Article by Eric Worrell
The real reason why Silicon Valley quietly quit the Democratic Party is that it quietly supports President Trump’s America First energy policy.
Let's get straight to it. A neural network wants to learn how to drive.
An explanation. Modern artificial neural networks usually start by generating a set of random neural networks.
As you can imagine, a randomly generated neural network is unlikely to be a good solution to the problem. But occasionally, some randomly generated networks are slightly less bad than their brethren.
These slightly less bad neural networks are used as guides for creating the next generation of neural network populations. The actual method of extracting the best from the previous generation varies, but it may involve backpropagation (mathematical optimization of expected results), genetic algorithms (“breeding” the best network like cattle, hoping that the subnetwork will inherit preferably from a parent), and mutation – copy the parent neural network but randomly change some parameters and see if these random changes produce a better neural network.
Eventually, if you stick with it long enough, some real skills will start to emerge.
Evaluating the performance of neural networks is a research hotspot in itself. The following article discusses some of the implications of different approaches to assessing performance.
Why do neural networks require so much energy?
The reason is that the software that artificial neural networks run is garbage. Artificial intelligence researchers are far from unlocking all the shortcuts and tricks the human brain uses to discover solutions.
But what it lacks in quality it makes up for in quantity—using acres of computers consuming hundreds of megawatts of power to accomplish tasks our human brains can accomplish with less power than is needed to run a television.
If you press “High Performance PC” in the above demonstration, you can see a lot of computing power help. The above neural network does not test one neural network at a time, but simultaneously evaluates the driving capabilities of all 10 neural networks in a single generation in “High Performance Computer” mode.
In a similar way, the simultaneous evaluation of millions of neural networks containing millions of neurons on huge data center computers can compensate for the low quality of the original software driving a single neural network.
What’s the point of spending all this cash and putting all these resources into neural networks?
Have you ever tried playing a chess app on your computer? Unless you are a very good chess player, that simple chess app will surprise you every time because the chess AI always knows the right move.
I believe that what chess apps do to you when you play chess is what artificial intelligence research does to the entire planet. Advanced neural networks will help tech giants always take the right action. They always know exactly what to put on the teleprompter to convince the audience.
Tech giants are also likely to invest heavily in medical life extension, and Google raised eyebrows in the industry when it hired Ray Kurzweil for a senior role in 2013. In addition to being a technical genius, Kurzweil is also one of the world's leading proponents of using artificial intelligence technology to research life extension and medical immortality.
A person will spend everything he has in his life.
Executives at Big Tech want to win and keep winning.
But to win, tech giants must abandon green energy. To compete with Asia, which is also developing AI technology, they need rock-solid, reliable energy supplies at prices comparable to those of Asia's tech giants. That's why tech giants now want to have their own in-house nuclear reactors.
I like to think I'm making a difference writing for WUWT, but when the dust settles, nothing I've ever written or said will kill the green energy movement.
The final blow to the green energy movement will be delivered by those who were once its strongest supporters.
But this silvery cloud also has a dark side. If the tech giants succeed, if they have the ability to always know the right actions to advance their goals, personal, political, social, and of course financial, maybe in the future we will be nostalgic for the good old days, when we All opponents are human beings like us.
Relevant
