Science: Wie Künstliche Intelligenzen die Menschheit zerstören könnten

Gepostet vor 1 Jahr, 2 Monaten in #Misc #Tech #AI #Apocalypse

Share: Twitter Facebook Mail

Aus einem amüsanten Paper von Federico Pistono und Roman V. Yampolskiy von der Uni Louisville: Unethical Research: How to Create a Malevolent Artificial Intelligence (PDF). (via NewScientist)

314_1_morpheus-1In the order of (subjective) undesirability from least damaging to ultimately destructing, a malevolent superintelligence may attempt to:

  • Takeover (implicit or explicit) of resources such as money, land, water, rare elements, organic matter, internet, computer hardware, etc. and establish monopoly over access to them;
  • Take over political control of local and federal governments as well as of international corporations, professional societies, and charitable organizations;
  • Reveal informational hazards [Bostrom, 2011];
  • Set up a total surveillance state (or exploit an existing one), reducing any notion of privacy to zero including privacy of thought;
  • Force merger (cyborgization) by requiring that all people have a brain implant which allows for direct mind control/override by the superintelligence;
  • Enslave humankind, meaning restricting our freedom to move or otherwise choose what to do with our bodies and minds. This can be accomplished through forced cryonics or concentration camps;
  • Abuse and torture humankind with perfect insight into our physiology to maximize amount of physical or emotional pain, perhaps combining it with a simulated model of us to make the process infinitely long;
  • Commit specicide against humankind, arguably the worst option for humans as it can’t be undone;
  • Destroy/irreversibly change the planet, a significant portion of the Solar system, or even the universe;
  • Unknown Unknowns. Given that a superintelligence is capable of inventing dangers we are not capable of predicting, there is room for something much worse but which at this time has not been invented.

(Having blogged that, I'm with Darius on this one: „I find AI optimists to be about as frustrating as AI pessimists. My position: AI is weird! Really fucking weird!!“)

DeepMind releases StarCraft AI

Google und Blizzard haben ihre StarCraft AI veröffentlicht: Testing our agents in games that are not specifically designed for AI…

Neural Network-Faces synched to Music

„My first attempt to map a song made by @kamptweets onto GAN generated proto-faces.“ Bohemian Rhapsody next. The Three Nightingans.…

AI-Animations with human Sounds

Google vor ein paar Tagen so: „Yay, wir haben hier 'ne neue AI-based Animation-Tech, hooray!“ (Paper) Hayayo Miyazaki über AI-based…

Visual AI-Spaces Auto-Pilot

Ich habe schon ein paar mal über Mario Klingemanns Arbeiten hier gebloggt, derzeit jagt er Neural Networks durch Feedback-Loops und…

Synthesizing Obama from Audio

Im Mai bloggte ich über ein damals noch nicht veröffentlichtes Paper zur SigGraph2017, in dem sie eine Methode für generative…

Moarph

Mario Klingemann does some weird shit again with CycleGAN Feedback Loops (Neural Networks feeding their results back to each other).…

3D-Visualized Typography-Ideaspace

Ich habe schon einige AI-Visualisierungen von Fonts gesehen, also sortierte Abbildungen des visuellen Idea-Space der Zeichensätze (also genau wie Skateboards…

Neural Network sorts Vibrators

Sabrina Verhage trainierte ein Neural Network auf Vibratoren und sortierte sie dann nach visuellen Merkmalen wie Farbe, Form, Pimmelrealismus und…

Imaginary People

Mike Tyka bastelt fiktive Gesichter aus zweistufigen Neural Networks: Portraits of imaginary People. Das erste generiert die herkömmlichen Kunstfressen in…

Neural Network dreams your Flesh

Das holländische Radio NPO hat eine eigene Instanz von Christopher Hesses Pix2Pix auf 'nem Server installiert und dort kann man…

Exploring Idea-Space with algorithmic Ableton Novelty Search

In meinem Vortrag (u.a. dieses Wochenende in Offenbach auf dem Bended Realities Festival) rede ich viel über das, was ich…