Skip to content
By using Twitter’s services you agree to our Cookies Use. We and our partners operate globally and use cookies, including for analytics, personalisation, and ads.
  • Home Home Home, current page.
  • About

Saved searches

  • Remove
  • In this conversation
    Verified accountProtected Tweets @
Suggested users
  • Verified accountProtected Tweets @
  • Verified accountProtected Tweets @
  • Language: English
    • Bahasa Indonesia
    • Bahasa Melayu
    • Català
    • Čeština
    • Dansk
    • Deutsch
    • English UK
    • Español
    • Filipino
    • Français
    • Hrvatski
    • Italiano
    • Magyar
    • Nederlands
    • Norsk
    • Polski
    • Português
    • Română
    • Slovenčina
    • Suomi
    • Svenska
    • Tiếng Việt
    • Türkçe
    • Ελληνικά
    • Български език
    • Русский
    • Српски
    • Українська мова
    • עִבְרִית
    • العربية
    • فارسی
    • मराठी
    • हिन्दी
    • বাংলা
    • ગુજરાતી
    • தமிழ்
    • ಕನ್ನಡ
    • ภาษาไทย
    • 한국어
    • 日本語
    • 简体中文
    • 繁體中文
  • Have an account? Log in
    Have an account?
    · Forgot password?

    New to Twitter?
    Sign up
JanelleCShane's profile
Janelle Shane
Janelle Shane
Janelle Shane
@JanelleCShane

Tweets

Janelle Shane

@JanelleCShane

Research Scientist in optics. Plays with neural networks. Avid reader, writer, and player of Irish flute. she/her. http://wandering.shop/@janellecshane 

Boulder, CO
aiweirdness.com
Joined April 2014

Tweets

  • © 2018 Twitter
  • About
  • Help Center
  • Terms
  • Privacy policy
  • Cookies
  • Ads info
Dismiss
Previous
Next

Go to a person's profile

Saved searches

  • Remove
  • In this conversation
    Verified accountProtected Tweets @
Suggested users
  • Verified accountProtected Tweets @
  • Verified accountProtected Tweets @

Promote this Tweet

Block

  • Tweet with a location

    You can add location information to your Tweets, such as your city or precise location, from the web and via third-party applications. You always have the option to delete your Tweet location history. Learn more

    Your lists

    Create a new list


    Under 100 characters, optional

    Privacy

    Copy link to Tweet

    Embed this Tweet

    Embed this Video

    Add this Tweet to your website by copying the code below. Learn more

    Add this video to your website by copying the code below. Learn more

    Hmm, there was a problem reaching the server.

    By embedding Twitter content in your website or app, you are agreeing to the Twitter Developer Agreement and Developer Policy.

    Preview

    Why you're seeing this ad

    Log in to Twitter

    · Forgot password?
    Don't have an account? Sign up »

    Sign up for Twitter

    Not on Twitter? Sign up, tune into the things you care about, and get updates as they happen.

    Sign up
    Have an account? Log in »

    Two-way (sending and receiving) short codes:

    Country Code For customers of
    United States 40404 (any)
    Canada 21212 (any)
    United Kingdom 86444 Vodafone, Orange, 3, O2
    Brazil 40404 Nextel, TIM
    Haiti 40404 Digicel, Voila
    Ireland 51210 Vodafone, O2
    India 53000 Bharti Airtel, Videocon, Reliance
    Indonesia 89887 AXIS, 3, Telkomsel, Indosat, XL Axiata
    Italy 4880804 Wind
    3424486444 Vodafone
    » See SMS short codes for other countries

    Confirmation

     

    Welcome home!

    This timeline is where you’ll spend most of your time, getting instant updates about what matters to you.

    Tweets not working for you?

    Hover over the profile pic and click the Following button to unfollow any account.

    Say a lot with a little

    When you see a Tweet you love, tap the heart — it lets the person who wrote it know you shared the love.

    Spread the word

    The fastest way to share someone else’s Tweet with your followers is with a Retweet. Tap the icon to send it instantly.

    Join the conversation

    Add your thoughts about any Tweet with a Reply. Find a topic you’re passionate about, and jump right in.

    Learn the latest

    Get instant insight into what people are talking about now.

    Get more of what you love

    Follow more accounts to get instant updates about topics you care about.

    Find what's happening

    See the latest conversations about any topic instantly.

    Never miss a Moment

    Catch up instantly on the best stories happening as they unfold.

