collapse

Author Topic: replacement is close, humans  (Read 18041 times)

Offline space otter

  • Hero Member
  • *****
  • Posts: 5659
  • Gold 691
replacement is close, humans
« on: March 20, 2017, 02:52:11 pm »

https://www.technologyreview.com/s/603868/how-deepminds-memory-trick-helps-ai-learn-faster/

Intelligent Machines
How DeepMind’s Memory Trick Helps AI Learn Faster

While AI systems can match many human capabilities, they take 10 times longer to learn. Now, by copying the way the brain works, Google DeepMind has built a machine that is closing the gap.

by Emerging Technology from the arXiv  March 16, 2017



Intelligent machines have humans in their sights. Deep-learning machines already have superhuman skills when it comes to tasks such as face recognition, video-game playing, and even the ancient Chinese game of Go. So it’s easy to think that humans are already outgunned.

But not so fast. Intelligent machines still lag behind humans in one crucial area of performance: the speed at which they learn. When it comes to mastering classic video games, for example, the best deep-learning machines take some 200 hours of play to reach the same skill levels that humans achieve in just two hours.

So computer scientists would dearly love to have some way to speed up the rate at which machines learn.

Today, Alexander Pritzel and pals at Google’s DeepMind subsidiary in London claim to have done just that. These guys have built a deep-learning machine that is capable of rapidly assimilating new experiences and then acting on them. The result is a machine that learns significantly faster than others and has the potential to match humans in the not too distant future.

First, some background. Deep learning uses layers of neural networks to look for patterns in data. When a single layer spots a pattern it recognizes, it sends this information to the next layer, which looks for patterns in this signal, and so on.

So in face recognition, one layer might look for edges in an image, the next layer for circular patterns of edges (the kind that eyes and mouths make), and the next for triangular patterns such as those made by two eyes and a mouth. When all this happens, the final output is an indication that a face has been spotted.

Of course, the devil is in the details. There are various systems of feedback to allow the system to learn by adjusting various internal parameters such as the strength of connections between layers. These parameters must change slowly, since a big change in one layer can catastrophically affect learning in the subsequent layers. That’s why deep neural networks need so much training and why it takes so long.

Pritzel and co have tackled this problem with a technique they call neural episodic control. “Neural episodic control demonstrates dramatic improvements on the speed of learning for a wide range of environments,” they say. “Critically, our agent is able to rapidly latch onto highly successful strategies as soon as they are experienced, instead of waiting for many steps of optimisation.”

The basic idea behind DeepMind’s approach is to copy the way humans and animals learn quickly. The general consensus is that humans can tackle situations in two different ways. If the situation is familiar, our brains have already formed a model of it, which they use to work out how best to behave. This uses a part of the brain called the prefrontal cortex.

But when the situation is not familiar, our brains have to fall back on another strategy. This is thought to involve a much simpler test-and-remember approach involving the hippocampus. So we try something and remember the outcome of this episode. If it is successful, we try it again, and so on. But if it is not a successful episode, we try to avoid it in future.

This episodic approach suffices in the short term while our prefrontal brain learns. But it is soon outperformed by the prefrontal cortex and its model-based approach.

Pritzel and co have used this approach as their inspiration. Their new system has two approaches. The first is a conventional deep-learning system that mimics the behaviur of the prefrontal cortex. The second is more like the hippocampus. When the system tries something new, it remembers the outcome.

But crucially, it doesn’t try to learn what to remember. Instead, it remembers everything. “Our architecture does not try to learn when to write to memory, as this can be slow to learn and take a significant amount of time,” say Pritzel and co. “Instead, we elect to write all experiences to the memory, and allow it to grow very large compared to existing memory architectures.”

They then use a set of strategies to read from this large memory quickly. The result is that the system can latch onto successful strategies much more quickly than conventional deep-learning systems.

They go on to demonstrate how well all this works by training their machine to play classic Atari video games, such as Breakout, Pong, and Space Invaders. (This is a playground that DeepMind has used to train many deep-learning machines.)

The team, which includes DeepMind cofounder Demis Hassibis, shows that neural episodic control vastly outperforms other deep-learning approaches in the speed at which it learns. “Our experiments show that neural episodic control requires an order of magnitude fewer interactions with the environment,” they say.

That’s impressive work with significant potential. The researchers say that an obvious extension of this work is to test their new approach on more complex 3-D environments.

