This issue sponsored by
Discover the A-B-Cs of BricsCAD at the inaugural Australasia BricsCAD Conference 19 March 2019 in Brisbane, Australia Book your tickets at http://www.trybooking.com/YIUA |
|
|
Readers Respond
Will Computers Revolt?
|
|
Best thing I've read in years! I've been deep into the MIT AGI courses (online) and absorbing the views and assertions of everyone interviewed on Lex Fridman's YouTube channel. So many angles and levels to this discussion.
It's as tough as defining "existence" itself. One aspect that rarely gets much attention is the legal ramifications. For example, it's been argued that once your thoughts/memories are uploaded outside your corporeal body, they no longer belong to you exclusively. Much like who has legal rights to your cloud content.
And if your original biological construct is augmented with mechanical, or even bio-engineered replacements, at what point can responsibility and liability be challenged for any subsequent actions?
In short, when I opened your newsletter this time it was like I won the lottery. Thank you! :) - David Stein
The editor replies: Thanks for your added insights. I hadn't thought about the legal aspects.
- - -
Personally I think the Swiss cheese metaphor, pointing to the obvious technology gaps, is not a particularly strong defense. Who could have anticipated how machines can now beat humans in games like Jeopardy and Go, the logic and intuition of game playing, captured in a computer program. Perhaps a simulated consciousness could also be programmable. - Dave Jeffries
The editor replies: It turns out that the problem with AI playing Jeopardy and Go is that the algorithms are "not scalable," to use the computer term. This means that the AI code for Go cannot teach itself beyond Go by playing something else, like checkers or chess, or going on to diagnose cancer.
This major admission was made just a few months ago by one of the leaders of AI programming. Here is what the problem is, known as Goedel's Incompleteness Theorem (see https://www.crisismagazine.com/2017/limits-artificial-intelligence):
Only humans can act in ignorance, or refuse to follow what truth they know, or go beyond a given set of rules. A computer must follow what truth is programmed into it, inexorably proceeding from one step to another, never able to transcend itself, never to recognize, and never go against or beyond its own consistency.
Mr Jeffries responds: I am certainly out of my depth here. i do agree that it is difficult to see how machines could ever become conscious in the way humans are. but I do think that, just as the skills of driving and playing Go have been simulated, it might be possible to program a simulated consciousness.
It seems that Alexa at least is headed that way: https://www.fastcompany.com/40479207/for-amazon-the-future-of-alexa-is-about-the-end-of-the-smartphone-era
i keep thinking of the movie version of I, Robot. Have you seen that?
The editor replies: I saw it when it first came out -- in a movie theatre, no less! There is, however, a vast difference between what is claimed as possible (a million people living on Mars -- I doubt that will ever happen -- or traveling faster than the speed of light, as in Star Trek), and what we can actually accomplish.
One of the flaws in humans is that we think linearly into the future. So we take a couple of possible actions, like: 1. We traveled to a space station 2. We traveled to the moon And then we extrapolate linearly: 3. We can also travel to Mars 4. We can also travel to other galaxies, really quickly
What we cannot predict are Black Swans (or chaos events, like the Yellow Jacket movement in France or spaceships exploding in space) that disrupt the present and so cause a severe swerve to a different future.
- - -
"In my lifetime, the singularity will happen. But why does everyone think they'd be hostile? That's our brain assuming it's evil. Why on earth would it need to be? People are just terrified of change."
This is wrong, and it must stop. It's not only emotionally disturbing, it's physically disturbing. Other than an academic achievement, there is little merit to it. The potential for being evil is there, with dire uncontrollable consequences for the human race. It's not worth the risk. It will be attempted by more than one "pristine" individual, with little or no oversight. It can become uncontrollable. Look at the recent flack the Chinese genetic expert, Dr. He, caused with world's first gene-edited babies. - Dik Coates Canada
- - -
I don't think computers will take over; rather, business leaders will see them as a means to cut costs, even in creative industries. Many of the manual functions of design work will be handed over to them. This includes things like building-layout AI for architecture, detailing using massive material and construction databases, and CAC (computer-aided construction).
