sawwit

Fair point. I somehow was under the false impression that no one ever uses these Flash Player projectors, but it seems to have been supported for a long time according to the WINE app DB. Still awesome. =D

I'm frequently surprised that wine just works even with the most obscure and odd programs, e.g. exe packed Flash SWFs and programs with weird custom GUIs.

> People start to become very irrelevant.

Please be precise here and say they will be irrelevant in economic terms. What will be left are things humans otherwise care about: producing art, consuming art, fun, games, sports, traveling, companionship, partying, building things, learning new things etc.

I'm looking forward to it, and I don't see a reason why anyone wouldn't.

> they will adhere primarily to the profit motive

You are assuming that there is be an obvious way of doing so, an obvious solution to the control problem.

No. Cooperations are not necessarily like artificial intelligences. They are cooperations of human intelligences and these two classes of intelligences have actually very little in common if you look past the similarity that they are potentially very powerful and intelligent. Cooperations are driven by material profit, but in the end there is a reasonably large possibility that they are shaped by human values (because they are run by humans and otherwise people would also refuse to buy their products). The same cannot be said about AIs with high certainty.

I think one of the best evidences we have is the level at which computers outperform the human mind in certain domains. An artificial general intelligence would have very low latency access to all these mathematical and computational tools we've invented (physical simulations, databases, theorem provers), and it would not need to mechanically enter program code on a keyboard, but it would be directly wired to the compiler. It could possibly learn to think in program code and execute it on the fly.

The computation environment of neurons is also extremely noisy (axons are not well insulated) and neurons only fire at 7-200Hz. Assuming noise and low firing rate do not fulfill a certain task in mammalian brains, this is another way in which silicon-based minds could potentially be vastly superior.

Thirdly, assuming sleep is not necessary for intelligence, artificial minds would never get exhausted. They could work 24 hours on a problem a day, which is possibly 5-10 time the amount of thinking time a human can do realistically.

And lastly, an AI could easily make copies of itself. Doing so it could branch a certain problem to many computers which run copies of it and eventually collect the best result, or just shorten the time it takes to get a result. It could also evolve at a much faster rate than humans, assuming it has a genetic description: possibly hours to seconds instead of 20 years. Anyhow, it could easily perform experiments with slightly changed version of itself.

Just because we can formulate an optimization objective, it is not guaranteed that we will find an algorithm that solves it within a reasonable amount of time. In case of humans, these objectives or preferred states are possibly very simple ones, like hunger and pain avoidance, reproduction and curiosity; and it is actually easy to write down an algorithm that optimizes these objectives (if you ignore how reality actually works): Simply try out all possible ways of reacting to the environment and choose the best one.

This works in theory, but in practice you only have a limited amount of chances to try something out (because of the arrow of time). This makes learning a necessity. You need to keep a record of all trials you have performed so that you can reuse this information later when the same situation reoccurs. How to do this in an optimal way is described by Bayes' theorem.

The key to AI will be a certain set of priors, biases and fixed function units that make this computationally tractable; we'll likely need things like invariance of the learned information to various changes so that it can be reused in different settings, segmentation of data coming from the world into episodes (hippocampus), attention, control (basal ganglia), mental rotation (cortex) and path integration (hippocampus, grid cells).

It's perhaps specifically a problem with YouTube, or the JS events for the video playback, or something related to that:

https://bugzilla.mozilla.org/show_bug.cgi?id=1163327

Oh well, I'll just use a bookmarklet/AppleScript to open YouTube videos in VLC, if is too sluggish (VLC supports YouTube URLs via menu item Open Network…).

You are right and I gave it a go: I've tested it with a 640x386 H264 movie clip averaged over the same time span with a blank Firefox profile:

    VLC: 14.0 %
    Firefox: 27.1 %

Edit: Interesting. I tried it with a different video (H264, 720p) and it turned the results around (33% VLC and 28% Firefox), there was more movement in this clip, so it turns out to be more complex to measure this.

This is my Firefox wish list:

* Make no distinction between tabs and bookmarks, so that I can search for tabs in the bookmark library and easily arrange them and move them to bookmark folders etc.

* Add optional tab auto-suspend for heavy JS web apps so that they don't eat all CPU cycles in the background.

* Please do something about graphics accelerated video playback. My machine runs hot with 460p YouTube videos, even with h264ify (110% CPU FF, and by comparison only 34% VLC!).

* Be more keyboard friendly: Escape to defocus the address and search bars, page up/down in the add-ons list.

It will be interesting to test whether one can react fast enough if a drone like this approaches from a shallow angle from the side. It could even approach in free fall the last few meters, so it would be almost noiseless: https://www.youtube.com/watch?v=VxOdXYCRAds

I think a new quality about this kind of weapon is that it can be controlled remotely or can even operate semi-autonomously. Deadly pipe bombs are certainly heavier than a crossbow and ignition mechanisms aren't trivial to build.

> Chairs with arms

I don't remember where, but I've heard that chair with arms are not optimal since they constantly apply pressure to the interior of your forearm.

