Tuesday, November 29, 2016

Google teaches robots to dream

ExtremeTech reports:
Following in the wake of recent neuroscientific discoveries revealing the importance of dreams for memory consolidation, Google’s AI company DeepMind is pioneering a new technology which allows robots to dream in order to improve their rate of learning. Not surprisingly given the company behind the project, the substance of these AI dreams consists primarily of scenes from Atari Video games. DeepMind’s earliest success involved teaching AI to play ancient videos games like Breakout and Asteroids. But the end game here is for robots to dream about much the same things humans do – challenging real world situations that play important roles in learning and memory formation. ...

One of the primary discoveries scientists made when seeking to understand the role of dreams from a neuroscientific perspective was that the content of dreams is primarily negative or threatening. Try keeping a dream journal for a month and you will likely find your dreams consist inordinately of threatening or awkward situations. It turns out the age old nightmare of turning up to school naked is the rule rather than the exception when it comes to dreams. ...

DeepMind is using dreams in a parallel fashion, accelerating the rate at which an AI learns by focusing on the negative or challenging content of a situation within a game. ... it does seem increasingly likely that AIs could soon dream of socially awkward situations like showing up to school naked.
Okay, I think that Google is trolling us. Next it will be announcing that quantum AI computers are dreaming in parallel universes.

Meanwhile, the British science journal Nature continues to whine about the American election:
he Oxford Dictionaries named ‘post-truth’ as their 2016 Word of the Year. It must sound alien to scientists. Science’s quest for knowledge about reality presupposes the importance of truth, both as an end in itself and as a means of resolving problems. How could truth become passé?

For philosophers like me, post-truth also goes against the grain. But in the wake of the US presidential election and the seemingly endless campaigns preceding it, author Ralph Keyes’s 2004 declaration that we have arrived in a post-truth era seems distressingly plausible.

Post-truth refers to blatant lies being routine across society, and it means that politicians can lie without condemnation. ...

The lack of public indignation when political figures claim disbelief in response to scientific consensus on climate change is part of this larger pattern.
There was a scientific consensus that Hillary Clinton should be, and would be, the next US President.

The elites got caught lying to us about immigration, trade deals, foreign wars, Russia, Islam, Common Core, urban crime-fighting, and an assortment of other issues.

The big-shot scientists also tell us to believe in supersymmetry, quantum computing, quantum gravity, entangled black holes, parallel universes, embryonic stem cell miracle cures, extraterrestial intelligence, and many other crazy things.

Apparently, Trump taught some ppl to question what the elites tell us, and this is very disturbing to a lot of professors.

Here is another wake-up call about modern science:
A bug in fMRI software could invalidate 15 years of brain research ...

Functional MRI (fMRI) is 25 years old, yet surprisingly its most common statistical methods have not been validated using real data. Here, we used resting-state fMRI data from 499 healthy controls to conduct 3 million task group analyses. ...

but instead we found that the most common software packages for fMRI analysis (SPM, FSL, AFNI) can result in false-positive rates of up to 70%. These results question the validity of some 40,000 fMRI studies and may have a large impact on the interpretation of neuroimaging results.
Oops. For some reason, both experts and the general public are extremely gullible on subject of scientific claims based on DNA or brain scans.


  1. Would you PLEASE stop being right all the time: You wrote:
    There was a scientific consensus that Hillary Clinton should be, and would be, the next US President...
    Apparently, Trump taught some ppl to question what the elites tell us, and this is very disturbing to a lot of professors.

  2. I've noticed on most of the science blogs I read that there is an incredibly strong political consensus and component to what is called 'science'. It is common to see the comment that conservatives are 'anti-science' and 'deniers', and just not good at science. When I ask the question "And how many people who do not share your political views are in your department?" I rarely get any answer. I suppose if you surround yourself with people who all have the same exact politics you might think everyone agrees with you merely because of your brilliance. What I find disturbing is the inability of the technical 'elite' to even acknowledge they even have a political bias, if they can't see this, what else are they blind to?
    This also makes me think there should be a firewall between academics in the sciences and politics, else all you get is political science.