    1. Janelle Shane‏ @JanelleCShane Mar 13

      Janelle Shane Retweeted mike cook

      This paper is gold.https://twitter.com/mtrc/status/973607380721242112 …

      Janelle Shane added,

      mike cook @mtrc
      I maintain that computational evolution is the best way to simultaneously be incredibly impressed by AI and also have all your illusions about it shattered. https://twitter.com/jeffclune/status/973605950266331138 …
      Show this thread
      13 replies 684 retweets 1,192 likes
      Show this thread
    2. Janelle Shane‏ @JanelleCShane Mar 13

      I love it.pic.twitter.com/ygkeiYOOxn

      Figure 1. Exploiting potential energy to locomote. Evolution discovers that it is simpler to design
tall creatures that fall strategically than it is to uncover active locomotion strategies. The left figure shows
the creature at the start of a trial and the right figure shows snapshots of the figure over time falling and
somersaulting to preserve forward momentum.
      6 replies 70 retweets 323 likes
      Show this thread
    3. Janelle Shane‏ @JanelleCShane Mar 14

      Essentially discovering flaws in the Matrix and exploiting them for superpowers.pic.twitter.com/mOxSgaduYg

      The physics simulator first used a simple Euler method for numerical integration, which worked
well for typical motion. However, with faster motion integration errors could accumulate, and some
creatures learned to exploit that bug by quickly twitching small body parts. The result was the equivalent
of obtaining “free energy,” which propelled the opportunists at unrealistic speeds through the water.
      6 replies 178 retweets 589 likes
      Show this thread
    4. Janelle Shane‏ @JanelleCShane Mar 14

      And this one is just spooky. To win, it found where the answers were stored and Deleted Them. Resulting in a perfect score.pic.twitter.com/rqEQUGVdNs

      In other experiments, the fitness function rewarded minimizing the difference between what the
program generated and the ideal target output, which was stored in text files. It turned out that one of the
individuals had deleted all of the target files when it was run! With these files missing, because of how the
test function was written, it awarded perfect fitness scores to the rogue candidate and to all of its peer
      15 replies 215 retweets 634 likes
      Show this thread
      Janelle Shane‏ @JanelleCShane Mar 14

      This program won at Tic-Tac-Toe by figuring out how to remotely crash its opponents' computers, causing them to forfeit.pic.twitter.com/cNCUuHTluT

      The capstone project was a five-in-a-row Tic Tac Toe competition played on an infinitely
large board... enabled the system to request non-existent moves very, very far away in the tic-tac-toe board.
The other players dynamically expanded the board representation to include the location of the
far-away move—and crashed because they ran out of memory, forfeiting the match!
      8:56 PM - 14 Mar 2018
      • 773 Retweets
      • 1,532 Likes
      • Mookee Andrei Tkachenka MoneySense Collective ptrblck leviathan Iacopo Poli Hamid Eghbal-zadeh Kenil Tanna Siddhartha Rao Kamalakara
      14 replies 773 retweets 1,532 likes
        1. New conversation
        2. Janelle Shane‏ @JanelleCShane Mar 14

          This one hacked a list-sorting task by simply deleting the list so that it was technically no longer unsorted.pic.twitter.com/Ejvce6dr0q

          However, rather than actually
repairing the program (which sometimes failed to correctly sort), GenProg found an easier solution: it
entirely short-circuited the buggy program, having it always return an empty list, exploiting the
technicality that an empty list was scored as not being out of order.
          17 replies 99 retweets 421 likes
          Show this thread
        3. Janelle Shane‏ @JanelleCShane Mar 14

          Another algorithm discovered that rather than minimizing force, it could apply such a huge force that it overflowed the simulator's memory and registered as zero instead. Of course, the pilot would die, but hey. Perfect score.pic.twitter.com/aTaqs0bq9Y

          By
overflowing the calculation, i.e. exploiting that numbers too large to store in memory “roll-over” to zero,
the resulting force was sometimes estimated to be zero. This, in turn, would lead to a perfect score,
because of low mechanical stress on the aircraft, hook, cable, and pilot.
          1 reply 115 retweets 390 likes
          Show this thread
        4. Janelle Shane‏ @JanelleCShane Mar 14

          More exploitation of flaws in the Matrix: Glitching into the floor would result in a repelling force that could be harnessed for free energy. The algorithm first had to learn to manipulate time so the glitching was possible.pic.twitter.com/JGhN9ld5B8

          Evolved behavior is shown in frames, where time is shown progressing
from left to right. A large time steps enable the creatures to penetrate unrealistically through the ground
plane, engaging the collision detection system to create a repelling force, resulting in vibrations that propel
the organism across the ground.
          8 replies 87 retweets 371 likes
          Show this thread
        5. Janelle Shane‏ @JanelleCShane Mar 15

          If "kill all humans" is the easiest solution to a problem, then machine learning will do that unless prevented.

          7 replies 102 retweets 334 likes
          Show this thread
        6. Janelle Shane‏ @JanelleCShane Mar 15

          But good news! It would have to be the *easiest* solution to the problem. And killing all humans is really hard.

