Showing posts with label predictions for a new year. Show all posts
Showing posts with label predictions for a new year. Show all posts

Sunday, March 10, 2019

Looking ahead... Science news, predictions and more... including lasers!


NBC News has web-featured "19 bold predictions for science and technology in 2019": Here's what thought leaders in innovation expect to see in the coming year." Nearly all the sages and experts made interesting but short-term forecasts about their own fields... with one or two exceptions. I took a risk... but then, I can afford to.

Possibly the most important thing I wrote or said in 2018 is this interview for Thomson-Reuters on the topic of Artificial Intelligence. Expertly edited by Paul Thies, it manages to convey – in a very brief space -- some important concepts currently ignored by AI researchers, societies and citizens.

In the science-related political news: As we celebrate amazing discoveries… and NASA recovers from the shut-down… there’s this. The U.S. Senate confirmed extreme-weather specialist Kelvin Droegemeier as President Donald Trump's top science and technology adviser, after a two year vacancy, the longest ever. His office (OSTP) has plummeted to just 35 staffers. What’s unclear is what it all means. He’s a meteorologist and expert on weather disruptions and not a denialist cultist. So what gives? Did someone apply leverage? 

Maybe a younger/smarter Koch who doesn't want to be rich in a world that is driven into radical revolution?

Here's an interesting survey-essay by Nobelist Charles Townes.  His 1997 perspectives on the big issues -- where the universe (and we) came from and where life (and we) may be going -- make for fascinating reading, especially from twenty years later. Among the things that have changed since then -- from pocket supercomputers to AIDS semi-cures to genomics and the stunning plummet in electorate-IQ -- few discoveries change the essence. But one of the biggest is our picture of the origin of life on Earth. Back then, Townes thought that up to a billion years passed between the Earth's formation and the first known evidence of living systems. We now know that life appears in the fossil record (or at least probable signs of it) within only a couple of hundred million years after the planet was molten and probably earlier. This puts a much higher supposition that life will emerge almost automatically, as soon as conditions are right. This, in turn, has huge implications for our mental image of Life in The Universe and questions like the Fermi Paradox. 

But overall a fascinating read. We need big perspectives, from time to time. (Some of us try to provide ;-)

== Intelligence Wonders ==

This Clever AI Hid Data From Its Creators To Cheat At Its Appointed Task.  Using tricks akin to steganography, a Machine Learning program embedded into images subtle color cues that it could later use to cheat at an image processing task.  In fact, the computer is so good at slipping these details into the street maps that it had learned to encode any aerial map into any street map. It doesn’t even have to pay attention to the “real” street map’ — all the data needed for reconstructing the aerial photo can be superimposed harmlessly on a completely different street map.  “A machine learning agent might even find ways to transmit information to itself, imperceptible to the human eye, in the interest of solving a problem more easily and efficiently. This doesn’t mean AIs are outmaneuvering us per se, but rather that AI scientists need robust measures to validate AI output at each stage of a program's process. Only then can we ensure that neural networks solve problems in the ways we intend them to do so.”

And now- Deep Squeak - computer visualization-learning systems  are being used to “translate” high-pitch vocalizations of lab mice. Perhaps helping ask the ultimate question of Life, The Universe, and Everything.

Researchers at the MIT-IBM Watson Lab are using General Adversarial Networks, or GANs, to help explain how artificial intelligence systems “think.” GANs are a form of AI that pits two neural networks against each other to achieve a larger goal, such as creating new pictures of dogs, human faces or swapping heads in photos. And this approach to using adversarial sub-intelligences to make the macros intelligence smarter is exactly what I talk about in “Disputation Arenas,” and in my novel EARTH.

== Defense… zappers? ==

We’ve entered the era of hypersonic missiles – Putin claims to have them and the U.S. almost certainly does. Is this the end of “strategic defense”? Well, there is a twist to SDI that no one mentions, except it hearkens back to 1967. But I won’t go there. Not here. Suffice it to say that I do not deem that money to be totally wasted. Still…

Some of you may want to get up to date on lasers and beams and such. They have been “the future” for so long that – for many of you – they may have faded into the background. Correct that with "Lasers,Death Rays, and the Long, Strange Quest for the Ultimate Weapon," by my old Caltech classmate Jeff Hecht.  You'll get up to date on what's NOT top secret... and that will help you have perspective about what's still under wraps. (I suspect - and there had better be - a lot.)

== Solar Wonders ==

First see this wonderful innovation.  Refugee tents that can collect rainwater & store solar power

As Russ Daggatt aptly put it: "China's development of its solar industry - both panels and power - may be "unfair" and violate free market fundamentalism. But it's the right thing to do for China - and the global environment. I wish our government was acting with comparable vigor. (Instead, we're promoting coal.)" The one swallowing a "hoax" is a puppet employed in the White House. The puppeteer is perfectly happy with melting tundra and an ice-free arctic.

Fascinating counterpoints! Says Venture pundit Mark Anderson: "China got its monopoly by dumping and thereby destroying the global leaders in the U.S. and Germany."  

To which Asia expert Scott Foster replied: "When I was sustainability analyst at BNP Paribas, our Chinese economist put it to me this way: 'It takes a lot of energy to make solar panels, which only reach energy break-even after one or two years of operation. In China, most of that energy comes from coal, which pollutes the air, which in turn makes solar panels less efficient. China produces more and more solar panels, driving down prices and profit margins. It floods the world market, of course, but you Americans get very cheap solar panels and make higher margins installing them. America gets cheap clean energy and clean service jobs. China gets low paid manual labor and pollution. And you Americans are complaining about it.'

Interesting way of viewing it. Donald Trump thinks solar panel imports should be taxed and solar subsidies cut so that this dynamic can be reversed. What a deal.

Wow, complicated. The crux? Yay, panel prices plummet! Boo to unfair trade practices and IP theft, Solution, let's make an allowance "for development" and then play fair.