What is Socrates.Digital™?
Socrates Digital™ is a new type of computer application that is described in Mark Salisbury’s new book, Socrates Digital for Learning and Problem-Solving. Socrates Digital™ will change how we solve problems -- and with Mark’s book, the software is practically free! In his book, Mark shows how to write a Socrates Digital™ application in a general-purpose programming language that leverages cloud-based artificial intelligence services such as natural language processing and sophisticated data processing techniques. Offerors of these cloud-based AI services include Microsoft, Apple, Google, IBM, and Amazon. Most of these cloud-based AI services come with no-code development features that provide a quick and easy means for creating a Socrates Digital™ application. For example, Mark has created a no-code implementation of Socrates Digital™ with Microsoft’s Power Virtual Agent that can be added to Microsoft Teams. Since a version of Power Virtual Agent comes with Teams, this implementation of Socrates Digital™ is essentially free to organizations that use Office 365.
This means that the technology aspect of Socrates Digital™ has been solved with Mark’s book and the wide availability of AI services that are easily employed. What has not been solved is how organizations can embrace and use a Socrates Digital™ application to solve their complex problems – and how we can teach this problem-solving to a new generation.
While Mark Salisbury’s book describes the design and implication of a Socrates Digital™ system, Socrates Digital™ is also a brand name for professional services offered to assist organizations in developing Socrates Digital™ systems and embrace them to improve problem-solving. Click on the “Services” tab to view these professional services. To learn more about Mark Salisbury, you can go to his website, here.
Why Use Socrates.Digital™?
Our ability to learn and problem-solve has increased very little since computers were invented. That’s because of the way we were taught to solve problems in school. Most of our time was spent on gaining factual knowledge about various phenomena and the problems associated with them. As for solving problems, we are taught to solve them by recalling all the problems we have memorized, select a problem that matches the problem which we are facing, and adapt and apply the associated solution of the selected problem to the problem that we are facing. From the field of logic, we call this deductive reasoning. We are applying a general rule to a problem in order to solve it.
This view of human problem-solving has tremendous implications for the quality of our work and the society we live in. In the workplace, it means that we focus our effects on learning about a problem. And we spend very little time on the problem-solving process itself. This view of human problem-solving also impacts our educational system. In fact, we reinforce this view in our educational systems and our students take it to the workplace. As a result, we spend most of our energy and time teaching students about the nature of problems. We spend very little time with them in learning a process to solve problems.
This misguided focus has finally caught up with us. Problems have become too complex to solve without a using a concentrated effort to follow a problem-solving process. In addition, this problem-solving process must be able to deal with big and disparate data. It must also be able to solve problems that don’t have a “rule” to for solving them. That means it needs to be inductive in nature. And finally, it must be able to deal with ambiguity and help humans use informed judgement to build on previous steps and create new understanding. This is where Socrates.Digital™ comes in.
Socrates Digital™, like the human Socrates, guides users with Socratic questions to examine information, concepts, assumptions, conclusions, and implications that go into formulating a solution to a problem. Furthermore, Socrates Digital™ can do more than a human in guiding users through the Socratic problem-solving process. As any teacher who has committed to Socratic problem-solving can tell you, it is challenging and time consuming to learn how to do it effectively. It also takes planning for its successful application and considerable guidance to keep it on track with a group of learners and problem-solvers. In addition, it is labor intensive to document the outcome of a Socratic problem-solving session and make those results available to the participants and a larger audience.
Socrates Digital™ can help with all these challenges. Like an expert teacher in Socratic problem-solving, it guides the conversation so that all the important questions are answered to solve the problem. Problem-solvers can immediately use Socrates Digital™ without having the expertise to conduct a Socratic problem-solving session by themselves. This greatly reduces the time and effort needed to get up to speed on how to conduct sessions, plan for the sessions, and guide those sessions. An added benefit is that a Socrates Digital application can use knowledge management techniques to capture results during problem-solving sessions and make those results immediately available to participants and a larger audience. This is knowledge management at its potential since it can provide the most important knowledge any organization will ever manage – how it solves problems -- to its members in an instant.
Socrates Digital™ also enhances the effectiveness of problem-solving in ways that only a computer can do. For example, Socrates Digital™, can guide human problem-solvers in assessing the quality of their reasoning. As part of this process, it can assist human problem-solvers in determining a numerical value for their level of confidence in a single step in a session – for example, how much confidence they have in the information analysis step. Socrates Digital™ can also combine the confidence levels from several steps into an overall confidence level for the solution produced in the session. As a result, Socrates Digital™ ushers in a new way to solve problems that greatly enhances the ability of problem-solvers to create solutions for complex problems and be able to communicate the reasoned judgement behind those solutions. This will change problem-solving as we know it and how we teach it in our educational institutions.
How does Socrates.Digital™ Work?
The following dialog shows how Socrates Digital™ can guide users to learn and solve problems with complex and disparate data in a business setting. In the following scenario, Socrates Digital™ helps guide users to combine data and reasoning from multiple sources such as spreadsheets, databases, survey data, and firsthand reports to determine if an investment opportunity is a good one. Not only is the data from multiple sources, but it has been analyzed with different methods. Some of the data has been tallied into totals and subtotals. Subsets of other data have been selected with criteria through search queries. Data has also been analyzed with statistical methods to measure central tendencies and correlations between values. In addition, data has also been analyzed with techniques from the field of artificial intelligence including textual analysis and the application of machine learning algorithms.
