Curated by Michael Junkroski
I have some good news! Despite the existence of a real-life Skynet, AI research is still extremely primitive. Yes, your Tesla can drive itself, more or less, but that’s less a function of artificial intelligence than the result of smoke-and-mirrors tricks with lasers, GPS, and really fast microprocessors.
Artificial Intelligence is a branch of computer science tasked with creating machines that operate more like the human brain. Today, the computer on your desk or in your pocket makes decisions in a very linear fashion. An Operating System is programmed with a very specific instruction set.
Computers designed to use AI require multidisciplinary research in neuroscience and games theory to improve a computer’s ability to “know” a topic well enough to make the same sort of informed decisions about it as an expert in their field.
AI systems are able to access and analyze information non-linearly. The computer can, to a certain extent, process many different concepts simultaneously to reach a “best outcome” result.
There are many ethical issues raised by the rise of artificially intelligent machines—at what point in their development do they have human rights? Can a machine believe in God? Can a machine be God?
That IBM’s Watson can beat a chess master is impressive, but is it a sign of intelligence, artificial or otherwise? Watson has near-instant access to every possible move and counter-move because unlike a human, Watson has perfectly stored. and can instantly access, pretty much every move ever played by every chess master in the history of the game—and every other Master’s pary.
A computer with that capability should win this sort of competition.
It’s impressive, but Watson is still merely a computer processing a larger series of choices exponentially faster than the IBM PC that is its forbear. The programming that makes Watson choose a winning move isn’t artificial intelligence in the Turing sense, it’s just insanely good programming and a ton of memory. Again, smoke and mirrors.
Alan Turing to
In the 1950s the remarkable Alan Turing devised a test designed to determine whether or not a computer could “think” rather than “parrot.” Most computer scientists agree that, despite all appearances, even IBMs Watson isn’t thinking.
In part to avoid the debate about whether or not a computer could ever actually “think,” Turing developed a test (The Turing Test) to prove that a computer could at least fool another human into thinking the machine itself is also a human. Anyone who’s used Apple’s Siri can vouch for just how far away we are from even that basic proof of AI.
As our ability to program computers that actually do make semi-conscious decisions improves (especially as biological computing evolves), it’s imperative for us to discuss the ethical, religious, and spiritual issues raised by a new class of computers—and their programmers.
One foundation leading the conversation is ai and faith. According to their website, ai and faith is “a cross-spectrum consortium of faith communities and academic institutions. Its mission is to bring the fundamental values of the world’s major religions into the emerging debate on the ethical development of Artificial Intelligence and related technologies.”
They have some excellent, thought-provoking articles at https://aiandfaith.org/contributors/
What is Artificial Intelligence? article in PC Mag
Ray Kuzweil is one of the pioneers of AI and a proponent of the idea of singularity, a time when humans and machines will become one, new organism. He is absolutely brilliant and has a series of YouTube videos available that are thought-provoking, especially for people of faith.
Hit up AI News for your daily dose of data about the industry
Wired Magazine’s AI section
Good articles on the ethics of AI at Future of Life Institute
What is biocomputing? A great primer.
The MIT Technology Review has a terrific article about biocomputing as a bridge to true, chemical, molecular computing. Fascinating.
In 2013 a group of scientists in Israel developed a simple biological transducer
Microsoft has a special group focused on designing and programming molecular circuits.
Biocomputing—and nanotechnology—rely on techniques such as protein folding to manipulate DNA, turning genes into simple binary switches that can manipulate certain biological functions. While it’s relatively easy to manipulate proteins in simple organisms, it’s much more difficult to attain consistent results in mammals. This new technique using enzymes instead of proteins might help.
BioVoice is THE journal for everyone interested in this burgeoning field.