Environmental and legal articles
updated in real time.

Jan 13 2020

AlphaGo . . . AlphaEarth: Is the environment just a complicated game of “Go”?

(Houston) – Editorial by Jed Anderson, Creator, Ceres AI  
—“Like AlphaGo, we can begin to build a machine that plays ‘environmental protection’ like AlphaGo plays Go.” – Jed Anderson, Creator of Ceres AI
AlphaEarth? Yes.
 
The earth has rules just like Go. Many of these rules (laws of physics) we do not fully understand yet. But they are eventually understandable–even if in their randomness. Can AI then be trained to play “environmental protection” like Google’s Deepmind AlphaGo plays Go? Of course. Whether its narrow AI or general AI, the answer is of course. It should be noted that if the game is “earth” its really not that narrow anymore. And if you have unlimited training data and unlimited computing power at some point, even narrow AI should be enough. Regardless, once AI is trained to play “environmental protection”, our expression will be very similar to Lee Sedol’s when AlphaGo played move number 37. It’s a move a human would never even see. If you have never seen this video footage, Lee Sedol’s expression will almost leave you breathless. A pivotal moment in human history. China after the defeat to its unbeatable national game decided it would try to become the world leader in AI.
AI doesn’t need thousands of years to learn how to play “environmental protection”.
 
Here’s what’s fascinating about AI. It doesn’t need to understand rules. We train it on outcomes, and it makes the rules to achieve the outcomes. We need to be extremely cautious here of unintended consequences, and of course understanding the “black box” decision making, but that isn’t a reason not to proceed. Why? First, because it’s inevitable. So directing it is the choice–not passively succumbing to fear and fatalism–deterministic procrastination that leads one past the point of being able to affect an outcome. And second, we need to protect our planet with more speed and less cost.  

Human beings are incredible . . . but none of us are that smart   The fastest way we can solve the climate change and air pollution problems is to build a machine that solves the problems for us. Relying on human minds as the primary means to find the solutions, and then all communicating and agreeing with each other on the solutions, will take forever. Forever. The fact is that none of us are that smart. We simply lack the computing power. Smart people especially have a hard time acknowledging that other people are smarter than they are, or that a computer can be smarter than they are. I’m not sure why. It’s like their ego or persona is completely tied up in their cognitive abilities. We are much more than our intellect. Much, much more. Please, please see that you are more than mere intellect. And let’s not argue here about what is intelligence, but let’s all agree that none of us have the power to calculate, store, or transmit information faster than a computer.
The limitations to human minds when it comes to computing power   Because we each individually lack the necessary computing power to solve the super-big data problems of this world, we then need to link up with other human minds to increase our collective computing power. The problem of course is that these other human minds, like our own mind, can only transmit information at about 39 bits per second to other minds–and have firewalls and run their own hardware and software programs. It makes it really difficult to perform combined computing tasks. People will say, it should take forever because this human collective conscious works best that way. It takes time to develop a consensus on what to do. Yes, but allowing a machine to help us get to a point where we can make a better and faster informed human judgment decision is a good thing. It’s what we’ve been doing with increasing effectiveness for the past 50 years.   —“If we have data, let’s look at data. If all we have are opinions, let’s go with mine.” – Jim Barksdale, former Netscape CEO

It’s faster to build a machine to understand air pollution and climate change than it is to try to understand it ourselves   It’s much faster than a concomitant morass of individual research projects, laws, treaties, policy efforts, and pleas for behavioral change.   skip most of the politicsskip most of the laws and regulationsskip much of the underlying/competing research (figuring out all the individual rules to the game)   Just build a machine that does it for us. That sees things we as individuals can’t see. Or even if one or two people can see, would need to convince thousands of other individuals to see.   —“As always, there’s good news and there’s bad news. The bad news is, we seem incapable of solving our more pressing or persistent problems. The good news is, we’re getting closer to building a machine that might do it for us.” – Jim Vibert
Combination of AI and quantum technology and we are there.   People have told me “Jed, quantum computing isn’t there yet”. Of course not. The Macintosh wasn’t there yet either, but you could still do things with it. Here’s a fascinating Cal Tech article on the use of quantum computers in the current NISQ era (noisy intermediate scale quantum era). And a recent article on quantum computing and climate change. And a really interesting article on research using AI to learn quantum mechanics to predict molecular wave functions.
Lee Sedol’s expression again really says it all.   Lee Sedol thought AlphaGo made a mistake in move 37. It was a move that a human would never make. Never would even see.   There are similar epiphanies that I think “AlphaEarth” will find . . . and our expression will be like Lee Sedols. I think we will be astonished by how profoundly simple some of the environmental solutions will be.
Click here for information on Ceres AI’s current environmental applications.

Call The AL Law Group, PLLC.

Talk to Texas Environmental Lawyers Now!

Call AL Law Group (281) 852-8064

Texas Administrative Code: Title 30: Part 1

Get Texas Environmental News by Email

Texas Environmental News