Support the Arctic Sea Ice Forum and Blog

Author Topic: Robots and AI: Our Immortality or Extinction  (Read 9719 times)

sidd

  • Young ice
  • Posts: 4935
    • View Profile
  • Liked: 340
  • Likes Given: 0
Re: Robots and AI: Our Immortality or Extinction
« Reply #100 on: September 25, 2019, 01:20:08 AM »
Link to the Tshitoyan paper on automated analysis of chemistry papers:

https://perssongroup.lbl.gov/papers/dagdelen-2019-word-embeddings.pdf

It is rather nice, but does get some things wrong .. positions on the periodic table is one example.

small writeup at

https://clarodatascience.com/2019/07/13/modelling-chemistry-2/

sidd

vox_mundi

  • Nilas ice
  • Posts: 1629
    • View Profile
  • Liked: 639
  • Likes Given: 113
Re: Robots and AI: Our Immortality or Extinction
« Reply #101 on: September 25, 2019, 01:34:12 AM »
Thanks for the link, sidd.  :)
“There are three classes of people: those who see. Those who see when they are shown. Those who do not see.” ― Leonardo da Vinci

Insensible before the wave so soon released by callous fate. Affected most, they understand the least, and understanding, when it comes, invariably arrives too late

vox_mundi

  • Nilas ice
  • Posts: 1629
    • View Profile
  • Liked: 639
  • Likes Given: 113
Re: Robots and AI: Our Immortality or Extinction
« Reply #102 on: September 26, 2019, 05:35:08 PM »
Pentagon Wants AI to Interpret ‘Strategic Activity’ Around the Globe
https://www.defenseone.com/technology/2019/09/pentagon-wants-ai-interpret-strategic-activity-around-globe/160157/

The Defense Department’s startup outreach office is looking for artificial intelligence tools that can keep help Pentagon officials understand trends in world events.

The Defense Innovation Unit last week put out a solicitation for commercial software that can automatically ingest, analyze and derive insights from publicly available information. The tool is expected to plug into a wide variety of global resources, including news outlets, blogs, social media platforms, patent databases and academic journals.

After crunching the data, the “AI-based knowledge graph” would generate visualizations and written reports to help Defense leaders better understand and interpret the “strategic activity” it uncovers, according to the solicitation. The tool would also need to show its work, allowing officials to trace its insights and conclusions back to the source data.

DIU officials want the system to be able to translate information from multiple languages, and detect and redact personally identifiable information.

... A spokesperson for the Office of the Secretary of Defense told Nextgov they hope to have a working prototype of the AI system within 12 months or less. After the prototype, the department will consider awarding a follow-up production contract and potentially scale the tech across the Pentagon, the spokesperson said.

“There are three classes of people: those who see. Those who see when they are shown. Those who do not see.” ― Leonardo da Vinci

Insensible before the wave so soon released by callous fate. Affected most, they understand the least, and understanding, when it comes, invariably arrives too late

nanning

  • Grease ice
  • Posts: 879
  • 0Kg CO2, 35 KWh/wk,130L H2O/wk, No heating
    • View Profile
  • Liked: 135
  • Likes Given: 5783
Re: Robots and AI: Our Immortality or Extinction
« Reply #103 on: September 26, 2019, 06:31:44 PM »
I think THIS is the wet dream of immortality by the powerful crazy people behind the scenes.
"It is preoccupation with possessions, more than anything else, that prevents us from living freely and nobly" - Bertrand Russell
   Simple: minimize your possessions and be free and kind    It's just a mindset.       Refugees welcome

vox_mundi

  • Nilas ice
  • Posts: 1629
    • View Profile
  • Liked: 639
  • Likes Given: 113
Re: Robots and AI: Our Immortality or Extinction
« Reply #104 on: September 26, 2019, 07:27:57 PM »
Agent Smith: Never send a human to do a machine's job.

-----------------------

The standard term in the Pentagon now for human involvement with AI and weapons now is “human ON the loop,” a shift from human IN the loop. That reflects greater stress on the advisory function of humans with AI and a recognition that domains like cyber require almost instantaneous responses that can’t wait for a human.

Next: ... 'humans OUT of the loop'


----------------------------

No AI For Nuclear Command & Control: JAIC’s Shanahan
https://breakingdefense.com/2019/09/no-ai-for-nuclear-command-control-jaics-shanahan/

The Joint Artificial Intelligence Center won't automate nuclear response -- but it is working towards AI on conventional weapons.

In movies like WarGames and Terminator, nuclear launch controls are the first thing fictional generals hand over to AI. In real life, the director of the Pentagon’s Joint Artificial Intelligence Center says, that’s the last thing he would integrate AI with. The military is beginning a massive multi-billion dollar modernization of its aging system for Nuclear Command, Control, & Communications (NC3), much of which dates to the Cold War. But the Joint Artificial Intelligence Center is not involved with it.

A recent article on the iconoclastic website War on the Rocks argued “America Needs A ‘Dead Hand’,” a reference to the Soviet system designed to automatically order a nuclear launch if the human leadership was wiped out. “I read that,” Shanahan told the Kalaris Intelligence Conference here this afternoon. “My immediate answer is ‘No. We do not.’
« Last Edit: September 26, 2019, 09:04:04 PM by vox_mundi »
“There are three classes of people: those who see. Those who see when they are shown. Those who do not see.” ― Leonardo da Vinci

Insensible before the wave so soon released by callous fate. Affected most, they understand the least, and understanding, when it comes, invariably arrives too late

vox_mundi

  • Nilas ice
  • Posts: 1629
    • View Profile
  • Liked: 639
  • Likes Given: 113
Re: Robots and AI: Our Immortality or Extinction
« Reply #105 on: September 30, 2019, 06:50:44 PM »
Robots & AI eliminating jobs ...

AI Just as Good at Diagnosing Illness as Humans
https://www.medicalnewstoday.com/articles/326460.php

----------------------

Even Good Looks Won't Get You a Job; 100,000 Free AI-generated Headshots
https://www.theverge.com/2019/9/20/20875362/100000-fake-ai-photos-stock-photography-royalty-free


None of these photos are of real people

--------------------------------------

Computer Says No: An Expression-Analysing AI has been Picking Out Job Candidates for Unilever
https://www.theregister.co.uk/2019/09/30/unilever_using_ai_and_facial_tracking_in_job_interviews/

Hirevue claims to have already carried out 100,000 interviews for companies in the UK. Most of the analysis is based on language – word choice, tone, speed of speech and sentence length are all considered. But the software also looks at facial expressions – smiling, brow raising and eye widening.

Loren Larsen, CTO at Hirevue, said: "There are 350-ish features that we look at in language: do you use passive or active words? Do you talk about 'I' or 'We.' What is the word choice or sentence length? In doctors, you might expect a good one to use more technical language.

"Then we look at the tone of voice. If someone speaks really slowly, you are probably not going to stay on the phone to buy something from them. If someone speaks at 400 words a minute, people are not going to understand them. Empathy is a piece of that."

People are then ranked on a 1-100 score system.

------------------------------------

Drones Now Dominate Israeli Flying Operations
https://breakingdefense.com/2019/09/exclusive-drones-now-dominate-israeli-flying-operations/

Israel’s Air Force, long dominated by some of the world’s best fighter pilots, has completed the shift from pilots to drones, with more than 80 percent of all Israeli Air Force flight hours executed by unmanned aircraft. ... "Last year 78 percent of the IAF's operational flight hours were performed by UAS. This year the number jumped and is 80 percent,"

Recent reports by Arab media describing Israeli attacks on Iranian targets far from Israel included details of unidentified drones. Israel has never confirmed that it performed such long-range attacks.

... What was missing in the view around us when we walked on the tarmac was the presence of the “outside pilot,” using controllers that are required for many other drones. This job has been eliminated because the Heron-TP uses an automatic takeoff and landing system, which also functions in inclement weather. To begin a mission, the commander simply pushes a button. The large drone taxis out of its hangar, continues to the runway and takes off. The same sequence is used once the mission is complete.
“There are three classes of people: those who see. Those who see when they are shown. Those who do not see.” ― Leonardo da Vinci

Insensible before the wave so soon released by callous fate. Affected most, they understand the least, and understanding, when it comes, invariably arrives too late

vox_mundi

  • Nilas ice
  • Posts: 1629
    • View Profile
  • Liked: 639
  • Likes Given: 113
Re: Robots and AI: Our Immortality or Extinction
« Reply #106 on: October 03, 2019, 07:03:40 PM »
'Goliath Is Winning': The Biggest U.S. Banks Are Set to Automate Away 200,000 Jobs
https://gizmodo.com/goliath-is-winning-the-biggest-u-s-banks-are-set-to-a-1838740347

Over the next decade, U.S. banks, which are investing $150 billion in technology annually, will use automation to eliminate 200,000 jobs, thus facilitating “the greatest transfer from labor to capital” in the industry’s history.

The job cuts are slated to hit front offices, call centers, and branches the hardest, where 20-30 percent of those roles will be on the chopping block. They will be replaced by better ATMs, automated chatbots, and software instruments that take advantage of big data and cloud computing to make investment decisions.

The report will contribute to an automated arms race between companies to cut staff and purchase enterprise financial software products that is already underway. This is how a lot of corporate automation unfolds.

In the current formulation, Goliath is winning, and 200,000 people are losing. ... Goliath is winning by replacing its human employees with instruments that direct capital more swiftly from us to it.

“There are three classes of people: those who see. Those who see when they are shown. Those who do not see.” ― Leonardo da Vinci

Insensible before the wave so soon released by callous fate. Affected most, they understand the least, and understanding, when it comes, invariably arrives too late

vox_mundi

  • Nilas ice
  • Posts: 1629
    • View Profile
  • Liked: 639
  • Likes Given: 113
Re: Robots and AI: Our Immortality or Extinction
« Reply #107 on: October 15, 2019, 11:53:19 PM »
Textron Unveils Revolutionary Unmanned Robot Light Tank
https://www.thedrive.com/the-war-zone/30428/the-movie-star-ripsaw-mini-tank-has-reemerged-unmanned-and-packing-a-big-30mm-cannon



Textron subsidiary, Howe Technologies has unveiled the latest variant of their Ripsaw, a line of tracked vehicles that look ripped straight from a G.I. Joe playset – 2013's G.I. Joe: Retaliation even featured an earlier model. The company hopes that the M5 model could meet the U.S. Army's requirements for future unmanned ground vehicles, including semi-autonomous "wingmen" to support its future, larger optionally-manned combat vehicles.