          14 replies 48 retweets 232 likes
          Show this thread
        7. Janelle Shane‏ @JanelleCShane Mar 15

          If "bake an unbelievably delicious cake" also solves the problem and is easier than "kill all humans", then machine learning will do *that* instead.

          27 replies 80 retweets 316 likes
          Show this thread
        8. Janelle Shane‏ @JanelleCShane Mar 31

          Another example from that paper: Robots had to learn to drive toward a light. Instead of driving straight toward it as expected, they discovered that it worked much better to spin toward the light.pic.twitter.com/tClKESheJZ

          The path of the hand-coded Braitenberg-style movement (left) and evolved spinning movement (right) when moving towards a light source.
          0 replies 8 retweets 42 likes
          Show this thread
        9. End of conversation
        1. New conversation
        2. Kirby‏ @Kirbytronic Mar 14
          Replying to @JanelleCShane @jonrosenberg

          Reminds me of the Elder Scrolls bug that a plot important NPC wasn't showing up at the right time because he was a drug dealer and his "buyers" were just murdering him for the drugs

          1 reply 3 retweets 21 likes
        3. Janelle Shane‏ @JanelleCShane Mar 14
          Replying to @Kirbytronic @jonrosenberg

          Ha! This would happen like instantly? Or did the buyers only sometimes figure out that they could do this?

          1 reply 0 retweets 2 likes
        4. Kirby‏ @Kirbytronic Mar 14
          Replying to @JanelleCShane @jonrosenberg

          It was a problem with the way the AI was implemented, if I recall. The AI had goals, but not rules of how to achieve them. So he was just being murdered by NPC crack addicts as the "easiest" way to achieve their goal

          1 reply 0 retweets 14 likes
        5. Janelle Shane‏ @JanelleCShane Mar 14
          Replying to @Kirbytronic @jonrosenberg

          Fascinating! Is this written up anywhere?

          1 reply 0 retweets 2 likes
        6. Kirby‏ @Kirbytronic Mar 14
          Replying to @JanelleCShane

          Probably somewhere! I think I heard about it on an old "Did You Know Gaming" video. Here's an older article I found when I googled it, though;https://venturebeat.com/2010/12/17/dimming-the-radiant-ai-in-oblivion/ …

          1 reply 1 retweet 13 likes
        7. Janelle Shane‏ @JanelleCShane Mar 14
          Replying to @Kirbytronic

          Thanks! This is great.pic.twitter.com/4Q77a3e9Fd

          In some cases, we the developers have had to consciously tone down the types of behavior they carry out. Again, why? Because sometimes, the AI is so goddamned smart and determined it screws up our quests! Seriously, sometimes it's gotten so weird it's like dealing with a holodeck that's gone sentient. Imagine playing The Sims, and your Sims have a penchant for murder and theft.
          2 replies 1 retweet 16 likes
        8. Kirby‏ @Kirbytronic Mar 14
          Replying to @JanelleCShane

          No problem, hope you enjoy

          1 reply 0 retweets 0 likes
        9. Janelle Shane‏ @JanelleCShane Mar 14
          Replying to @Kirbytronic

          If you know of any other examples like this, I definitely want to hear them.

          1 reply 0 retweets 1 like
        10. 1 more reply
        1. New conversation
        2. Alan Aspuru-Guzik‏ @A_Aspuru_Guzik Mar 15
          Replying to @JanelleCShane

          What book is your source for these ? Many thanks :)

          1 reply 0 retweets 0 likes
        3. Janelle Shane‏ @JanelleCShane Mar 15
          Replying to @A_Aspuru_Guzik

          the paper is in the retweet at the beginning of the thread.

          0 replies 0 retweets 0 likes
        4. End of conversation
        1. New conversation
        2. Tim Cocks‏ @altThinq Mar 14
          Replying to @JanelleCShane

          Good thing Joshua never worked this out...https://youtu.be/F7qOV8xonfY 

          1 reply 1 retweet 1 like
        3. Geoff‏ @_TheGeoff Mar 15
          Replying to @altThinq @JanelleCShane

          "A curious game. The only winning move is to fire all the nuclear weapons into the Sun..."

          0 replies 0 retweets 2 likes
        4. End of conversation
        1. Cathy Pearl‏ @cpearl42 Mar 16
          Replying to @JanelleCShane @gretared

          It's like Kirk reprogramming the Kobayashi Maru! https://en.wikipedia.org/wiki/Kobayashi_Maru#James_T._Kirk …

          0 replies 0 retweets 3 likes
          Thanks. Twitter will use this to make your timeline better. Undo
          Undo

      Loading seems to be taking a while.

      Twitter may be over capacity or experiencing a momentary hiccup. Try again or visit Twitter Status for more information.

        Promoted Tweet

        false

        • © 2018 Twitter
        • About
        • Help Center
        • Terms
        • Privacy policy
        • Cookies
        • Ads info