Once BIM is married to programmatical front ends (Python, AI scripting), you will see whole new fields evolve. - Dave Edwards Dave Edwards Consulting
- - -
My personal thinking is that there is no way human brain can be computerized, simply by the fact that each human being has a unique, personal model. We will see how readers will react. Frankly, I got lost in some of the expressions used -- they are simply beyond me. - Herb Grabowski
The editor replies: The Singularity suffered a setback last week when nVIdia ceo Jensen Huang declared Moore's Law dead.
- - -
I don't think AI will be hostile by design or by default, but it is entirely possible that (after careful analysis) AI will just conclude the world, as a whole, would be better off without us!
Not my idea, but one that I agree with. It is perhaps best expressed in Marshall Brain’s (yes, really!) book The Second Intelligent Species: How Humans Will Become as Irrelevant as Cockroaches, which looks at the potential social and economic impact of AI. See Amazon http://a.co/d/iqU0Lb6 - Robin Capper New Zealand
- - -
You wrote, “Some estimate the brain's power to be 30 terraflops at the low end, and 1,000 terraflops at the upper end.”
More important than big numbers is that it has to be an analogue computer, not a digital one. However hi-res a digital simulation is, there still are no-data gaps between the [discrete] data points, which have to be extrapolated. In real life, the universe, brains, etc don’t have such gaps – their data is continuous.
Chaos Theory (which isn’t really chaos, but stupendously fine-grained determinism complicated by multiple-solution nonlinear equations) shows that the tiny fluctuations (the flap of the butterfly wings) are the very things that tip a fine balance into a major phenomenon (hurricanes).
Digital simulations (or neural nets) can never be so hi-res as to capture all, most, or even any of the subtlest fluctuations which happen undetected in those data gaps, and almost never precisely on the data point. So digital has to assume that the tiniest fluctuations are insignificant.
Lesson no.1 of Chaos Theory is that those tiniest fluctuations are (or may prove to be) in fact the all-important ones. The only computer powerful enough to simulate the universe is the universe itself. Not quite the same size-logic, but probably the only computer capable of really, fully simulating the human brain, is the brain itself. - Tom Foster Tom Foster Architecture, England
The editor replies: Your statement that the brain is analogue is better than my statement that it doesn't perform floating point operations. Thanks for that!
- - -
That's a great book review and some good philosophy. I shared it around with a couple of non-CAD, non-tech relatives. - Jess Davis, president Davis Precision Design, Inc.
The editor responds: I am glad you found it useful. It is an aspect of AI that people in technology have little interest in addressing, and that the mainstream media doesn't understand.
- - -
Your Ship of Theseus reference reminded me of a visit to the famous USS Constitution in Boston a few years ago. I turned to one of the Navy people and jokingly said "Wow! I didn't know they had polypropylene rope in the 1700's!" He replied that modern air pollution was destroying the original rope material. I asked how much of the ship was still original and he looked around carefully before whispering, "A few planks down near the keel." - Bill Fane
The editor replies: Similarly, all dinosaurs skeletons on public display at places like the Royal Tyrrell Museum are fiberglass replicas, and most of the castles in Europe were expertly reconstructed in the last century, except for a few rows of stone down near the ground. |
|
ARES KUDO: MODERN DWG EDITING IN THE CLOUD
ARES Kudo is the only online CAD solution offering a full set of 2D features to read, create, modify, and share DWG and DXF drawings.
Nothing to install. It runs in your Web browser and your files follow you on any device. ARES Kudo is integrated with cloud storage services like Google Drive, Dropbox, Box, OneDrive, OneDrive for Business -- and industry-specific solutions such as Onshape and Trimble Connect.