The vehicle is the firm's entry for what the Army is presently calling the Robotic Combat Vehicle-Medium (RCV-M) program, which envisions a vehicle between 10 and 20 tons.



Howe and Howe's miniature "tank" has the same 30mm XM813 Bushmaster main gun, as the Army's up-gunned 8-wheeled M1296 Stryker Dragoon, making it a very well-armed vehicle for its weight.

Howe and Howe even demonstrated a variant with M5 Modular Crowd Control Munitions (MCCM), an obscure less-than-lethal cousin of the M18A1 Claymore directional anti-personnel mine that has a smaller explosive charge and rubber pellets instead of steel balls, mounted on the sides for riot control missions.

--------------------------

Wasn't M-5 the Computer that Killed Humans in Star Trek
https://en.m.wikipedia.org/wiki/The_Ultimate_Computer

... The Federation starship Enterprise is summoned to a space station without explanation. Commodore Wesley (Barry Russo) explains that the Enterprise will be a test vessel for the M-5 Multitronic System, a revolutionary tactical and control computer designed by Dr. Richard Daystrom (William Marshall). The M-5 is to handle all ship functions without human assistance.

The M-5 succeeds at its first tasks, performing ship functions more quickly and efficiently than a living crew. However, M-5 also exhibits unexpected behavior, such as turning off power and life support to unoccupied parts of the ship, and drawing increased power for unknown reasons; Daystrom maintains M-5 is working properly.

In its first tactical drill, M-5 defends the Enterprise against mock attacks from Starships Excalibur and Lexington. The Enterprise is declared the victor, and Wesley jokingly refers to Kirk as "Captain Dunsel", employing a Starfleet Academy slang term for a part serving no useful purpose. Kirk is troubled by this.

Some time later, M-5 detects the Woden, an unmanned freighter, and attacks with live torpedoes, destroying it. Kirk orders M-5 taken offline, but on attempting to do so, finds it protected by a powerful force field. M-5 has made communication with the fleet impossible. Chief Engineer Scott orders Ensign Harper to disconnect its power source, but the M-5 creates a direct connection to the ship's warp engines, vaporizing Harper in the process. Spock and Scott attempt a manual override, but discover that the M-5 has rerouted all controls. Spock questions Daystrom on his computer design, and Daystrom reveals that he has imprinted human engrams onto M-5's circuits, creating what amounts to a human mind operating at the speed of a computer.


Poor Ensign Harper

... Kirk then tries to persuade the M-5 to stop its attacks. The M-5 acknowledges Kirk, who asks M-5 what its purpose is. M-5 responds that its purpose is to protect lives. Kirk rejoins that it acted contrary to its purpose by murdering people. M-5 acknowledges that it has committed murder and must therefore die, and shuts itself down. In so doing, it also cripples the Enterprise. ... (... today, it would double down and shutoff life support like HAL9000)
« Last Edit: October 16, 2019, 02:43:40 AM by vox_mundi »
“There are three classes of people: those who see. Those who see when they are shown. Those who do not see.” ― Leonardo da Vinci

Insensible before the wave so soon released by callous fate. Affected most, they understand the least, and understanding, when it comes, invariably arrives too late

vox_mundi

  • Nilas ice
  • Posts: 1629
    • View Profile
  • Liked: 639
  • Likes Given: 113
Re: Robots and AI: Our Immortality or Extinction
« Reply #108 on: October 16, 2019, 02:20:11 AM »
OpenAI Teaches Robot Hand to Solve Rubik's Cube
https://spectrum.ieee.org/automaton/robotics/robotics-hardware/openai-demonstrates-sim2real-by-with-onehanded-rubiks-cube-solving



In a preprint paper published online today, OpenAI has managed to teach its robot hand to solve a much more difficult version of in-hand cube manipulation: single-handed solving of a 3x3 Rubik’s cube.



“While the video makes it easy to focus on the physical robot, the magic is mostly happening in simulation, and transferring things learned in simulation to the real world. The key to this is domain randomization—jittering parts of the simulation around so that your system has to adapt to different situations similar to those that might be encountered in the real-world.”

... The researchers point out that the method they’ve developed here is general purpose, and you can train a real-world robot to do pretty much any task that you can adequately simulate. You don’t need any real-world training at all, as long as your simulations are diverse enough, which is where the automatic domain randomization comes in. The long-term goal is to reduce the task specialization that’s inherent to most robots, which will help them be more useful and adaptable in real-world applications.

---------------------------

Self-Taught AI Masters Rubik’s Cube Without Human Help
https://gizmodo.com/self-taught-ai-masters-rubik-s-cube-without-human-help-1836420294

New research published this week in Nature Machine Intelligence describes DeepCubeA, a system capable of solving any jumbled Rubik’s Cube it’s presented with. More impressively, it can find the most efficient path to success—that is, the solution requiring the fewest number of moves—around 60 percent of the time. On average, DeepCubeA needed just 28 moves to solve the puzzle, requiring 1.2 seconds to calculate the solution.

DeepCubeA, on the other hand, taught itself to solve Rubik’s Cube using an approach to artificial intelligence known as reinforcement learning.

... “The solution to the Rubik’s Cube involves more symbolic, mathematical and abstract thinking, so a deep learning machine that can crack such a puzzle is getting closer to becoming a system that can think, reason, plan and make decisions.”

------------------------

A Self-Solving Rubik’s Cube That Floats In the Air is Completely Hands-Free
https://gizmodo.com/a-self-solving-rubik-s-cube-that-floats-in-the-air-is-c-1838143319

“There are three classes of people: those who see. Those who see when they are shown. Those who do not see.” ― Leonardo da Vinci

Insensible before the wave so soon released by callous fate. Affected most, they understand the least, and understanding, when it comes, invariably arrives too late

vox_mundi

  • Nilas ice
  • Posts: 1629
    • View Profile
  • Liked: 639
  • Likes Given: 113
Re: Robots and AI: Our Immortality or Extinction
« Reply #109 on: October 16, 2019, 02:42:13 AM »
Agility Robotics Unveils Upgraded Digit Walking Robot
https://spectrum.ieee.org/automaton/robotics/humanoids/agility-robotics-digit-v2-biped-robot



... This is v2 hardware, so there’s one more full version in development prior to the 2020 launch, which will expand the autonomy envelope significantly.

... It will be a while before Digit (or any other humanoid robot) is operating fully autonomously in crowds of people, but there are so many large market opportunities (think indoor factory/warehouse environments) to address prior to that point that we expect to mature the operational safety side of things well in advance of having saturated the more robot-tolerant markets.
“There are three classes of people: those who see. Those who see when they are shown. Those who do not see.” ― Leonardo da Vinci

Insensible before the wave so soon released by callous fate. Affected most, they understand the least, and understanding, when it comes, invariably arrives too late

vox_mundi

  • Nilas ice
  • Posts: 1629
    • View Profile
  • Liked: 639
  • Likes Given: 113
Re: Robots and AI: Our Immortality or Extinction
« Reply #110 on: October 16, 2019, 03:31:34 PM »
Assembler Robots Assembles Large Structures from Cellular Units
https://techxplore.com/news/2019-10-robots-large-pieces.html


Prototype versions of these robots can assemble small structures and even work together as a team to build up a larger assemblies

... According to Gershenfeld, the key difference with this system lies in the relationship between the robotic device and the materials that it is handling and manipulating. With these new kinds of robots, "you can't separate the robot from the structure—they work together as a system," he says. For example, while most mobile robots require highly precise navigation systems to keep track of their position, the new assembler robots only need to keep track of where they are in relation to the small subunits, called voxels, that they are currently working on. Every time the robot takes a step onto the next voxel, it readjusts its sense of position, always in relation to the specific components that it is standing on at the moment.

The underlying vision is that just as the most complex of images can be reproduced by using an array of pixels on a screen, virtually any physical object can be recreated as an array of smaller three-dimensional pieces, or voxels, which can themselves be made up of simple struts and nodes. The team has shown that these simple components can be arranged to distribute loads efficiently; they are largely made up of open space so that the overall weight of the structure is minimized. The units can be picked up and placed in position next to one another by the simple assemblers, and then fastened together using latching systems built into each voxel.

... For practical assembly applications, swarms of such units could be working together to speed up the process, thanks to control software developed by Abdel-Rahman that can allow the robots to coordinate their work and avoid getting in each other's way.



... "For a space station or a lunar habitat, these robots would live on the structure, continuously maintaining and repairing it," says Jenett.

Benjamin Jenett et al. Material–Robot System for Assembly of Discrete Cellular Structures, IEEE Robotics and Automation Letters (2019)
https://ieeexplore.ieee.org/document/8769886
“There are three classes of people: those who see. Those who see when they are shown. Those who do not see.” ― Leonardo da Vinci

Insensible before the wave so soon released by callous fate. Affected most, they understand the least, and understanding, when it comes, invariably arrives too late

Sigmetnow

  • Multi-year ice
  • Posts: 15650
    • View Profile
  • Liked: 452
  • Likes Given: 215
Re: Robots and AI: Our Immortality or Extinction
« Reply #111 on: October 16, 2019, 03:59:44 PM »
... "For a space station or a lunar habitat, these robots would live on the structure, continuously maintaining and repairing it," says Jenett.

That was my first thought.  ;) :)  Autonomous constructor-bots are planned for the next-generation space habitats, such as Von Braun Station.  ( https://gatewayspaceport.com )

But such robots could also be included on uncrewed missions to the moon and Mars, where they could set up a field of solar panels (and on Mars, help keep them clean) and begin constuction of habitats.  Monitoring their progress from earth would assist in assuring sustainability of future crewed missions.
People who say it cannot be done should not interrupt those who are doing it.

Sigmetnow

  • Multi-year ice
  • Posts: 15650
    • View Profile
  • Liked: 452
  • Likes Given: 215
Re: Robots and AI: Our Immortality or Extinction
« Reply #112 on: October 30, 2019, 07:44:19 PM »
Rather than arguing whether aircraft should exist, I post this here as a look at advances in automation.  Autoland capability in airliners is trickling down to light aircraft.  And passengers are the ones to activate it.