Admittedly the more interesting part will be the software flying the drone. It seems we are going to need a lot of these [1] and targeted people will need to wear safety goggles.

[1]: http://i.imgur.com/aAh4jwq.gifv

I wrote "almost autonomously".

> It's not much more of a threat to society than a handgun.

These drones could be programmed to target specific groups of people, for example of a certain ethnicity, and attack them almost autonomously. Short range slingshot mechanisms are several orders of magnitude cheaper to build than firearms. Moreover, the inhibition threshold is much lower if you are not involved in first-hand violence. There is also a much lower risk of getting busted and no need for intricate escape planning.

> The risks of things like rogue drones made from cheap commodity hardware, which might be sensible as well …

In Prof. Tegmark’s recent presentation at the UN he mentioned the possibility of extremely cheap drones that approach the victim's face very quickly and pierce a bolt into their brain through one of their eyes. Such a drone wouldn't require high precision projectiles which would make it cheap to build.

> Of course the reason governments don't do this is because almost nobody sees the risks of AI "acting on its own"…

It is near impossible to enforce something like this globally and forever. At best, it would be a near-term solution, especially so because there is a huge military and economic interest in technology and AI. Quite possibly, the only long-term solution is solving the control problem.

Very interesting:

> “Dopamine encodes what are called reward-prediction errors – the ongoing difference between reward expectations and the actual rewards experienced,” Montague said. “From just dopamine signals, we can see when a person expects a reward and whether the person receives the reward. But in our most recent study, we found this earlier model of reward-prediction error to be incomplete. Rather, dopamine pulses appear to combine information about what might have happened with information about what actually happened. This is an entirely new way of viewing the role of dopamine signaling in the human brain.” The idea that “what could have been” is part of how people evaluate actual outcomes is not new. But no one expected that dopamine would be doing the job of combining this information in the human brain. “We married two known computational models into something new,” Montague said. “In doing so, we found dopamine tracking and combining two streams of information into one chemical pulse.”

In Prof. Tegmark’s presentation at the UN he mentioned the possibility of extremely cheap drones that approach the victim's face very quickly and pierce a bolt into the victim's brain through one of their eyes. That wouldn't require a high precision projectile so it will be easy and cheap to build.

New users quickly reach a comment limit. You can create a new account on a different IP and continue.

Paper: https://www.sciencemag.org/content/350/6266/1332.full

Code: https://github.com/brendenlake/BPL

Abstract: People learning new concepts can often generalize successfully from just a single example, yet machine learning algorithms typically require tens or hundreds of examples to perform with similar accuracy. People can also use learned concepts in richer ways than conventional algorithms—for action, imagination, and explanation. We present a computational model that captures these human learning abilities for a large class of simple visual concepts: handwritten characters from the world’s alphabets. The model represents concepts as simple programs that best explain observed examples under a Bayesian criterion. On a challenging one-shot classification task, the model achieves human-level performance while outperforming recent deep learning approaches. We also present several “visual Turing tests” probing the model’s creative generalization abilities, which in many cases are indistinguishable from human behavior.

Bonus link: A panel on automating science which I greatly enjoyed watching.

[1]: https://www.youtube.com/watch?v=VchrEoimSO8

The snake oil?

For more practical persons I recommend skipping the historical parts of philosophy and go straight to Bayesianism. I recommend "Rationality" by E. Yudkowsky as an entry point, "Probabilistic Inference" by J. Pearl, "Artificial Intelligence" by Russel & Norvig for the theory and "Formal Theory of Creativity & Fun & Intrinsic Motivation" by J. Schmidhuber for inspiration.

[1]: http://people.idsia.ch/~juergen/creativity.html

Off topic: This seems to be an odd emphasis on the word "good". Can a native English speaker confirm?

How come they aim for GPU technology rather than for neuromorphic computing ASICs?

Here is a time lapse video of the 9 year construction, quite fascinating to watch: http://www.ipp.mpg.de/115632/zeitraffer_w7x

YouTube mirror: https://www.youtube.com/watch?v=u-fbBRAxJNk

A more technical video explaining how it works: https://www.youtube.com/watch?v=lyqt6u5_sHA

I wouldn't find it too surprising if the brain would abstract actions and actors, perhaps even the tense from thoughts and sensory input and map them to auditive features and motor programs for (sub)vocalization and the internal monologue. I think, visual thinking is different from that. Mental rotation and visualizing objects have no actors and actions in it. It is just recall of movement, shape and texture patterns that have been previously encountered. However, I think visual thinking is often enhanced by thinking in language. The auditive memory can be leveraged as an additional buffer and as a means of verifying and constructing visual thoughts.

An obvious difference is that we do have scientific models of global climate change that give us at least some evidence, but we lack any such information for the case of the existence of a god.

You seem to be saying that ANNs are good at pattern recognition because BNNs are. That just defers it to the question why BNNs are good at it.

OTOH, the parent comment suggested we should draw our attention to the processes that produce the data and find correspondences to how NNs decompose it.

Indeed, this course was great. It's pretty much a summary of the state of the art of learning science.