How will the super intelligences (SIs) of the future function? I define a SI as a single entity that can communicate about a broad range of topics as fast as any machine and as correctly as any well-organized team of topic-specific human experts. So how does it work? The SI is a Six State Logic (6SL) based expert system/machine that teams of humans massively parallel program to communicate on a broad range of topics. A well-functioning SI presents an always available interface for any human or machine that wants to communicate about any programmed topic.
Consider the example of a team that wants to teach an SI how to sell their product. The sales team starts by creating counter responses to prospective customer’s Six State Logic (6SL) answers to the question “Will you purchase my product?” The sales team’s counter responses include (1) information for Uncertain prospects, (2) purchase instructions for Known Positive prospects, (3) please tell us why requests for Known Negative prospects, and combined and/or multiple level responses for (4) Mixed, (5) Seems Positive, and (6) Seems Negative prospects.
Once programed, the SI becomes the first point of contact for all prospective customers. A typical SI and prospect interaction involves (1) the prospect or the SI initiates communications by asking about product purchase, (2) the prospect provides a Six State Logic status regarding its desire to purchase, (3) the SI matches its response with the prospects purchase status, (4) responses and counter responses continue to until the transaction is concluded, put on hold pending new information, or abandoned.
For example, a prospect could approach the SI with a Seems Positive status for purchase with the reason for uncertainty being the need to check current price. That SI provides the price information and the prospect changes its status to Known Positive. The SI then proceeds to conclude the transaction.
An alternative example is when an SI with a Known Positive status for selling a product ask a prospect to purchase. The prospect answers with a Known Negative response with the reason that a similar product purchase was made recently. The SI and prospect could then agree to communicate in the future when the prospect needs the product again. The SI could try to gather competitive information if the prospect is willing or able to communicate on that topic. The SI could then store this prospect and product sales status as Mixed because the prospect has a similar product purchase history, but has recently purchased a competitive product.
The sales team refines SI programming by identifying cases where a sales prospect asks questions the team has not prepared answers for or cases when the sales process does not progress smoothly to a purchase, conditional purchase, do not purchase, or conditional do not purchase conclusion. The sales support team uses 6SL consensus teamwork tools (https://youtu.be/NtKVA2GitY4) to rapidly answer questions the SI can not answer and to prepare SI programming enhancements.
The question and answer script prepared for a SI can require multiple layers of 6SL responses. For example, if in response to the “Will you purchase my product?” question a prospect with a Seems Positive answer could ask the counter question, “Can you deliver tomorrow?” In response the SI could provide a Seems Positive answer and the counter question, “Can you pay for express shipping?” If the answer to this counter question from the prospect is Known Positive then the SI would change its original question to “Will you purchase my product if we ship it to arrive tomorrow and you pay for shipping?”
The primary difference between 6SL expert systems and a yes/no answer based expert system is the enhanced information value of 6SL responses. For example, a Known Negative response with a well understood reason effectively ends the SI interaction. This is because the prospect did not select any other 6SL states where follow up questions are required. This short path between question and justified action contrasts with a simple “no” response to the purchase question which would requires many follow up questions to generate a similar level of understanding as the single 6SL answer provided.
In summary, the 6SL value points that help enable SIs are: (1) SI interactions are limited to prepared topics, (2) the SI only interacts with machines and humans using 6SL format questions and responses, (3) each topic is supported by a combination of human created and updated SI programming and real time interactions with human programmers when the SI’s programming is insufficient to conclude a transaction. It is expected that a properly functioning SIs will create a prioritized backlog of required programming updates for its human programmers.
The key value of the SI is the ability to communicate quickly across a broad range of topics in a continuously improving way. These abilities are particularly helpful when two well supported SIs interact and identify a broader range of opportunities in less time than any similar human managed interaction system could. Consider all the economic opportunities that an Apple SI could identify when interacting with a Samsung SI or possibly a USA national SI when interacting with a Japanese national SI.
Michael Klasen, PhD
Creator of Six State Logic