Quote
Jason Rabinowitz (@AirlineFlyer) 10/30/19, 7:59 AM
Whoa. @Cirrus_Aircraft announced a feature on its G2 Vision Jet called Safe Return. It's literally a button that automagically lands the airplane when pressed by a passenger. It uses GPS, radar, weather data etc to pick an airport and runway, communicates to ATC and lands itself. [Photos at the link]
https://twitter.com/airlineflyer/status/1189512136650825730
[h/t to Jason for the photo and info below from his Twitter thread:]

- "Safe Return utilizes all available aircraft data streams, including available fuel, to safely navigate the Vision Jet to the identified airport and transition on to final approach for the appropriate runway based on ground winds."

Here's a video of how the Cirrus Aircraft Safe Return feature works.
Vision Jet: Safe Return™ Emergency Autoland - YouTube
https://m.youtube.com/watch?t=86&v=PiGkzgfR_c0   Click to view (it skips the Cirrus intro).

Piper Aircraft announces a similar feature today. They both appear to be enabled by the @Garmin avionics suite.
Piper Announces New M600 SLS. First GA Aircraft to be Standard Equipped with HALO™ Safety System and Autoland Capability. Available Q4 2019.
https://www.piper.com/press-releases/piper-announces-new-m600-sls/
Article contains another video.

Quote
Jon Ostrower (@jonostrower) 10/30/19, 9:13 AM
Earlier this summer, I was given a live aerial demonstration of Garmin’s new (nearly certified) Emergency Autoland system. What I saw has the power to change aviation forever. One button to land. More to come on @theaircurrent.
https://twitter.com/jonostrower/status/1189530858585116672 
At the link:  30-sec video, in flight, showing screens; brief audio of Autoland activation announcement, over engine noise

- Here's a detailed video from @AINonline on the @Garmin Autoland system demonstrated on the @piperaircraft M600.
Garmin’s Autoland Can Land an Airplane with an Incapacitated Pilot – AIN
« Last Edit: October 30, 2019, 07:52:12 PM by Sigmetnow »
People who say it cannot be done should not interrupt those who are doing it.

vox_mundi

  • Nilas ice
  • Posts: 1629
    • View Profile
  • Liked: 639
  • Likes Given: 113
Re: Robots and AI: Our Immortality or Extinction
« Reply #113 on: October 30, 2019, 08:21:43 PM »
Leveling up: DeepMind’s AlphaStar AI  achieves Grandmaster level in StarCraft II
https://arstechnica.com/science/2019/10/leveling-up-deepminds-alphastar-achieves-grandmaster-level-in-starcraft-ii/

Back in January, Google's DeepMind team announced that its AI, dubbed AlphaStar, had beaten two top human professional players at StarCraft. But as we argued at the time, it wasn't quite a fair fight. Now AlphaStar has improved on its performance sufficiently to achieve Grandmaster status in StarCraft II, using the same interface as a human player. The AI system mastered the game after 44 days of training. The team described its work in a new paper in Nature.

http://www.natureasia.com/en/research/highlight/13130

The DeepMind team restricted AlphaStar’s capabilities, ensuring for example that it could not perform moves at superhuman speeds.

... As chess grandmaster Gary Kasparov wrote in an editorial in Science last year, "Instead of processing human instructions and knowledge at tremendous speed, as all previous chess machines, AlphaZero generates its own knowledge."



... StarCraft II is a game of incomplete information, and there is no single best strategy, much like playing rock-paper-scissors. It requires long-range planning ability and real-time decision-making in a large action space. Not only is the gameplay map hidden to players, but they must also control hundreds of units (mobile game pieces that can be built to influence the game) and buildings (used to create units or technologies that strengthen those units) simultaneously.

StarCraft is particularly challenging for an AI because players must carry out long-term plans over several minutes of gameplay, tweaking them on the fly in the face of enemy counterattacks.

“AlphaStar has become the first AI system to reach the top tier of human performance in any professionally played e-sport on the full unrestricted game under professionally approved conditions,” said David Silver, a researcher at DeepMind.

... "The key insight of the league is that playing to win is insufficient," Vinyals and Czarnecki wrote of the improvements to this latest incarnation of AlphaStar. "Instead, we need both main agents whose goal is to win versus everyone, and also exploiter agents that 'take one for the team.' focusing on helping the main agent grow stronger by exposing its flaws, rather than maximizing their own win rate. Using this training method, the current league learns all its complex StarCraft II strategy in an end-to-end fashion—as opposed to the earlier incarnation of our work, which stitched together agents produced by a variety of methods and algorithms."

.... While DeepMind states that it will never be involved in military work, and Starcraft II is not a realistic war simulation, the results will be of interest to the military, said Noel Sharkey, emeritus professor of AI and robotics at the University of Sheffield. In March, a US government report described how AI enriched battlefield simulations and allowed wargamers to assess the potential outcomes of different tactics.

“Military analysts will certainly be eyeing the successful AlphaStar real-time strategies as a clear example of the advantages of AI for battlefield planning.
“There are three classes of people: those who see. Those who see when they are shown. Those who do not see.” ― Leonardo da Vinci

Insensible before the wave so soon released by callous fate. Affected most, they understand the least, and understanding, when it comes, invariably arrives too late

vox_mundi

  • Nilas ice
  • Posts: 1629
    • View Profile
  • Liked: 639
  • Likes Given: 113
Re: Robots and AI: Our Immortality or Extinction
« Reply #114 on: November 05, 2019, 12:21:58 AM »
The DOE's Supercomputers Are Getting Smarter
https://www.popularmechanics.com/technology/design/a29629616/department-energy-artificial-intelligence/



- The U.S. Department of Energy is planning to ask Congress for $3 to 4 billion to turn its existing network of supercomputers into high-performance AI machines, that it's referring to as a next-generation "exascale" machine.

- In part, the DOE's move is a bid to compete with countries like China that are heavily investing in artificial intelligence at the governmental level.

- The Department of Defense is seeking a similar amount of funding for military AI programs, much of which will take place at Carnegie Mellon University.

- Having a network of AI machines could improve the rate at which scientific discoveries are made, keeping the U.S. competitive in innovation.

- ... or exterminate humanity.



-----------------------

MIT-Air Force AI Accelerator Takes Off: Magnetic Field GPS & A DoD Smart Speaker
https://breakingdefense.com/2019/11/mit-air-force-ai-accelerator-takes-off/

WASHINGTON: The Air Force and MIT are kicking off 10 joint research projects aimed at finding Artificial Intelligence (AI) solutions fast for some of the service’s key operational problems, such as building a simulated co-pilot (dubbed R2D2) and navigating without GPS.

The Air Force signed a contract in May with MIT to create the MIT-Air Force AI Accelerator, a research partnership on fundamental AI research such as computational intelligence, reasoning and decision-making.

Other projects include: 

- Using AI to predict weather in regions where the Air Force has no current data

- Applying AI to personnel recovery missions using swarming technology;

- Using AI to develop algorithms to reduce interference in communications;

- Using AI for to improve operational scheduling (he noted that this is “one of the major pain points for airmen throughout the Air Force”);

- AI for imagery fusion, especially synthetic aperture radar (SAR) imagery; and

- Figuring out how to run AI on legacy software systems “rather than having to dump all that IT infrastructure out.”

Finally, the MIT-Air Force team is looking at how AI can help build “an Alexa that can understand our military jargon,” Brown said, adding with a laugh that “there are a million use cases out there for that.”

--------------------

Artificial Intelligence Will Detect Hidden Targets In 2020 Wargame
https://breakingdefense.com/2019/10/artificial-intelligence-will-detect-hidden-targets-in-2020-wargame/

The Army has developed AI to spot hidden targets in reconnaissance photos and will field-test it in next year’s massive Defender 20 wargames in Europe, the head of the service’s said.

Shooting down drones, aiming tank guns, coordinating resupply and maintenance, planning artillery barrages, stitching different sensor feeds together into a single coherent picture, analyzing how terrain blocks units’ fields of fire and warning commanders where there are blind spots in their defenses are all uses that will be tested.

Gen. Easley said, the new application looks for tanks and other targets of interest in a major-power war, he said, in keeping with the Pentagon’s increasing focus on Russia and China.

For example, Easley elaborated to reporters at a follow-on roundtable, you would feed the AI surveillance imagery of “a forested area” and ask it, “show me every tank you see in the image.” The AI would rapidly highlight the targets – far faster than a human imagery analyst could go through the same volume of raw intelligence – and pass them along to humans to take action.
« Last Edit: November 05, 2019, 12:54:30 AM by vox_mundi »
“There are three classes of people: those who see. Those who see when they are shown. Those who do not see.” ― Leonardo da Vinci

Insensible before the wave so soon released by callous fate. Affected most, they understand the least, and understanding, when it comes, invariably arrives too late

nanning

  • Grease ice
  • Posts: 879
  • 0Kg CO2, 35 KWh/wk,130L H2O/wk, No heating
    • View Profile
  • Liked: 135
  • Likes Given: 5783
Re: Robots and AI: Our Immortality or Extinction
« Reply #115 on: November 05, 2019, 11:28:22 AM »
Unintended consequences with pure AI

POTUS:
   "Alexa, start pre-emptive strike on enemy"
Alexa:
   "White House target locked in. Missiles launched."
   "... peace \/"
"It is preoccupation with possessions, more than anything else, that prevents us from living freely and nobly" - Bertrand Russell
   Simple: minimize your possessions and be free and kind    It's just a mindset.       Refugees welcome

gerontocrat

  • First-year ice
  • Posts: 6551
    • View Profile
  • Liked: 1504
  • Likes Given: 21
Re: Robots and AI: Our Immortality or Extinction
« Reply #116 on: November 05, 2019, 12:21:43 PM »
- The Department of Defense is seeking a similar amount of funding for military AI programs, much of which will take place at Carnegie Mellon University.[/b]

- Having a network of AI machines could improve the rate at which scientific discoveries are made, keeping the U.S. competitive in innovation.

- ... or exterminate humanity.


