Artificial Superintelligence Coordination & Strategy (Roman V. Yampolskiy & Allison Duettmann, 2020)

-


  • Artificial intelligence-enabled adaptive learning systems (AI-ALS)
  • Intelligent tutoring systems (ITS)
  • Bayesian Knowledge Tracing (BKT)
  • Vicarious trial and error (VTE)
  • Null hypothesis significance testing (NHST)
  • Probabilistic structural equation modeling (PSEM)
Artificial General Intelligence (AGI)
  • Intelligence amplification—AI can improve its own intelligence
  • Strategy—optimizing chances of achieving goals using advanced techniques, e.g., game theory, cognitive
  • psychology, and simulation
  • Social manipulation—psychological and social modeling e.g., for persuasion
  • Hacking—exploiting security flaws to appropriate resources
  • R&D—create more powerful technology, e.g., to achieve ubiquitous surveillance and military dominance
  • Economic productivity—generate vast wealth to acquire resources
Safety on AGI
  • Non-hackability and non-censurability via decentralization (storage in multiple distributed servers),
  • encryption in standardized blocks, and irrevocable transaction linkage (the “chain”)
  • Node-fault tolerance: Redundancy via storage in a decentralized ledger of (a) rules for transactions, (b) the transaction audit trail, and (c) transaction validations
  • Transparency of the transaction rules and audit trail in the DLT
  • Automated “smart” contracts
  • Decentralized applications (“dApps”), i.e., software programs that are stored and run on a distributed network and have no central point of control or failure
  • Validation of contractual transactions by a decentralized consensus of validators
Forecasting
  • Trend-impact analysis (TIA)
  • Cross-impact analysis (CIA)
  • The intuitive logics school
  • The probabilistic modified trends (PMT) school
  • La Prospective, a.k.a. the French school
Single-agent overoptimization failure modes outlined
  • Tails Fall Apart, or Regressional inaccuracy, where the relationship between the modeled goal and the true goal is inexact due to noise (for example, measurement error,) so that the bias grows as the system is optimized.
  • Extremal Model Insufficiency, where the approximate model omits factors which dominate the system’s behavior after optimization.
  • Extremal Regime Change, where the model does not include a regime change that occurs under certain (unobserved) conditions that optimization creates.
  • Causal Model Failure, where the agent’s actions are based on a model which incorrectly represents causal relationships, and the optimization involves interventions that break the causal structure the model implicitly relies on.

Comments

Post a Comment

Popular posts from this blog

Kokology Questions & Answers

-
[These are extracts only, to see the full details kindly purchase the original book] Kokology is the study of kokoro (Japanese: 心) which in the aforementioned language means "mind" or "spirit". ... The books were published in 1998 in Japan and became a Japanese bestselling phenomenon. The books were translated and became available in the United States in 2000. Below are some questions and answers for sharing: A Desert Journey Boring class settings, you drifted away and enter the eternal desert... 1. You are riding a camel and you are near exhaustion, what will you say to the camel? 2. Before you die of thirst, a traveler point to you a oasis, who is this person you know? 3. You finally come to your destination which is a town, what is your feeling? 4. After you dismount, another person took over, who is this guy? - 1. What happened when love is lost? 2. Who always help or comforted you? 3. What is your feeling after getting past lost love?...
Read more

Neuro-Linguistic Programming Models Summary (02 of 14)

-
GeneraliSation The The Meta Model generaliSations make explicit the way we summarise and categorise our experiences. Generalisations are basically the rules we use to manage life. There are usually exceptions or counter examples to rules. These rules and categories create self-imposed or company wide limitations. Examining these limitations is a powerful problem solving strategy. By generalising one or a few experiences, we can figure out what to do in a different experience. We learn to drive a particular car, and then use that experience to drive a new one. If we weren’t able to generalise, we would have to learn how to drive all over again. Problem Solving Examining categories and generalisations is an excellent creative problem solving strategy. What are the experiences that come to mind when you think of money for instance? We typically use an emotionally charged experience to stand in for all our experiences in a category. We generalise and categorise all kinds ...
Read more

Neuro-Linguistic Programming Models Summary (11 of 14)

-
The Inner Hero Pattern Beat the drum and let the poets speak. This is the day of purification for those who Are already mature and initiated into what love is. No need to wait until we die! There’s more to want here than money And being famous and bites of roasted meat. --Rumi Life can be lived in many ways. You can make it about making money or winning at all costs, or pleasing other people, or perhaps never standing out. Or you can live your life as a great journey of consciousness, one filled with many challenges and surprises, one that makes a positive contribution to the world. I want to talk here a bit about these different paths, emphasising that the Self-Relations approaches of Generative Self and Generative Trance are especially tools for supporting the latter path. Life as a journey has been described by many people, most notably the mythologist Joseph Campbell.  Campbell (1949; Gilligan and Dilts, 2009) studied the stories of many different cultur...
Read more