It’ll be interesting to see what environments the team chooses and the impact this will have on the real world. We’ll look forward to seeing how that works out.

Ref: Neural Episodic Control : arxiv.org/abs/1703.01988



....................................


http://www.msn.com/en-us/news/technology/deepmind-ai-learns-to-remember-previous-knowledge/ar-AAoneKY

DeepMind AI learns to 'remember' previous knowledge
 Engadget Engadget
Jon Fingas
5 days ago


© Provided by Engadget
For all the talk of artificial intelligence becoming increasingly brain-like, there's one area where it frequently falls short: memory. Neural networks usually have to learn everything they need to know about their duties, rather than building on top of existing experiences like real brains do. Alphabet's DeepMind team hopes to fix that. They've crafted an algorithm that lets a neural network 'remember' past knowledge and learn more effectively. The approach is similar to how your own mind works, and might even provide insights into the functioning of human minds.

Much like real synapses, which tend to preserve connections between neurons when they've been useful in the past, the algorithm (known as Elastic Weight Consideration) decides how important a given connection is to its associated task. Ask the neural network to learn a new task and the algorithm will safeguard the most valuable connections, linking them to new tasks when relevant. In tests with 10 classic Atari video games, the AI didn't need learn how to play each game in isolation. It could learn them sequentially, taking the knowledge accrued in one game and applying it to the other.

The technology is more than a little rough around the edges. It's a jack of all trades, but a master of none. A single-task neural network is still better when limited to one game, DeepMind's James Kirkpatrick says toWired. It's also not ready to adapt to situations on the spot. The algorithm shows that it's at least possible to give AI memory-like functions, however. And what DeepMind has learned here could shed light on how real brains consolidate information -- it may well validate theories that have existed for years.

DeepMind, PNAS
« Last Edit: March 20, 2017, 02:55:07 pm by space otter »

Offline The Seeker

  • grouchy, old, but inquisitive...
  • Administrator
  • Hero Member
  • *****
  • Posts: 3757
  • Gold 426
  • The one-armed Bandit
Re: replacement is close, humans
« Reply #1 on: March 20, 2017, 03:19:56 pm »
Sounds like Skynet is growing faster and getting closer to becoming self-aware  8)
Look closely: See clearly: Think deeply; and Choose wisely...
Trolls are crunchy and good with ketchup...
Seekers Domain

Offline Amaterasu

  • The Roundtable
  • Hero Member
  • *****
  • Posts: 6713
  • Gold 276
  • Information Will Free Us
    • T.A.P. - You're It
Re: replacement is close, humans
« Reply #2 on: March 21, 2017, 05:46:46 am »
Sounds like Skynet is growing faster and getting closer to becoming self-aware  8)

Mimick "Self awareness," maybe, but all machines are reliant on their programming.  I suggest that machines will never achieve true Self awareness.
"If the universe is made of mostly Dark Energy...can We use it to run Our cars?"

"If You want peace, take the profit out of war."

Offline Somamech

  • Order of the Mystic Star
  • Hero Member
  • *****
  • Posts: 3355
  • Gold 337
Re: replacement is close, humans
« Reply #3 on: March 21, 2017, 12:27:23 pm »
Can Machines Party?  :o

Nope but the Beastie boys can :P



I lament the day's when True Thinker's were featured in the MSM... They were pushing for Tibetan Freedom and getting Banned by China...now our Military are too scared to scuttle that bullshit even though it has been declared BS legally. 








Offline Somamech

  • Order of the Mystic Star
  • Hero Member
  • *****
  • Posts: 3355
  • Gold 337
Re: replacement is close, humans
« Reply #4 on: March 21, 2017, 12:33:56 pm »
If anyone here has Netflix, please do watch the TV Serie's called Black Mirror.

It's brilliant in depicting the near future and the drama's that may lay ahead... serious!


Offline space otter

  • Hero Member
  • *****
  • Posts: 5659
  • Gold 691
Re: replacement is close, humans
« Reply #5 on: March 31, 2017, 08:11:55 pm »
this guy is a bit hard to listen too but if you can stay awake the info is verry interesting
and scary

min 18  smart dust / magnetically activated nano particles
and  it gets scarier









DARPA Insider says Scientist on Verge of Cataclysmic Discovery FULL VIDEO

Published on Mar 22, 2017
DARPA Insider Explains Where Were At Today.