Artificial Intelligence Will Detect Hidden Targets In 2020 Wargame
https://breakingdefense.com/2019/10/artificial-intelligence-will-detect-hidden-targets-in-2020-wargame/

The Army has developed AI to spot hidden targets in reconnaissance photos and will field-test it in next year’s massive Defender 20 wargames in Europe, the head of the service’s said.

Shooting down drones, aiming tank guns, coordinating resupply and maintenance, planning artillery barrages, stitching different sensor feeds together into a single coherent picture, analyzing how terrain blocks units’ fields of fire and warning commanders where there are blind spots in their defenses are all uses that will be tested.

Gen. Easley said, the new application looks for tanks and other targets of interest in a major-power war, he said, in keeping with the Pentagon’s increasing focus on Russia and China.

For example, Easley elaborated to reporters at a follow-on roundtable, you would feed the AI surveillance imagery of “a forested area” and ask it, “show me every tank you see in the image.” The AI would rapidly highlight the targets – far faster than a human imagery analyst could go through the same volume of raw intelligence – and pass them along to humans to take action.

"and pass them along to humans to take action."

And how many generals and other assorted lunatics are telling Trump "But much more efficient for the AI machine to take the action itself?

I don't think Isaac Asimov's 3 laws of robotics have a chance.

"Para a Causa do Povo a Luta Continua!"
"And that's all I'm going to say about that". Forrest Gump
"Damn, I wanted to see what happened next" (Epitaph)

Archimid

  • Nilas ice
  • Posts: 2360
    • View Profile
  • Liked: 284
  • Likes Given: 184
Re: Robots and AI: Our Immortality or Extinction
« Reply #117 on: November 05, 2019, 12:36:18 PM »
Quote
I don't think Isaac Asimov's 3 laws of robotics have a chance.

No, they don't. They are meant to protect humans from AI, not from humans controlling AIs.  AIs are more dangerous as tools of the powerful than as self-aware algorithms.
I am an energy reservoir seemingly intent on lowering entropy for self preservation.

Sigmetnow

  • Multi-year ice
  • Posts: 15650
    • View Profile
  • Liked: 452
  • Likes Given: 215
Re: Robots and AI: Our Immortality or Extinction
« Reply #118 on: November 05, 2019, 07:12:42 PM »
FRONTLINE investigates how artificial intelligence is disrupting life as we know it — for good and ill. Tune in or stream “In The Age of AI”
SEASON 2019: EPISODE 5
FRONTLINE investigates the promise and perils of artificial intelligence, from fears about work and privacy to rivalry between the U.S. and China. The documentary traces a new industrial revolution that will reshape and disrupt our lives, our jobs and our world, and allow the emergence of the surveillance society.

Episode Trailer: https://www.pbs.org/wgbh/frontline/film/in-the-age-of-ai/
People who say it cannot be done should not interrupt those who are doing it.

vox_mundi

  • Nilas ice
  • Posts: 1629
    • View Profile
  • Liked: 639
  • Likes Given: 113
Re: Robots and AI: Our Immortality or Extinction
« Reply #119 on: November 06, 2019, 07:35:16 PM »
DoD Should Consider Truly Autonomous Weapons: Bipartisan AI Commission Reports
https://breakingdefense.com/2019/11/bipartisan-ai-commission-dod-should-consider-truly-autonomous-weapons/

WASHINGTON: The US military should adopt artificial intelligence urgently without letting debates over ethics and human control “paralyze AI development,” a congressionally mandated panel says. “In light of the choices being made by our strategic competitors, the United States must also examine AI through a military lens, including concepts for AI-enabled autonomous operations.”

The interim report released yesterday by the bipartisan National Security Commission on Artificial Intelligence is full-throated in its defense of the pursuit of autonomous, AI-driven military systems as not only ethical but essential for future US military operations. Even in the military, some commanders have been publicly reluctant to trust AI — especially for anything related to nuclear weapons.

... It also posits the benefits of AI for homeland defense, the Intelligence Community (IC) and the military:

  • For homeland defense, the report says, AI-enable tools can assist with border protections, cybersecurity, protection of critical infrastructure and natural disaster response.
  • For the Intelligence Community, “AI algorithms can sift through vast amounts of data to find patterns, detect threats, and identify correlations. AI tools can make satellite imagery, communications signals, economic indicators, social media data, and other large sources of information more intelligible. AI-enabled analysis can provide faster and more precise situational awareness that supports higher quality decision-making.”
  • On future battlefields, the military “could use AI-enabled machines, systems, and weapons to understand the battlespace more quickly; develop a common joint operating picture more rapidly; make relevant decisions faster; mount more complex multi-domain operations in contested environments; put fewer U.S. service members at risk; and protect innocent lives and reduce collateral damage.”

Of course, if those applications, however appealing, require relinquishing or even reducing human control, the controversy will be intense.

Notably, nowhere does commission use the phrase ‘human in the loop,’ the language currently favored by the Pentagon to assert that a human would always have ultimate control over any autonomous system.



--------------------------------------

SecDef: China Is Exporting Killer Robots to the Mideast
https://www.defenseone.com/technology/2019/11/secdef-china-exporting-killer-robots-mideast/161100/
https://breakingdefense.com/2019/11/china-seeks-ai-without-limits-ethics-secdef-esper/

China is exporting drones that it advertises as having lethal autonomy to the Middle East, Defense Secretary Mark Esper said Tuesday. It’s the first time that a senior Defense official has acknowledged that China is selling drones capable of taking life with little or no human oversight.

“As we speak, the Chinese government is already exporting some of its most advanced military aerial drones to the Middle East, as it prepares to export its next-generation stealth UAVs when those come oneline,” Esper said today at the National Security Commission on Artificial Intelligence conference. “In addition, Chinese weapons manufacturers are selling drones advertised as capable of full autonomy, including the ability to conduct lethal targeted strikes.”

The Chinese company Ziyan, for instance, markets the Blowfish A3, essentially a helicopter drone outfitted with a machine gun. Ziyan says it “autonomously performs more complex combat missions, including fixed-point timing detection, fixed-range reconnaissance, and targeted precision strikes.”



Last year, Zeng Yi, a senior executive at NORINCO, China’s third-largest defense company, forecast that, ... In future battlegrounds, there will be no people fighting” —as early as 2025.

Quote
... it’s much easier to create an indiscriminate weapon than one that avoids collateral damage. An AI that can detect, for example, a roughly human-sized object with a roughly human-like body temperature is much easier to program than one that can tell a civilian from an armed combatant. And an unmanned system that opens fire on its own, without seeking approval from a human overseer, doesn’t need the kind of secure, jam-proof, long-range communications networks that are required for human oversight.

Esper also said Chinese surveillance software and hardware networks could help China develop AI. “All signs point to the construction of a 21st-century surveillance state designed to censor speech and deny basic human rights on an unprecedented scale. Look no further than its use of surveillance to systematically repress more than a million Muslim Uighurs,” he said. “Beijing has all the power and tools it needs to coerce Chinese industry and academia into supporting its government-led efforts.”

He said it was “equally troubling are the outside firms or multinational corporations that are inadvertently or tacitly providing the technology or research behind China’s unethical use of AI.

“Let me be clear: The question is not whether AI will be used by militaries around the world – it will be,” Esper concluded. ...

------------------------------------

Navy, Marines Moving Ahead with Unmanned Vessel Programs
https://news.usni.org/2019/10/31/navy-marines-moving-ahead-with-unmanned-vessel-programs

The Navy is gaining enough experience with unmanned vehicles on and below the water’s surface that it’s becoming easier to kick off new programs, as each can build on previous program’s lessons learned, service officials said last week.

On larger unmanned surface vessels, the Navy and Pentagon’s Ghost Fleet Overlord program transitioned from Phase 1 to Phase 2 at the beginning of October, further building upon the base of knowledge that will inform two future programs of record, the Large Unmanned Surface Vehicle (LUSV) and Medium USV (MUSV).

“We’ve completed over 600 hours of autonomous testing, we’ve demonstrated autonomy, we’ve demonstrated navigation, and we’ve launched into the Phase 2 with our two vendors now,” Berkof said last week at the National Defense Industrial Association’s annual Expeditionary Warfare Conference.

After his speech, he told USNI News that “Phase 2, we integrate a government supplied
[command, control, communications, computers and intelligence] system into the vessel, we do more complex autonomy, more complex navigation, and we start also some payload work. And then so we get through Phase 2, and then we will have a number of demonstrations out there where we really ramp up the capability on that front.”

Once the integration of a government supplied C4I system onto a hull rigged to operate autonomously is complete, the only real remaining work will be to add in a vertical launching system to give the USV a strike capability.

For the Navy’s LUSV program, we envision integrating vertical launch into that vessel. … That’s really the biggest difference” between Ghost Fleet Overlord Phase 2 and LUSV, Berkof said.

...  And if you’re the adversary, well I guess you’re going to need to target everything,” the general said.

So I hope you bought enough DF-XXs [Chinese anti-ship weapons] because I’m going to try to spread you out. That I don’t mind saying publicly. So long-range unmanned surface vessels, for us, vitally important because they’re lethal. They’re not just connectors; they’re sniffers, they’re out there telling me what’s going on, they’re passing that information back to me, and they’re spreading out the enemy because at some point you’ve got to target everything that moves because the one thing that does get through is carrying the lethal package



---------------------------------

Brain-Like Computer Chips Developed
https://techxplore.com/news/2019-11-brain-like-chips-privacy-greenhouse-emissions.html

A team lead by Professor Simon Brown at the University of Canterbury (UC) has developed computer chips with brain-like functionality, that could significantly reduce global carbon emissions from computer energy consumption.

Published this week in prestigious peer-reviewed journal Science Advances, the paper proves signals on the chips are remarkably like those that pass through the network of neurons in the brain. This is important for building new kinds of computers because the brain is incredibly good at processing information using very small amounts of energy. Brain-like computing could enable "edge computing" and address the ever increasing energy consumption of computers.

The chips are based on self-organisation of nanoparticles – taking advantage of physical principles at unimaginably small scales, a hundred thousand times smaller than the thickness of a human hair, to make brain-like networks.

The components of this new chip are at the atomic level and are so small they cannot be seen with the naked eye or conventional microscopes, and can only be seen in electron microscopes.

