural language means anyone could advise the AI, not just computer programmers.
Post has shared content
Just assume you're being watched, and that anything you say on camera can be analyzed afterward for truthfulness.
Another one for the "privacy is dead" file. This type of software really doesn't bother me. While I understand the fears expressed over it, I feel like the software has a better chance of being unbiased than humans attempting to do the same job. This is how I really see most crime going away when the odds of being caught go up to 100%. Of course, the challenge then becomes making sure that you do a good job of defining what is criminal. I would argue that should be the goal from the beginning, and that fighting systems like this instead of pushing to make sure that we don't make it illegal to be different in ways that don't actually harm others is a waste of resources.
Post has shared content
Greatly encouraging to see the political progress we've made here.
"California, the largest U.S. car market, plans to allow testing on public roads of self-driving vehicles without human backup drivers by the end of the year, state officials said Friday."
Wow.
Wow.
Post has shared content
"Rethinking recurrent neural networks." "We decided to take our approach one step further. We realized that the output of the attention mechanism could be feed back into the attention mechanism at the next processing step. The resulting model represents a new kind of RNN model. Given that the weighted average is recursively defined, we decided to call this approach a recurrent weighted average (RWA) model."
"We started testing the RWA model on several toy problems and compared its performance to the LSTM model. On every task except one, the RWA model learned much faster using fewer training steps. Moreover, each training step for the RWA model required less clock-time."
"We started testing the RWA model on several toy problems and compared its performance to the LSTM model. On every task except one, the RWA model learned much faster using fewer training steps. Moreover, each training step for the RWA model required less clock-time."
Post has shared content
Google’s Finally Offering Rides in Its Self-Driving Minivans http://bit.ly/2q2eTnC
Wait while more posts are being loaded