Robert Duncan is a scientist with multiple degrees from Dartmouth College and Harvard University in Applied Sciences (e.g., artificial intelligence, pattern recognition and robotics) and Business. He has worked on projects for the CIA, Department of Defense, and Department of Justice. In 2004 he testified before the Senate Intelligence Committee, Judiciary Committee, and 23 members of Congress regarding non-consensual experimentation on American citizens.

In 2010 he authored, in partnership with an anonymous consortium of scientists calling themselves the “Mind Hacking Strategy Group,” a book entitled Project: Soul Catcher, Secrets of Cyber and Cybernetic Warfare Revealed. He has appeared on the radio show Coast To Coast and was interviewed by Jesse Ventura for the “Brain Invaders” episode of the television program Conspiracy Theory. [1]

I cited Duncan’s work in an article I wrote for Veterans Today and Nexus Magazine called “Mind Control In The 21st Century” (Part I and Part II). Recently I had the opportunity to speak with him.

In the interview posted below we discuss, among other topics, remote “mind reading,” CIA hive mind experimentation (multiple people sharing the same mental “space”), how people can be targeted and tracked remotely via their energy signature or “brain print,” the existence and use of a technology for remotely “cloning” or “copying” thoughts, emotions and other states (including intense pain) onto a target, [2] and how a weapons system capable of totalitarian control has been developed (with non-consensual and often-times brutal experimentation on civilians ongoing).

Duncan also reveals that many people are now in wireless mental contact with artificial intelligence run on supercomputers.

It is worth noting that current research projects, according to Duncan, have targeted people from all walks of life, not merely (or mainly) the “downtrodden,” as was done in the past.

Some of these targets are threatened with harm to themselves or family members if they do not carry out the instructions of their “handler,” or are promised release from their tortures if they perform some action. [3]

For example, Myron May, the Florida State University shooter, a former assistant district attorney, posted on Facebook before the shooting:

Has anyone been asked to kill promise of freedom

Duncan, who has interviewed many targets, has said that he immediately recognized the “torture script” used on May after listening to a recording in which May discusses his ordeal.

Perhaps most frightening is Duncan’s contention that the scalability of this technology is being tested: “…[A] lot of people won’t believe me and that’s fine. But they’re increasing this. They are scaling this technology to the entire population.

[1] A partial summary history of his work (classification prevents full disclosure) includes: Tank recognition from satellites, submarine signature recognition from acoustic underwater phased arrays, stock market statistical arbitrage, voice and vascular signature recognition, truth detection systems, robotic surgery in medicine, EEG controlled robots, and “brain print” recognition.

[2] The technology is called “EEG Cloning” or “EEG Heterodyning.” The brain patterns (electromagnetic signals) “cloned” onto the target may be catalogued brain patterns stored in computers, though live signals from a “real person’s” brain could also be transmitted to the target.

[3] The handler communicates with the target through a technology for transmitting voices into people’s heads which goes by many names, including “voice to skull,” synthetic telepathy,” “microwave hearing,” and the “Frey effect.” The communication may also be accomplished via EEG heterodyning, which involves “re-mapping the auditory cortex of the human brain.” Other technologies for transmitting voices include “ultrasonic heterodyning” and bone conduction techniques.
Category
People & Blogs
License
Standard YouTube License

« Last Edit: March 31, 2017, 08:40:19 pm by space otter »

Offline space otter

  • Hero Member
  • *****
  • Posts: 5659
  • Gold 691
Re: replacement is close, humans
« Reply #6 on: March 31, 2017, 09:42:55 pm »


Quote
Can Machines Party? 

hummmmmmmmm, you decide  bwhahahahahaha




Robert M. Trammell Music Conservatory and the Sharon Wick School of Engineering-Farm Machine Music





Gotye - Somebody That I Used to Know - Old School Computer Remix


and just a a bit creepy



"Resonant Chamber" - Animusic.com

Offline space otter

  • Hero Member
  • *****
  • Posts: 5659
  • Gold 691
Re: replacement is close, humans
« Reply #7 on: March 31, 2017, 10:24:57 pm »

 i've been looking around and this stuff and sorry  too late to add this to reply 6

just an observation
looking at his picture  at this link after watching the video of him makes me think he is medicated..the face flush and the face swelling


https://www.drrobertduncan.com/

Subjects needed for research


We have a limited number of prototypes for which we are collecting data to alleviate electromagnetic sensitivity or prove to the government that regulations must be passed to protect a certain population.