"The research shows that this type of chip really does mimic the signalling behaviour of the brain. We were surprised at the extent to which the avalanches or cascades of voltage pulses on our chips replicate the avalanches of 'action potentials' that are observed in the brain.

"These chips might provide a different kind of artificial intelligence. By understanding the underlying fundamental physical processes, we believe we can design these chips and control their behaviour to do things like pattern or image recognition," he says. "The key is that processing on-chip and with low power consumption opens up new applications that are not currently possible."

J. B. Mallinson et al. Avalanches and criticality in self-organized nanoscale networks, Science Advances (2019
“There are three classes of people: those who see. Those who see when they are shown. Those who do not see.” ― Leonardo da Vinci

Insensible before the wave so soon released by callous fate. Affected most, they understand the least, and understanding, when it comes, invariably arrives too late

vox_mundi

  • Nilas ice
  • Posts: 1629
    • View Profile
  • Liked: 639
  • Likes Given: 113
Re: Robots and AI: Our Immortality or Extinction
« Reply #120 on: November 07, 2019, 05:42:04 PM »
Abusing a Robot Won't Hurt It, But It Could Make You a Crueler Person
https://techxplore.com/news/2019-11-abusing-robot-wont-crueller-person.html

Set in a dystopian 2019, the sci-fi classic Blade Runner explores how artificial humans could impact our humanity. Harrison Ford's character experiences powerful emotional and moral effects as he goes about hunting "replicants".

Now, in the real 2019, the influence of robots on human behaviour is increasingly relevant. Killer military robots and sex robots, for example, might alter attitudes to killing and to women, respectively.

In our research, we explored the potential link between social robots and human character.

Could treating social robots kindly make us kinder people? And could cruelty towards them make us more callous?

... If social robots could shape our characters in significant ways, it may be young children who are most affected, as childrens' characters are especially impressionable.

Just as we have animal anti-cruelty laws, some say we'll soon need robot anti-cruelty laws.

Simon Coghlan et al. Could Social Robots Make Us Kinder or Crueler to Humans and Animals?, International Journal of Social Robotics (2019)



--------------------------------

Showing Robots 'Tough Love' Helps Them Succeed
https://techxplore.com/news/2019-11-robots-tough.html

In a computer-simulated manipulation task, the researchers found that training a robot with a human adversary significantly improved its grasp of objects.

"Picture it like playing a sport: if you're playing tennis with someone who always lets you win, you won't get better. Same with robots. If we want them to learn a manipulation task, such as grasping, so they can help people, we need to challenge them."

... "I've always worked on human-robot collaboration, but in reality, people won't always be collaborators with robots in the wild," said Nikolaidis.

As an example, he points to a study by Japanese researchers, who set a robot loose in a public shopping complex and observed children "acting violently" towards it on several occasions.

So, thought Nikolaidis, what if we leveraged our human inclination to make things harder for the robot instead? Rather than showing it how to better grasp an object, what if we tried to pull it away? By adding challenge, the thinking goes, the system would learn to be more robust to real world complexity.

The researchers found the system trained with the human adversary rejected unstable grasps, and quickly learned robust grasps for these objects. In an experiment, the model achieved a 52 percent grasping success rate with a human adversary versus a 26.5 percent grasping success rate with a human collaborator.

This begs the question: how far are we willing to take adversarial learning? Would we be willing to kick and beat robots into submission? The answer, said Nikolaidis, lies in finding a balance of tough love and encouragement with our robotics counterparts.

Jiali Duan, et.al. Robot Learning via Human Adversarial Games, International Conference on Intelligent Robots and Systems (2019)
“There are three classes of people: those who see. Those who see when they are shown. Those who do not see.” ― Leonardo da Vinci

Insensible before the wave so soon released by callous fate. Affected most, they understand the least, and understanding, when it comes, invariably arrives too late

TerryM

  • First-year ice
  • Posts: 5225
    • View Profile
  • Liked: 407
  • Likes Given: 5
Re: Robots and AI: Our Immortality or Extinction
« Reply #121 on: November 07, 2019, 06:05:46 PM »
^^
I wonder if there have been studies comparing those who grew up in ~egalitarian societies with those from apartheid cultures? Would the results be applicable to future cultures when androids walk amongst us?
Terry

vox_mundi

  • Nilas ice
  • Posts: 1629
    • View Profile
  • Liked: 639
  • Likes Given: 113
Re: Robots and AI: Our Immortality or Extinction
« Reply #122 on: November 07, 2019, 08:43:15 PM »
Since AI seems to be learning all our biases and faults  through the data they are trained on, the question might be: Will androids show an egalitarian tolerance towards us?



---------------------

It's a hard life
It's a hard life
It's a very hard life
It's a hard life wherever you go
If we poison our children with hatred
Then, the hard life is all they'll ever know

- Nanci Griffith


“There are three classes of people: those who see. Those who see when they are shown. Those who do not see.” ― Leonardo da Vinci

Insensible before the wave so soon released by callous fate. Affected most, they understand the least, and understanding, when it comes, invariably arrives too late

NeilT

  • Nilas ice
  • Posts: 1225
    • View Profile
  • Liked: 103
  • Likes Given: 8
Re: Robots and AI: Our Immortality or Extinction
« Reply #123 on: November 07, 2019, 08:47:47 PM »
I'd say a more pertinent question is.

"Given that Androids work on logic today, what would a world of AI look like which worked on emotion?"

Food for thought..
Being right too soon is socially unacceptable.

Robert A. Heinlein

Sigmetnow

  • Multi-year ice
  • Posts: 15650
    • View Profile
  • Liked: 452
  • Likes Given: 215
Re: Robots and AI: Our Immortality or Extinction
« Reply #124 on: November 07, 2019, 09:14:37 PM »
I'd say a more pertinent question is.

"Given that Androids work on logic today, what would a world of AI look like which worked on emotion?"

Food for thought..

And what might a world look like if it embraced pure logic?
People who say it cannot be done should not interrupt those who are doing it.

NeilT

  • Nilas ice
  • Posts: 1225
    • View Profile
  • Liked: 103
  • Likes Given: 8
Re: Robots and AI: Our Immortality or Extinction
« Reply #125 on: November 07, 2019, 09:23:00 PM »

And what might a world look like if it embraced pure logic?

I'd suggest there would be a LOT less humans..
Being right too soon is socially unacceptable.

Robert A. Heinlein

Sigmetnow

  • Multi-year ice
  • Posts: 15650
    • View Profile
  • Liked: 452
  • Likes Given: 215
Re: Robots and AI: Our Immortality or Extinction
« Reply #126 on: November 07, 2019, 10:46:51 PM »

And what might a world look like if it embraced pure logic?

I'd suggest there would be a LOT less humans..

In a logical world that took care of its existing citizenry and reduced the need for manual labor for subsistence, humans would have less need to procreate for the purpose of having children to support themselves.
People who say it cannot be done should not interrupt those who are doing it.

vox_mundi

  • Nilas ice
  • Posts: 1629
    • View Profile
  • Liked: 639
  • Likes Given: 113
Re: Robots and AI: Our Immortality or Extinction
« Reply #127 on: November 08, 2019, 12:21:52 AM »
Something to keep in mind. Generally, and especially with neuromorphic chips, AI is not logical - it's probabilistic and heuristic.

Classic rule-based software is logical - sometimes.

And, as far as AI (and humans) are concerned, I'd prefer that they were empathic rather than emotional. There's a fair amount of crazy in emotional.
“There are three classes of people: those who see. Those who see when they are shown. Those who do not see.” ― Leonardo da Vinci

Insensible before the wave so soon released by callous fate. Affected most, they understand the least, and understanding, when it comes, invariably arrives too late

NeilT

  • Nilas ice
  • Posts: 1225
    • View Profile
  • Liked: 103
  • Likes Given: 8
Re: Robots and AI: Our Immortality or Extinction
« Reply #128 on: November 08, 2019, 03:54:27 PM »
Something to keep in mind. Generally, and especially with neuromorphic chips, AI is not logical - it's probabilistic and heuristic.

But based on an initial foundation of digital logic from which it can then extend into analogue abstraction.  Unlike Humans who are analogue from day 1.  We inherit our foundation genetically from our DNA, AI inherits it from digital logic and then builds on it.

It is far more likely that an AI would go Schizophrenic than totally insane.  Humans are different and, until we design a totally analogue AI which must be born and learn the way a baby does, this will remain the same.

Taking a foray back into current AI research tells met that most of it is about getting more bang for the buck with multiple on and off states for the same power. This is quite different from a purely analogue and abstract computer.
Being right too soon is socially unacceptable.

Robert A. Heinlein

SteveMDFP

  • Nilas ice
  • Posts: 1436
    • View Profile
  • Liked: 183
  • Likes Given: 15
Re: Robots and AI: Our Immortality or Extinction
« Reply #129 on: November 08, 2019, 04:24:06 PM »


But based on an initial foundation of digital logic from which it can then extend into analogue abstraction.  Unlike Humans who are analogue from day 1.  We inherit our foundation genetically from our DNA, AI inherits it from digital logic and then builds on it.

It is far more likely that an AI would go Schizophrenic than totally insane.  Humans are different and, until we design a totally analogue AI which must be born and learn the way a baby does, this will remain the same. ...

This is largely true, but not quite.  Each of the 100 billion neurons in the brain does operate in a digital fashion.  It either fires or it doesn't.  "Firing" means a wave of depolarization moves along the cell surface, speeding down myelinated axons (in many cases) and releasing a packet of neurotransmitters to signal the next nerve/muscle/endocrine cell.

I'm not suggesting that individual neurons are a good model for building AI hardware.  I am suggesting that the idea that "human" traits (like intuition/personality/consciousness) cannot possibly built up from innumerable digital components is just wrong.  To a large degree, we are ourselves built up from innumerable digital components.

vox_mundi

  • Nilas ice
  • Posts: 1629
    • View Profile
  • Liked: 639
  • Likes Given: 113
Re: Robots and AI: Our Immortality or Extinction
« Reply #130 on: November 08, 2019, 05:11:55 PM »
Robots 'Not Evil' Says Boston Dynamics As Humanoids Go Viral
https://techxplore.com/news/2019-11-robots-evil-boston-dynamics-humanoids.html

As videos of robot-like dogs made by Boston Dynamics go viral on the internet, the humanoids' uncanny abilities have also sparked worries that they could become a threat to humans.

