Steps Toward Super Intelligence Part I: How We Got Here
rodneybrooks.com
rodneybrooks.com
"Note that neural nets are neither. There has been a relatively small amount of non-mainstream work of getting neural nets to control very simple robots, mostly in simulation only."
This is just plain false. Playing Arcade Games involves controlling an agent and there is a lot of widely published, very mainstream work from DeepMind, OpenAI and many other less known research groups. The big part of hype is about using deep neural networks in tandem with reinforcement learning. It's a very reactive system.
In complex environment a work from Nvidia is notable "End to End Learning for Self-Driving Cars": https://arxiv.org/abs/1604.07316
Also, Alpha Zero is very clever combination of reinforcement learning, neural networks and symbolic reasoning (tree search). Alpha Zero is not only reactive, but also deliberative. Interestingly Alpha Zero, in very limited domain, managed to surpass not only unaided humans, but many years of programmers effort based on purely symbolic methods by beating Stockfish in Chess.
The author is no doubt a world class expert in the Good Old AI, but to my surprise he seems to be out of loops with more recent advancements.
Or am I missing something / misunderstood something?
As far as we could tell, Stockfish was used in a situation it is not optimized for (very many cores) and wrong settings.
What we can say is that Alpha Zero is comparable after throwing a lot of specialised compute power at it.
Most importantly, neither system works at a symbolic level or does anything even close to thinking. They handle statistics and see searches extremely well, sometimes tactics, but most of it is still hardcoded - they are fed a high level symbolic representation of the game.
When Alpha Zero can learn playing chess by experimenting with board physics and observing a few videos or at least reading a book, then we can say it can actually produce some form of symbolic thought.
(OpenAI is closer but nowhere near close - it falls back on memory of play as a full state model.)
The effort is lead by Gary Linscott. He is also one of the main developers of Stockfish (I believe the framework for testing Stockfish).
https://github.com/LeelaChessZero/lczero
Here is the current state of the effort: https://docs.google.com/spreadsheets/d/18UWR4FVhPi0vNwwPreu_...
Estimated date of surpassing Stockfish with settings and hardware reported in the spreadsheet is the end of this year.
@edit
I have missed your statements about symbolic reasoning. Yes, I'm referring to the same type of symbolic reasoning as the author of the article. I.e. system that manipulates high-level symbols without any grounded understanding of them in the surrounding us reality. Deliberating whether it is thinking or not is pointless. It's effective and that matters.
Most probably because the vast majority of its applications is in game-playing and not human-level AI (as in general intelligence, capable of tackling free-form tasks in the real world), which is the subject of the article.
To substantiate this intuition further:
a) Agents playing arcade games, Go, chess, etc, are still constrained in "blocks worlds", even if they can beat humans in those worlds.
b) To my knowledge, self-driving car systems actually deployed are not trained using reinforcement learning, or even deep learning (for anything other than vision) but instead make use of the SLAM approach discussed in the article. I might be wrong on this and would welcome a correction.
I can never get past the structural similarities between the singularity prediction and the apocalypse of St. John the Divine. This is not the place to parse it out, but the key thing they have in common is the idea of a rapture, in which some chosen humans will be taken up and made one with the infinite while others will be left behind.
This pretty much exactly what reaching enlightment/nirvana implies. Except it is not a specific point in history for everyone at the same time, but rather something that each individual person can choose to accomplish during their own lives.
Although unfortunately now only existing in the Internet Archive, here's a pretty decent rebuttal of this standard cognitive stopper by Steven Kaas:
http://web.archive.org/web/20110718031848/http://www.acceler...
After pointing out that the similarity is mostly structural (read: accidental, relevant at best as a heuristic) and listing a set of important differences between the two ideas, he finally concludes:
> It’s also interesting to think about what would happen if we applied “Rapture of the Nerds” reasoning more widely. Can we ignore nuclear warfare because it’s the Armageddon of the Nerds? Can we ignore climate change because it’s the Tribulation of the Nerds? Can we ignore modern medicine because it’s the Jesus healing miracle of the Nerds? It’s been very common throughout history for technology to give us capabilities that were once dreamt of only in wishful religious ideologies: consider flight or artificial limbs. Why couldn’t it happen for increased intelligence and all the many things that would flow from it?
Didn't this already happened once? Some chosen animals were lifted to unbelievable heights, while the others were left behind and now are basically at the mercy of the chosen one?
Edit: Similar ideas also appear in Anathem - although obviously in a somewhat different context.
This is an unfair characterization of the the most prominent figures in this area. This article by Allan Dafoe and Stuart Russell from 2016 refutes it: https://www.technologyreview.com/s/602776/yes-we-are-worried...
Man created god(s) in his own image. Man created AI in his own image.