Summary of Let's build an ACOG (artificial cognitive entity) - Part 5: preparing to act

00:00:00 - 00:35:00

This video discusses how artificial cognition works by breaking it down into three steps: assessment, deliberation, and action. It demonstrates how this process can be summarized into a "loose approximation" of what happens in your head when you're deciding what to do.

• 00:00:00 In this video, David Shapiro explains how artificial cognition works by breaking it down into nested loops. The first step is assessment, which is the automatic inference of what is happening in the world. Next, memories are recalled based on what is seen and heard. Finally, ideas are generated and discernment is used to decide which one to pursue.
• 00:05:00 In this video, a programmer explains how to create artificial cognitive entities (ACOs), using three steps: creating a scenario generator, feeding in scope and severity modifiers, and generating a story. Each ACO is based on a different context, ranging from a single person to a global company.
• 00:10:00 In this video, a model is explained for artificial cognition, in which a situation is assessed and information is gathered to help improve the situation. The context is the same for both scenarios. In the first scenario, a woman is having a bad day at work, and her husband is preoccupied with taking care of the kids. In the second scenario, an oil spill is occurring.
• 00:15:00 The video discusses how to create an artificial cognitive entity, or ACO, by analyzing a scenario and making inferences about what could have happened. The ACO is designed to help with decision-making, and the video discusses how to create an effective prompt for it.
• 00:20:00 In the video, the presenter discusses how to build an artificial cognitive entity that can anticipate and evaluate events, predict outcomes, and make decisions. The presenter then presents a memory example to illustrate how the artificial cognitive entity would work.
• 00:25:00 The video discusses the need for an artificial cognitive entity, or "ACO," to be able to make decisions and carry out actions in a given situation. It explains the process of creating an ACO, including the need for agents to be understandable, explainable, and observable. It finishes by giving an example of an ACO, a humanoid robot with average human strength, speed, and intelligence.
• 00:30:00 This YouTube video explains how to create an artificial cognitive entity, or ACOE, by first preparing it with specific agent models of identity and purpose. The ACOE is then allowed to read a scenario, and is asked what to do next. If the ACOE is not adopting the eye, it is explained that this is because the scenario is too long.
• 00:35:00 In this video, the creator of a "Let's build an ACOG (artificial cognitive entity)" explains how the cognitive process works, from assessment (what should the company's robot do), to deliberation (deciding that's not helpful), to action (calling someone in charge at that company). He then demonstrates how this process can be summarized into a "loose approximation" of what happens in your head when you're deciding what to do.