The first 1,000 Spot models are to be shipped to customers by the summer of 2020.

... We have a business plan for going into the black, but it's a few years out. So today it's Spot. We're already shipping it.

"The next thing is a logistics robot. We're working on one called Handle, which is designed to work in warehouses, moving boxes around. It looks like a bird. I think that that is going to be a bigger scale activity, but much narrower focus than Spot.



"And then the future is things like Atlas. I don't think Atlas will ever be shipped. But what we learned on Atlas will make its way into the other product things eventually. (... maybe a T-800)
“There are three classes of people: those who see. Those who see when they are shown. Those who do not see.” ― Leonardo da Vinci

Insensible before the wave so soon released by callous fate. Affected most, they understand the least, and understanding, when it comes, invariably arrives too late

gerontocrat

  • First-year ice
  • Posts: 6551
    • View Profile
  • Liked: 1504
  • Likes Given: 21
Re: Robots and AI: Our Immortality or Extinction
« Reply #131 on: November 08, 2019, 05:13:14 PM »
One thing is for sure - the human brain works completely differently from a  computer, because physically the brain just cannot do what a computer does.

"- the neuron fires a large electrical spike — an action potential. This jolt whizzes at speeds of up to 150 metres per second "
-Speed of electricity through a computer approaches the speed of light - 300,000 kms per second.
- 3 GHZ cpu can process 3,000 million cycles per second.

Computers can process vast amounts of data very quickly - but that is not intelligence as it is an auto-response to instructions given to it. Even when it "learns" to produce new analyses it is following instructions to do so.

So we do it different - and better. I always thought that that was the key to AI - a computer that could process stuff the way we do instead of people writing longer & longer code.

After all, a computer is not as clever as the Nile Crocodile with 80.5 million neurons in the brain+the whole nervous system.
____________________________________________________
https://www.nature.com/articles/d41586-018-06694-6
Quote
"High-speed process
The average human brain contains about 120 billion neurons, which constantly receive and send information through branch-like appendages called dendrites. Chemical or electrical signals that reach the dendrites produce small voltage changes across the cell’s membrane, which are routed to the cell body. When the sum of the voltage changes reaches a point of no return, called a threshold, the neuron fires a large electrical spike — an action potential. This jolt whizzes at speeds of up to 150 metres per second along a neuronal branch, known as an axon, to another set of branching appendages. Here, chemical or electrical signals pass the information on to the next set of dendrites.
"Para a Causa do Povo a Luta Continua!"
"And that's all I'm going to say about that". Forrest Gump
"Damn, I wanted to see what happened next" (Epitaph)

vox_mundi

  • Nilas ice
  • Posts: 1629
    • View Profile
  • Liked: 639
  • Likes Given: 113
Re: Robots and AI: Our Immortality or Extinction
« Reply #132 on: November 08, 2019, 05:34:19 PM »
Quote from: gerontocrat
... Computers can process vast amounts of data very quickly - but that is not intelligence as it is an auto-response to instructions given to it. Even when it "learns" to produce new analyses it is following instructions to do so.

That part of your statement may be true for a classical computer system. 

But, an AI system does not use an auto-response (... which would always give the same answer to a given input.) It reacts probabilisticly, meaning for a given input there are a distribution of possible answers.

And as far as 'learning' is concerned; there are no specific (if-then-else, case) instructions for an AI. It uses a sieve of 1-1.5 million algorithms to refine an input into an output. Much of what goes on in-between is a black box even for the computer scientists that are training it. And many of the algorithms are generated by the AI.
“There are three classes of people: those who see. Those who see when they are shown. Those who do not see.” ― Leonardo da Vinci

Insensible before the wave so soon released by callous fate. Affected most, they understand the least, and understanding, when it comes, invariably arrives too late

gerontocrat

  • First-year ice
  • Posts: 6551
    • View Profile
  • Liked: 1504
  • Likes Given: 21
Re: Robots and AI: Our Immortality or Extinction
« Reply #133 on: November 08, 2019, 06:03:42 PM »
Quote from: gerontocrat
... Computers can process vast amounts of data very quickly - but that is not intelligence as it is an auto-response to instructions given to it. Even when it "learns" to produce new analyses it is following instructions to do so.

That part of your statement may be true for a classical computer system. 

But, an AI system does not use an auto-response (... which would always give the same answer to a given input.) It reacts probabilisticly, meaning for a given input there are a distribution of possible answers.

And as far as 'learning' is concerned; there are no specific (if-then-else, case) instructions for an AI. It uses a sieve of 1-1.5 million algorithms to refine an input into an output. Much of what goes on in-between is a black box even for the computer scientists that are training it. And many of the algorithms are generated by the AI.
It uses a sieve of 1-1.5 million algorithms - it needs all that massive weight of crap because it can't do what the human brain does. The day it does, goodbye humanity.
"Para a Causa do Povo a Luta Continua!"
"And that's all I'm going to say about that". Forrest Gump
"Damn, I wanted to see what happened next" (Epitaph)

vox_mundi

  • Nilas ice
  • Posts: 1629
    • View Profile
  • Liked: 639
  • Likes Given: 113
Re: Robots and AI: Our Immortality or Extinction
« Reply #134 on: November 08, 2019, 07:38:11 PM »
^
That's exactly how a human brain makes sense of the world.

This is the cascade that allows human vision to occur:


https://en.m.wikipedia.org/wiki/Visual_system



Each of those boxes contains 100,000's of neurons that pass on signals to the next box; and in turn receive feedback. And it all works on 20 watts of energy.

Neuromorphic chips are using the same principle.
“There are three classes of people: those who see. Those who see when they are shown. Those who do not see.” ― Leonardo da Vinci

Insensible before the wave so soon released by callous fate. Affected most, they understand the least, and understanding, when it comes, invariably arrives too late

Sigmetnow

  • Multi-year ice
  • Posts: 15650
    • View Profile
  • Liked: 452
  • Likes Given: 215
Re: Robots and AI: Our Immortality or Extinction
« Reply #135 on: November 09, 2019, 12:36:17 AM »
Quote
Popular Science (@PopSci)n11/8/19, 11:40 AM
MIT’s nine new Mini Cheetahs frolicking in fall leaves.
Adorable or terrifying?
https://twitter.com/popsci/status/1192844465238618118
Brief video at the link.
People who say it cannot be done should not interrupt those who are doing it.

vox_mundi

  • Nilas ice
  • Posts: 1629
    • View Profile
  • Liked: 639
  • Likes Given: 113
Re: Robots and AI: Our Immortality or Extinction
« Reply #136 on: November 09, 2019, 02:11:48 AM »
Yikes!  :o

What next?

“There are three classes of people: those who see. Those who see when they are shown. Those who do not see.” ― Leonardo da Vinci

Insensible before the wave so soon released by callous fate. Affected most, they understand the least, and understanding, when it comes, invariably arrives too late

vox_mundi

  • Nilas ice
  • Posts: 1629
    • View Profile
  • Liked: 639
  • Likes Given: 113
Re: Robots and AI: Our Immortality or Extinction
« Reply #137 on: November 09, 2019, 06:10:40 PM »
AI Copernicus ‘Discovers’ that Earth Orbits the Sun
https://www.nature.com/articles/d41586-019-03332-7

Astronomers took centuries to figure it out. (... AI took hours) But now, a machine-learning algorithm inspired by the brain has worked out that it should place the Sun at the centre of the Solar System, based on how movements of the Sun and Mars appear from Earth. The feat is one the first tests of a technique that researchers hope they can use to discover new laws of physics, and perhaps to reformulate quantum mechanics, by finding patterns in large data sets. The results are due to appear in Physical Review Letters.

Physicist Renato Renner at the Swiss Federal Institute of Technology (ETH) in Zurich and his collaborators wanted to design an algorithm that could distill large data sets down into a few basic formulae, mimicking the way that physicists come up with concise equations like E = mc2. To do this, the researchers had to design a new type of neural network, a machine-learning system inspired by the structure of the brain.

One of the first tests was to give the network simulated data about the movements of Mars and the Sun in the sky, as seen from Earth. From this point of view, Mars’s orbit of the Sun appears erratic, for example it periodically goes ‘retrograde’, reversing its course. ... The team’s neural network came up with Copernicus-style formulas for Mars’s trajectory, rediscovering “one of the most important shifts of paradigms in the history of science”, says Mario Krenn, a physicist at the University of Toronto in Canada who works on applying artificial intelligence to scientific discovery.

Raban Iten, Tony Metger, Henrik Wilming, Lidia del Rio, Renato Renner, Discovering physical concepts with neural networks, Physical Review Letters 2019
“There are three classes of people: those who see. Those who see when they are shown. Those who do not see.” ― Leonardo da Vinci

Insensible before the wave so soon released by callous fate. Affected most, they understand the least, and understanding, when it comes, invariably arrives too late

vox_mundi

  • Nilas ice
  • Posts: 1629
    • View Profile
  • Liked: 639
  • Likes Given: 113
Re: Robots and AI: Our Immortality or Extinction
« Reply #138 on: November 09, 2019, 06:12:54 PM »
Russia Is Planning A “Ground Force” of Armed Military Robots
https://futurism.com/russia-ground-force-armed-military-robots

... “The evolution of combat robots is on the path of increasing the ability to perform tasks in autonomous mode with a gradual reduction in the role of the operator,” a Russian spokesperson told C4ISRNET.



Russia isn’t the only country building semi-autonomous weapons — The U.S. is developing weaponized autonomous boats and the U.K. has a program in place to build autonomous killer drones.

-----------------------------

Military Robotics Innovation – Comparing the US to Other Major Powers
https://emerj.com/ai-sector-overviews/military-robotics-innovation/


Autonomous Border Control



-----------------------------------

Army Eyes Groups of Autonomous 'Morphing' Robots for 2040
https://www.news.gatech.edu/2019/04/09/researchers-awarded-625-million-study-collective-emergent-behavior

The Army Research Laboratory, working with Northwestern University and Georgia Tech, is experimenting with groups of small robots to observe “emergent collective behavior” -- wherein otherwise disparate individual robots operate in a coordinated, synchronized fashion in ways they cannot do by themselves.