In this study we are conducting tests for a double blind study for a particular group which will give us the data we need to proceed with this design.

These are the qualifications that you must meet to participate:

You must currently and for the foreseeable future live in a house in the part of the United States that has moist soil. The house is necessary because if it is wired correctly, the third prong of your electrical outlet should be Earth grounded. Trailers and dry soil/sand will not work correctly for this study.
You must submit a $200 deposit for borrowing the prototype to ensure that theft does not occur of these prototypes. This is fully refundable – shipping costs within 90 days. We realize that some people may not want to give them back if they work for them.
Instructions of use are included with the device. Follow them exactly.
Send the survey results back in the pre-paid envelope provided.
Important! For this study we need “Targeted Individuals” that fit this profile exactly. The person must have a reasonably good bill of health including mental health other than hearing voices. The voices must say things like “We are giving you cancer. We are going to kill you. We are going to kill your family. You are ugly. Etc.” Typical psychological breakdown techniques. The voices should be somewhat mild to moderate. This is due to the sensitivity of the individual. The dampening algorithm and field cannot fully detune the fully tortured and scrambled individual. We need this data pure so please be honest or you will destroy the truth that we seek for further progress.

If you fit this profile and would like to participate in this study, send your name, email address, physical address, phone #, and a one paragraph description of your symptoms to HigherOrderThinkers@gmail.com. Your information will be recorded so you must feel comfortable with that.

Prototype bio-field dampener:


 also  money list of parts needed
...................................
Continue to drrobertduncan.com now

A holiday message from Dr. Robert Duncan A.B., S.M., M.B.A., Ph.D. and Todd Giffen

December 12, 2016

this seems to take you to ctc
https://www.drrobertduncan.com/index.html

......................................

http://www.declassifieddocuments.com/2016/10/cia-engineer-dr-robert-duncan-targeted-individual-cia-whistleblower-black-project-scientist.html

Monday, October 24, 2016
CIA Engineer Dr. Robert Duncan
Dr. Robert Duncan is an Ex-CIA black project scientist who figured out they were targeting innocent people domestically and became a whistleblower. He has some great advice for targeted individuals and some great information in some areas.

This page is dedicated to information about or from Ex-CIA Engineer / Whistleblower Dr. Robert Duncan. Anyone willing to courageously blow the whistle on sources like the CIA has earned my respect.


............................

well i don't know how to just  list the you tube  so sorry here's the whole thing




.........................................


search with his name gives  lots of links

https://www.google.com/#q=dr.+robert+duncan&*




Offline zorgon

  • Administrator
  • Hero Member
  • *****
  • Posts: 21309
  • Gold 903
Re: replacement is close, humans
« Reply #8 on: April 01, 2017, 01:58:19 am »
Mimick "Self awareness," maybe, but all machines are reliant on their programming.  I suggest that machines will never achieve true Self awareness.

Ah but you are WRONG  :P  We are very close to self aware machines




Offline ArMaP

  • Administrator
  • Hero Member
  • *****
  • Posts: 13171
  • Gold 770
Re: replacement is close, humans
« Reply #9 on: April 01, 2017, 05:39:07 am »
Mimick "Self awareness," maybe, but all machines are reliant on their programming.
Yes, but you can program a machine to change its programming, I did that some years ago (with a very limited list of changes for a few specific commands), and that's how polymorphic computer viruses work.

Offline The Seeker

  • grouchy, old, but inquisitive...
  • Administrator
  • Hero Member
  • *****
  • Posts: 3757
  • Gold 426
  • The one-armed Bandit
Re: replacement is close, humans
« Reply #10 on: April 01, 2017, 05:48:43 pm »
Mimick "Self awareness," maybe, but all machines are reliant on their programming.  I suggest that machines will never achieve true Self awareness.
You seem to not be cognizant of the fact that you are a biologic machine, programmed from birth, and that 2 pounds of grey matter in your noggin is nothing more than a highly advanced computer that performs millions of functions a second, reliant on all your programming you have amassed since birth
 ::)
Seeker
Look closely: See clearly: Think deeply; and Choose wisely...
Trolls are crunchy and good with ketchup...
Seekers Domain

Offline Phedre

  • The Matrix
  • Full Member
  • *****
  • Posts: 232
  • Gold 14
Re: replacement is close, humans
« Reply #11 on: April 01, 2017, 09:30:59 pm »



There is one thing that we have, that cannot be replicated . That is the Soul.  I would say most are not even aware they have one. And we are the only ones that can destroy our own.  They are not placed in all species. There is nothing religious about the soul, it is way more mysterious, than any religion could ever be.

