110 karma · joined January 24, 2017
It would be interesting if that independence were regarded as a threat.
Widespread availability of high quality media capture and technologies such as facial manipulation will inevitably lead to a future with many, many unverifiable snippets that go viral. This video is unique in that verification is possible, but the future will likely hold far more unverifiable content than otherwise. It is more positive to expect that humans will gain the ability to balance disinformation in the face of further information than to expect that benign dictators will censor perfect knowledge into cognisance.
The final report is available here: http://www.vpri.org/pdf/tr2012001_steps.pdf
I haven't kept up with subsequent research the Institute has produced. I do know they had an aversion to producing software artifacts and were far more concerned with the written reports (which in some senses restricted the ability of amateurs like me to play with the interesting output by the group). I did play with OMeta (a meta-parser) and the COLA / Id code - which was enlightening!
That aside, I think work has been debased because the values of currencies have been debased by central banks. The 1% fully and significantly invested in the stock market may be feeling somewhat differently, though.
Also, I think to some extent that the tech dreams promised by my childhood have largely been replaced by corporate simulacra that are in no way as satisfying.
As much as I detest the MATLAB licensing reaming (and sometimes the inelegance of the language itself), there is very little out there approaching the sheer depth of the associated toolboxes and Simulink models.
Both are effectively becoming meaningless and set my pretentiousness detector on high alert.
There is a lot of hype regarding deep learning, but I have struggled to find a concise definition apart from the fact that it is now relatively easy to work with monstrously big networks. Backprop and related algorithms have been around for decades. From what I remember of neural nets, one huge drawback was that they would be close to un-debuggable. The learning contained in the net would be inscrutable to a human, to all intents and purposes. Failure data could be recorded and replayed, but any actual reason for failure would frequently not be found. So, tweak the network, resize some layers and try again... I can think of several reasons why that's fundamentally unsuitable to the problem of driving.
I was in Egypt recently, and the sheer amount of lane crossing, merging, pedestrians ducking through multiple lanes of traffic, roadside obstacles, donkey-drawn vehicles etc would be 100% impervious to even a level 3 solution today. I believe the same would be true in India and many other parts of Africa.
So, I really hope that we are not falling blindly into another 5th Generation sinkhole here. History should have taught us better.