... Stanton likened the research aims to replicating natural phenomena observed within ant colonies which exhibit a certain collective synergy; individual ants participate in purposeful coordinated group functions, such as building anthills, transporting tiny pieces of food along a pheromone trail and regularly moving in an organized group manner to accomplish tasks. These group tasks, showing “emergent collective behavior,” far exceed the capabilities of any given individual ant, something which has inspired intrigue and focus among Army Research Laboratory scientists.

Although this work, exploring new frontiers in scientific research, is described as “very far out,” it raises interesting questions about the kinds of mission possibilities it might inspire. Perhaps a swarm of mini-drones blanketing an area with surveillance cameras could autonomously align themselves to complete a larger, collective task -- such as combining sensors on a single area, or forming a collective “mass” of explosives in the air to descend upon and explode together on a specific target ...

https://www.tradoc.army.mil/Portals/14/Documents/RAS_Strategy.pdf

--------------------------------

Army Testing Robot Trucks to Supply Troops, Reduce Casualties
https://www.trucks.com/2019/10/14/army-testing-robot-trucks-supply-troops-reduce-casualties/

... The Army Research Labs Coalition Assured Autonomous Resupply project, or CAAR, successfully combined a convoy of autonomous heavy trucks with smaller robotic ground vehicles that provided last-mile delivery to troops in the field.

The Army conducted the test at the U.S. National Guard’s sprawling Camp Grayling in northern Michigan. It used four Army trucks and two from the U.K. digitally tethered into a “semi-autonomous” platoon. The lead truck had a human safety driver who didn’t take control of the vehicle.

The six-truck convoy included two Humvees, formally known as High Mobility Multipurpose Wheeled Vehicles, or HMMWVs. It also had a pair of Light Medium Tactical Vehicles, each with 2.2 tons of cargo capacity. Two large 19-ton-capacity HX-60 utility trucks from the U.K. also participated.

All of the trucks were equipped with the Army’s drive-by-wire robotic kits, called the Autonomous Mobility Appliqué System.

In the future, those vehicles could include large cargo drones for airlifting material. The Army tested several before the demonstration, but weather conditions kept them grounded during the operation.

--------------------------------

The US Army is Creating Robots That Can Follow Orders
https://www.technologyreview.com/s/614686/the-us-army-is-creating-robots-that-can-follow-ordersand-ask-if-they-dont-understand/

For robots to be useful teammates, they need to be able to understand what they’re told to do—and execute it with minimal supervision.

... Young gives the example of the command “Go behind the farthest truck on the left.” As well as recognizing objects and their locations, the robot has to decipher “behind” and “left,” which depend on where the speaker is standing, facing, and pointing. Its hard-coded knowledge of the environment gives it further conceptual clues as to how to carry out its task.

The robot can also ask questions to deal with ambiguity. If it is told to “go behind the building,” it might come back with: “You mean the building on the right?”

Current AI and robotics systems are brittle and prone to misunderstanding—think Alexa or Siri,” says Etzioni. “So if we put them in the battlefield, I sure hope we don’t give them any destructive capabilities.”



... But two other big challenges remain. First, the robot is currently too slow for practical use. Second, it needs to be far more resilient. All AI systems can go wrong, but military robots have to be reliable in life-and-death situations. These challenges will be tackled in a follow-on ARL program.

... DARPA’s “Machine Common Sense” (MCS) program is combining deep learning with a knowledge-base-centered approach so a robot can learn and show something like human judgment.)

------------------------------

Humanity is Well On Its Way To a Real-Life Terminator Uprising
https://www.engadget.com/2019/11/08/terminator-uprising-dark-fate-robotics/
“There are three classes of people: those who see. Those who see when they are shown. Those who do not see.” ― Leonardo da Vinci

Insensible before the wave so soon released by callous fate. Affected most, they understand the least, and understanding, when it comes, invariably arrives too late

TerryM

  • First-year ice
  • Posts: 5225
    • View Profile
  • Liked: 407
  • Likes Given: 5
Re: Robots and AI: Our Immortality or Extinction
« Reply #139 on: November 10, 2019, 05:05:00 PM »
Asimov is spinning in his grave. :-\
Terry

vox_mundi

  • Nilas ice
  • Posts: 1629
    • View Profile
  • Liked: 639
  • Likes Given: 113
Re: Robots and AI: Our Immortality or Extinction
« Reply #140 on: November 10, 2019, 05:47:18 PM »
“There are three classes of people: those who see. Those who see when they are shown. Those who do not see.” ― Leonardo da Vinci

Insensible before the wave so soon released by callous fate. Affected most, they understand the least, and understanding, when it comes, invariably arrives too late

TerryM

  • First-year ice
  • Posts: 5225
    • View Profile
  • Liked: 407
  • Likes Given: 5
Re: Robots and AI: Our Immortality or Extinction
« Reply #141 on: November 10, 2019, 06:15:38 PM »
^^
And Killbot Hellscape has lapped the field!! :'(
Terry

vox_mundi

  • Nilas ice
  • Posts: 1629
    • View Profile
  • Liked: 639
  • Likes Given: 113
Re: Robots and AI: Our Immortality or Extinction
« Reply #142 on: November 12, 2019, 11:08:29 PM »
AI and Compute
https://openai.com/blog/ai-and-compute/
https://www.technologyreview.com/s/614700/the-computing-power-needed-to-train-ai-is-now-rising-seven-times-faster-than-ever-before/

We’re releasing an analysis showing that since 2012, the amount of computer power used in the largest AI training runs has been increasing exponentially with a 3.4-month doubling time (by comparison, Moore’s Law had a 2-year doubling period). Since 2012, this metric has grown by more than 300,000x (a 2-year doubling period would yield only a 7x increase). Improvements in compute have been a key component of AI progress, so as long as this trend continues, it’s worth preparing for the implications of systems far outside today’s capabilities.



The chart also notably does not include some of the most recent breakthroughs, including Google’s large-scale language model BERT, OpenAI’s language model GPT-2,  or DeepMind’s StarCraft II-playing model AlphaStar.

We see multiple reasons to believe that the trend in the graph could continue. Many hardware startups are developing AI-specific chips, some of which claim they will achieve a substantial increase in FLOPS/Watt (which is correlated to FLOPS/$) over the next 1–2 years. There may also be gains from simply reconfiguring hardware to do the same number of operations for less economic cost. On the parallelism side, many of the recent algorithmic innovations described above could in principle be combined multiplicatively—for example, architecture search and massively parallel SGD.

... The world’s total hardware budget is 1 trillion dollars a year, so absolute limits remain far away. Overall, given the data above, the precedent for exponential trends in computing, work on ML specific hardware, and the economic incentives at play, we think it’d be a mistake to be confident this trend won’t continue in the short term.



---------------------------

NVIDIA Touts Jetson Xavier NX as ‘World’s Smallest Supercomputer for AI’
https://www.roboticsbusinessreview.com/ai/nvidia-touts-jetson-xavier-nx-as-worlds-smallest-supercomputer-for-ai/



...NVIDIA said the energy-efficient Jetson Xavier NX module can deliver “server-class performance” of up to 21 Tera Operations per Second (TOPS) for running modern AI workloads, consuming as little as 10 watts of power. (... equal to the fastest National Lab supercomputer from 1993, using 1/10,000 the power)

-----------------------------------

New Research Suggests Robots Appear More Persuasive When Pretending to be Human
https://techxplore.com/news/2019-11-robots-persuasive-human.html

Recent technological breakthroughs in artificial intelligence have made it possible for machines, or bots, to pass as humans. A team of researchers led by Talal Rahwan, associate professor of Computer Science at NYU Abu Dhabi, conducted an experiment to study how people interact with bots whom they believe to be human, and how such interactions are affected once bots reveal their identity. The researchers found that bots are more efficient than humans at certain human-machine interactions, but only if they are allowed to hide their non-human nature.

Talal Rahwan, et.al. Behavioral Evidence for a Transparency-Efficiency Tradeoff in Human-Machine Cooperation, Nature Machine Intelligence (2019).



------------------------------------

Computers Evolve a New Path Toward Human Intelligence
https://www.quantamagazine.org/computers-evolve-a-new-path-toward-human-intelligence-20191106/

Neural networks that borrow strategies from biology are making profound leaps in their abilities. Is ignoring a goal the best way to make truly intelligent machines?

... Evolutionary algorithms have been around for a long time. Traditionally, they’ve been used to solve specific problems. In each generation, the solutions that perform best on some metric — the ability to control a two-legged robot, say — are selected and produce offspring. While these algorithms have seen some successes, they can be more computationally intensive than other approaches such as “deep learning,” which has exploded in popularity in recent years.

The steppingstone principle goes beyond traditional evolutionary approaches. Instead of optimizing for a specific goal, it embraces creative exploration of all possible solutions. By doing so, it has paid off with groundbreaking results. Earlier this year, one system based on the steppingstone principle mastered two video games that had stumped popular machine learning methods. And in a paper published last week in Nature, DeepMind — the artificial intelligence company that pioneered the use of deep learning for problems such as the game of Go — reported success in combining deep learning with the evolution of a diverse population of solutions.

The steppingstone’s potential can be seen by analogy with biological evolution. In nature, the tree of life has no overarching goal, and features used for one function might find themselves enlisted for something completely different. Feathers, for example, likely evolved for insulation and only later became handy for flight.

-------------------------------------

Is the Will To Survive the Road To the Next Level For AI?
https://www.medicaldaily.com/will-survive-next-level-ai-445616

... some scientists, like neuroscientists Kingson Man and Antonio Damasio, thinks that there might be a way to give real-life robot feelings, and that it could be the next step for A.I. The gist is simple: Simply build a robot that would have the ability to be aware of its existence, and the perils of it. From there, a modern A.I. brain could easily develop feelings and behavior that will help guide it to self-preservation and survival.

“Today’s robots lack feelings. They are not designed to represent the internal state of their operations in a way that would permit them to experience that state in a mental space,” Man and Damasio wrote in a recent paper published in Nature Machine Intelligence.