Maybe if we where more aware of the "special cargo " we carry with-in us it would give us "pause" for true thought. :)

Offline A51Watcher

  • Administrator
  • Hero Member
  • *****
  • Posts: 4429
  • Gold 593
Re: replacement is close, humans
« Reply #12 on: April 18, 2017, 12:27:36 am »

IBM Watson, Smartest Machine ever built




Jeopardy Watson IBM Fast Computer Artificial Intelligence Software Win Million Donation




Ken Jennings: Watson, Jeopardy and me, the obsolete know-it-all





Offline robomont

  • Inventor's Group
  • Hero Member
  • *
  • Posts: 4415
  • Gold 215
Re: replacement is close, humans
« Reply #13 on: April 18, 2017, 03:30:08 am »
im still not sure about watson,may be a giant psyop.i want to test it myself.
ive never been much for rules.
being me has its priviledges.

Dumbledore

Offline Irene

  • 🐾 👀 🐾
  • Global Moderator
  • Hero Member
  • *****
  • Posts: 1420
  • Gold 519
Re: replacement is close, humans
« Reply #14 on: April 18, 2017, 07:49:15 am »
A51,

The sound is wonky on that first Watson video.

Irene  :)
Shipwrecked by the laughter of the gods.....

 


Wal-Mart.com USA, LLC
affiliate_link
Free Click Tracking
Wal-Mart.com USA, LLC

* Recent Posts

Re: kits to feed your family for a year by Shasta56
[March 17, 2024, 12:40:48 pm]


Re: kits to feed your family for a year by space otter
[March 16, 2024, 08:45:27 pm]


Re: kits to feed your family for a year by Shasta56
[March 16, 2024, 07:24:38 pm]


Re: kits to feed your family for a year by space otter
[March 16, 2024, 10:41:21 am]


Re: Full Interview - Lance Corporal Jonathan Weygandt (1997) by RUSSO
[March 12, 2024, 07:22:56 pm]


Re: Full Interview - Lance Corporal Jonathan Weygandt (1997) by RUSSO
[March 09, 2024, 03:25:56 am]


Re: Full Interview - Lance Corporal Jonathan Weygandt (1997) by RUSSO
[March 09, 2024, 02:33:38 am]


Re: Music You Love by RUSSO
[March 09, 2024, 01:10:22 am]


Re: The Man Who Built UFOs For The CIA (Not Bob Lazar!) by RUSSO
[March 09, 2024, 12:14:14 am]


Re: Full Interview - Lance Corporal Jonathan Weygandt (1997) by RUSSO
[March 09, 2024, 12:08:46 am]


Re: A peculiar stone in DeForest by Canine
[March 03, 2024, 11:54:22 am]


Re: The Man Who Built UFOs For The CIA (Not Bob Lazar!) by kevin
[March 03, 2024, 11:30:06 am]


Re: The Man Who Built UFOs For The CIA (Not Bob Lazar!) by kevin
[March 03, 2024, 11:21:15 am]


Re: The Man Who Built UFOs For The CIA (Not Bob Lazar!) by kevin
[March 03, 2024, 11:16:05 am]


Re: Music You Love by RUSSO
[March 02, 2024, 07:58:09 pm]


Re: Full Interview - Lance Corporal Jonathan Weygandt (1997) by RUSSO
[March 02, 2024, 07:50:59 pm]


Re: The Man Who Built UFOs For The CIA (Not Bob Lazar!) by RUSSO
[March 02, 2024, 07:43:03 pm]


Re: The Man Who Built UFOs For The CIA (Not Bob Lazar!) by RUSSO
[March 02, 2024, 07:41:30 pm]


Re: The Man Who Built UFOs For The CIA (Not Bob Lazar!) by kevin
[March 01, 2024, 11:54:23 am]


Re: The Man Who Built UFOs For The CIA (Not Bob Lazar!) by kevin
[March 01, 2024, 11:34:15 am]

affiliate_link