Visit www.graebert.com/areskudo to learn more and activate your own free trial at kudo.graebert.com |
|
|
Sponsor: Okino Graphics
== Okino's PolyTrans|CAD Software for Professional 'Load & Go' 3D Conversions ==
For over three decades, mission-critical 3D conversion software from Okino of Toronto has been used effectively by tens of thousands of professionals. We develop, support, and convert between all major CAD, DCC, and VisSim formats. CTO Robert Lansdale and his team tailors each package to the specific conversion requirements or problems of each customer.Popular CAD data sources we support include SolidWorks, ProE/Creo, Inventor, AutoCAD, Revit, Navisworks, DGN, IGES, STEP, Parasolid, and JT. DCC data sources are Cinema-4D, 3ds Max, Maya, FBX/Collada, and many more.Popular CAD data sources include SolidWorks, ProE/Creo, Inventor, AutoCAD, Revit, Navisworks, DGN, IGES, STEP, Parasolid, and JT. DCC data sources are Cinema-4D, 3ds Max, Maya, FBX/Collada, and many more.Perfected over three decades, we know 3D data translation intimately, providing you with highly personalized solutions, education, and communication. Contact CTO Robert Lansdale at [email protected] http://www.okino.com
|
|
Re: Issue #1,000
Congratulations on publishing your 1,000th newsletter. - Don Beaton
- - -
Congratulations Ralph! That's an unbelievable milestone to reach. I must admit I haven't read them all, but I read a lot of them, since CAD has been my profession from the very beginning. I still own a very old AutoCAD user manual. I think it's 1986 or so. Thanks for all the good work and keep it up, old fellow! ;-) - Siem Eikelenboom The Netherlands
- - -
This is a remarkable achievement, particularly when one considers the consistently high level of quality information you deliver with every issue.
I retired a few years ago and have no direct use for the information you provide. I scan some of your articles and am always impressed by where you go for a story and the quality of your presentation. I would not have missed receiving this issue or writing this complimentary note for anything.
God bless you. I wish you and yours the very best. - Carl Howk
- - -
Congrats on a 1,000 issues -- seems like only yesterday. - Dave Edwards Dave Edwards Consulting
- - -
My colleagues and I admire your unique and fruitful activity on promotion of all that is meaningful in the domain of engineering software. Sending you our congratulations and best wishes on your 1000th anniversary. - David Levin Ledas, Russia
- - -
Congratulations on 1,000 issues. I didn't see much CAD content in this one, aside from the duck drawing complete with an intestinal spline, yet an important topic to consider. I found it very worthwhile reading, as usual, and am looking forward to many more issues! - Doug Vander Wilt
The editor replies: I had considered stopping at issue #999 or maybe #1,001, but there is too much happening in the CAD world to stop right now. In anniversary issues, I usually look back in time. But this time I thought I would look forward to the most forward thinking topic there is: AI replacing the human brain.
- - -
Wow! Congratulations on reaching the 1K milestone Ralph!
And what a topic to explore. I’m sure there will be a lot of comments on that piece -- very interesting to see where modern thought takes us. I rather side with the apostle Paul who wrote to Christians in Corinth, “And now abide faith, hope, love, these three; but the greatest of these is love.” Those are the things worth being human for.
Thanks again for the CAD news. - Ron Powell
- - -
Congratulations on issue #1,000! That is a great accomplishment. That has to make you #1 in this arena. - Jim Balding The ANT Group |
|
Thank you to readers who donate towards the operation of upFront.eZine:
- Risto Juhani Saarikko, Canada
- Key Glass LLC, USA
- Jim Balding at The ANT Group: "Ralph, I don't know how you do it. You never fail to bring straight-shooting CAD information from all over the globe. Thanks for your continued dedication to us, the readers."
Should you wish to support upFront.eZine through PayPal, then the suggested amounts are like these:
Should Paypal.me not operate in your country, then please use www.paypal.com and the account of [email protected].
Or mail a cheque (US$ or CDN$ only, please) to upFront.eZine Publishing, Ltd., 34486 Donlyn Avenue, Abbotsford BC, V2S 4W7, Canada. |
|
"All my life I wanted to be someone; I guess I should have been more specific." - Jane Wagner, Slashdot 2018-12-01; h.t. to D.B. |
|
|
|
Recent Comments