------------------------------------

AI Can Tell If You're Going to Die Soon. We Just Don't Know How It Knows.
https://www.popularmechanics.com/science/health/a29762613/ai-predict-death-health/


  • Researchers found that a black-box algorithm predicted patient death better than humans.
  • They used ECG results to sort historical patient data into groups based on who would die within a year.
  • Although the algorithm performed better, scientists don’t understand how or why it did.
The algorithm performed better than the traditional measures used by cardiologists. The study was done by researchers in Pennsylvania’s Geisinger regional healthcare group, a low-cost and not-for-profit provider.

The most surprising axis may be the number of people cardiologists believed were healthy based on normal ECG results: “The AI accurately predicted risk of death even in people deemed by cardiologists to have a normal ECG,” New Scientist reports.

----------------------------------------------

13 Mind-Blowing Things Artificial Intelligence Can Already Do Today
https://www.forbes.com/sites/bernardmarr/2019/11/11/13-mind-blowing-things-artificial-intelligence-can-already-do-today/#786d1a656502
« Last Edit: November 13, 2019, 12:18:51 AM by vox_mundi »
“There are three classes of people: those who see. Those who see when they are shown. Those who do not see.” ― Leonardo da Vinci

Insensible before the wave so soon released by callous fate. Affected most, they understand the least, and understanding, when it comes, invariably arrives too late

vox_mundi

  • Nilas ice
  • Posts: 1629
    • View Profile
  • Liked: 639
  • Likes Given: 113
Re: Robots and AI: Our Immortality or Extinction
« Reply #143 on: November 13, 2019, 12:01:13 AM »
Conservatives Most Likely to Be Duped by Scary New AI Text Generator Released By Salesforce
https://lawsuit.org/robotrump-an-ai-trump-experiment/

Text generated by Artificial Intelligence can now be used to replicate a specific speaker's style of writing. This technology represents great risk to those who are unaware of it's potential for manipulation and fake news generation.

Quote
... So, that’s why I’ve been saying all along that, yes, I’d love to win. But, boy, do these guys want me to. These guys, they don’t talk about it. They say, “Donald Trump, please, please run.” Because they’ll take away your tax cuts, because they’ll take away your regulation cuts. They’ll take them away. And, frankly, really, really bad things will happen with our country. Our country would go down very quickly. Very quickly, very, very rapidly. The Democrats want to turn back the clock, which is, essentially, what they’ve done. They’ve turned back. We’ve gone much further left than anybody thought possible.

The above paragraph was written by an artificial intelligence. We call it “RoboTrump.”

The analysis said, “While Trump’s rambling style probably makes differentiating between real and fake more difficult than it would be for a more eloquent and talented speaker, today’s new natural language generation AI models have reached a tipping point in their ability to generate fake, real-sounding text.” The AI system used bragging and repetitive phrases to stylistically mimic Trump.

The key findings from the analysis were:
  • On average, most of the 1,000 participants did “worse than chance,” getting it right just 40% of the time.
  • Trump supporters were more likely to think that AI-generated text was real.
  • Younger people were overall more likely to believe the AI-generated content was real.



In September, Salesforce released “CTRL,” a new, state-of-the-art natural language model. This model is larger (likely better), than the full-size GPT-2 model, which Elon Musk’s OpenAI said was “too dangerous to publicly release because of its potential for abuse.”

Then, a few weeks ago, the CTRL model code was updated to allow anyone to “fine-tune” the model. This enabled the ability to train the model on text written by an individual person, which the model could then mimic incredibly well when generating its own made-up text.

... Bots leveraging this tech could spam blogs, social media posts, etc. The volume of realistically generated text in Trump’s (or anyone else’s) style will enable massive disinformation campaigns of staggering volume and reach.”

----------------------------------

OpenAI Has Published the Text-Generating AI It Said Was Too Dangerous to Share
https://www.theverge.com/2019/11/7/20953040/openai-text-generation-ai-gpt-2-full-model-release-1-5b-parameters

The lab says it’s seen ‘no strong evidence of misuse - so far’

The research lab OpenAI has released the full version of a text-generating AI system that experts warned could be used for malicious purposes.

The institute originally announced the system, GPT-2, in February this year, but withheld the full version of the program out of fear it would be used to spread fake news, spam, and disinformation. Since then it’s released smaller, less complex versions of GPT-2 and studied their reception. Others also replicated the work.

It’s tricky to convey exactly how good GPT-2’s output is, but the model frequently produces eerily cogent writing that can often give the appearance of intelligence (though that’s not to say what GPT-2 is doing involves anything we’d recognize as cognition). Play around with the system long enough, though, and its limitations become clear. It particularly suffers with the challenge of long-term coherence; for example, using the names and attributes of characters consistently in a story, or sticking to a single subject in a news article.

The best way to get a feel for GPT-2’s abilities is to try it out yourself. You can access a web version at https://talktotransformer.com/ and enter your own prompts. (A “transformer” is a component of machine learning architecture used to create GPT-2 and its fellows.)
“There are three classes of people: those who see. Those who see when they are shown. Those who do not see.” ― Leonardo da Vinci

Insensible before the wave so soon released by callous fate. Affected most, they understand the least, and understanding, when it comes, invariably arrives too late

SteveMDFP

  • Nilas ice
  • Posts: 1436
    • View Profile
  • Liked: 183
  • Likes Given: 15
Re: Robots and AI: Our Immortality or Extinction
« Reply #144 on: November 13, 2019, 01:43:15 AM »

 
... Bots leveraging this tech could spam blogs, social media posts, etc. The volume of realistically generated text in Trump’s (or anyone else’s) style will enable massive disinformation campaigns of staggering volume and reach.”


Now I wonder how many ASIF posters are actually AI bots.

Archimid

  • Nilas ice
  • Posts: 2360
    • View Profile
  • Liked: 284
  • Likes Given: 184
Re: Robots and AI: Our Immortality or Extinction
« Reply #145 on: November 13, 2019, 03:06:41 AM »
"Surely, the smart thing should control the dumb thing, but actually it's the dumb thing that controls the smart thing." - E. Musk



The whole conversation is very much on topic. Lex Fridman's channel is fantastic too.
I am an energy reservoir seemingly intent on lowering entropy for self preservation.

vox_mundi

  • Nilas ice
  • Posts: 1629
    • View Profile
  • Liked: 639
  • Likes Given: 113
Re: Robots and AI: Our Immortality or Extinction
« Reply #146 on: November 14, 2019, 08:02:40 PM »
New Artificial Intelligence System Automatically Evolves To Evade Internet Censorship
https://techxplore.com/news/2019-11-artificial-intelligence-automatically-evolves-evade.html

Internet censorship by authoritarian governments prohibits free and open access to information for millions of people around the world. Attempts to evade such censorship have turned into a continually escalating race to keep up with ever-changing, increasingly sophisticated internet censorship. Censoring regimes have had the advantage in that race, because researchers must manually search for ways to circumvent censorship, a process that takes considerable time.

New work led by University of Maryland computer scientists could shift the balance of the censorship race. The researchers developed a tool called Geneva (short for Genetic Evasion), which automatically learns how to circumvent censorship. Tested in China, India and Kazakhstan, Geneva found dozens of ways to circumvent censorship by exploiting gaps in censors' logic and finding bugs that the researchers say would have been virtually impossible for humans to find manually.

The researchers will introduce Geneva during a peer-reviewed talk at the Association for Computing Machinery's 26th Conference on Computer and Communications Security in London on November 14, 2019.

"With Geneva, we are, for the first time, at a major advantage in the censorship arms race," said Dave Levin, an assistant professor of computer science at UMD and senior author of the paper. "Geneva represents the first step toward a whole new arms race in which artificial intelligence systems of censors and evaders compete with one another. Ultimately, winning this race means bringing free speech and open communication to millions of users around the world who currently don't have them."



------------------------------------------

Facebook Says AI Getting Better at Spying Unwanted Content
https://techxplore.com/news/2019-11-facebook-ai-spying-unwanted-content.html

Facebook on Wednesday said that its software is getting more skilled at spying banned content at the social network, then working with humans to quickly remove terrorist videos and more.

Facebook has been investing heavily in artificial intelligence (AI) to automatically spot banned content, often before it is seen by users, and human teams of reviewers who check whether the software was on target.

Facebook has more than 35,000 people working on safety and security, and spends billions of dollars annually on that mission, according to Zuckerberg.

"Our efforts are paying off," Zuckerberg said. "Systems we built for addressing these issues are more advanced."
« Last Edit: November 14, 2019, 08:35:23 PM by vox_mundi »
“There are three classes of people: those who see. Those who see when they are shown. Those who do not see.” ― Leonardo da Vinci

Insensible before the wave so soon released by callous fate. Affected most, they understand the least, and understanding, when it comes, invariably arrives too late

sidd

  • Young ice
  • Posts: 4935
    • View Profile
  • Liked: 340
  • Likes Given: 0
Re: Robots and AI: Our Immortality or Extinction
« Reply #147 on: Today at 12:34:53 AM »
There is a neat game you can play with neural networks. The term of art is "Generative Adversarial Network."

Briefly, in the censorship case you train a network X to classify objectionable and unobjectionable media. From X write a generative model G, to generate media and label them as objectionable or unobjectionable. Then use X as a testbed to train the network G into an an adversarial network A that generates media that X misclassifies.

Rinse and repeat, you can retrain X on the media generated by A so X gets better. Then you can retrain A on the new X to A gets better.

As I said, neat. Here is an example of use in improving maps. It could, of course, help in degrading them ...

https://clarodatascience.com/2019/04/02/battling-toward-truth/

sidd

Sigmetnow

  • Multi-year ice
  • Posts: 15650
    • View Profile
  • Liked: 452
  • Likes Given: 215
People who say it cannot be done should not interrupt those who are doing it.

NeilT

  • Nilas ice
  • Posts: 1225
    • View Profile
  • Liked: 103
  • Likes Given: 8
Re: Robots and AI: Our Immortality or Extinction
« Reply #149 on: Today at 06:42:08 AM »
There is a neat game you can play with neural networks. The term of art is "Generative Adversarial Network."

But, in fact, the Holy grail is to teach the AI social values and why we classify something as objectionable and then it will make informed decisions for itself.

But that is an entirely different order of magnitude in AI.
Being right too soon is socially unacceptable.

Robert A. Heinlein