For this scenario, Socrates Digital™ was initially configured by investment analysts for a conversation with end users – who are also analysts – about how to analyze a commercial real estate investment opportunity. This initial configuration includes investment concepts and assumptions that investment analysts routinely use to evaluate investment opportunities in the motel real estate industry. In this example dialog, Socrates Digital™ begins with these concepts and assumptions and guides two groups of users to learn about a particular investment opportunity and solve the problem of whether to invest in that opportunity. The user groups are identified as the “User Persona,” abbreviated as “UP.” The first group of users are abbreviated as “UP1” and the second group of users is abbreviated as “UP2.” During the dialog, Socrates Digital™, uses questions from Socratic Questioning categories based on critical thinking concepts and tools taken from Paul and Elder’s The Thinker’s Guide to Socratic Questioning.
SOCRATES DIGITAL: What can I help you with today? (Questions that Target the Parts of Thinking -- Questioning Goals and Purposes)
UP1: Deciding whether to invest in the Rosebud Motel1 chain.
SOCRATES DIGITAL: So, the question at hand is deciding whether to invest in the Rosebud Motel chain? (Questions that Target the Parts of Thinking – Question at Issue)
UP1: Yes.
SOCRATES DIGITAL: What is the main concept that you will use to make for your decision? (Questions that Target the Parts of Thinking -- Questioning Concepts and Ideas)
UP1: We need to know if the investment will boost total revenue enough to pay back the investors.
SOCRATES DIGITAL: How does it relate to the question at hand? (Questions that Target the Quality of Reasoning -- Questioning Relevance)
UP1: We need to know if the investment will increase the total revenue for the Rosebud Motel chain so that it can repay the investment at the projected rate of return.
SOCRATES DIGITAL: OK, of the following, what would you like to work on? (Questions that Target the Parts of Thinking -- Questioning Questions)
Information
Assumptions
Concepts
UP1: Concepts, I believe.
SOCRATES DIGITAL: Which concept do you want to use to explain how to answer the question at hand? “We need to know if the investment will boost total revenue enough to pay back the investors.” (Questions that Target the Parts of Thinking -- Questioning Concepts and Ideas)
The investment analysts, who initially configured Socrates Digital™, provided these concepts as widely used concepts for evaluating investment opportunities in the motel real estate industry.
Occupancy Rate – a measure of the percentage of total rooms rented
Average Daily Rate – a measure of average room price
Revenue per Available Room – a measure of average room price, including vacant rooms
Another Concept
UP1: Occupancy rate.
SOCRATES DIGITAL: How well does this concept explain how to answer the question at hand? (Questions that Target the Quality of Reasoning -- Questioning Significance)
The values for each option that Socrates Digital will use to calculate the strength of the concept have been displayed for this example dialog. These values would not be typically displayed to users.
Always Explains (0.75)
Mostly Explains (0.5)
Sometimes Explains (0)
Rarely Explains (-0.5)
Never Explains (-0.75)
UP1: Sometimes explains.
SOCRATES DIGITAL: OK, of the following, what would you like to work on? (Questions that Target the Parts of Thinking -- Questioning Questions)
Information
Assumptions
UP1: Information.
SOCRATES DIGITAL: What information, data, or experience do you have that supports or informs this analysis? (Questions that Target the Parts of Thinking – Information, Data, and Experience)
UP1: I have spreadsheet data on occupancy rates and revenue for the Rosebud Motel chain.
SOCRATES DIGITAL: Can you give me an example data point about this? (Questions that Target the Quality of Reasoning -- Questioning Clarity)
UP1: As the spreadsheet shows, the Dundalk Inn & Suites motel has low occupancy rates and low total revenue.
SOCRATES DIGITAL: Do any of the following assumptions underlie this observation? (Questions that Target the Parts of Thinking -- Questioning Assumptions)
The investment analysts, who initially configured Socrates Digital™, provided these assumptions as commonly accepted for evaluating investment opportunities in the motel real estate industry.
Occupancy Rate is positively correlated with total revenue.
Average Daily Rate is positively correlated with total revenue.
Revenue per Available Room is positively correlated with total revenue.
UP1: Occupancy rate is positively correlated with total revenue.
SOCRATES DIGITAL: Do you want to apply this assumption in your reasoning? (Questions that Target the Quality of Reasoning -- Questioning Significance)
As with presenting concepts to users, the values for each option that Socrates Digital will use to calculate the confidence in the assumption have been displayed for this example dialog. These values would not be typically displayed to users.
Definitely Apply It (0.75)
Probably Apply It (0.5)
Maybe Apply it (0)
Disregard It for Now (-0.5)
Delete It (-0.75)
UP1: Probably keep it.
SOCRATES DIGITAL: Does this assumption hold for the larger dataset? (Questions that Target the Parts of Thinking -- Questioning Information, Data, and Experience)
UP1: Yes. As the spreadsheet scatterplot shows, the Rosebud Motel chain occupancy rate is positively correlated with total revenue.
SOCRATES DIGITAL: How would you rate the accuracy of this information? Which following statement more closely represents your confidence in its accuracy? (Questions that Target the Quality of Reasoning -- Questioning Accuracy)
As with presenting concepts and assumptions to users, the values for each option that Socrates Digital will use to calculate the accuracy of the information have been displayed for this example dialog. These values would not be typically displayed to users.
Accurate (0.75)
Probably Accurate (0.5)
Unknown Accuracy (0)
Probably Not Accurate (-0.5)
Inaccurate (-0.75)
UP1: I think that it is probably accurate.
SOCRATES DIGITAL: Given these concepts, assumptions, and information, what is the most reasonable conclusion? (Questions that Target the Parts of Thinking -- Questioning Inferences and Conclusions)
UP1: Raising the occupancy rate for all the motels in the Rosebud Motel chain would raise total revenue for the chain.
SOCRATES DIGITAL: Given this is all the evidence that you currently have, how would you rate the logic of this conclusion? (Questions that Target the Quality of Reasoning -- Questioning Logic)
The same is true for rating the logic of conclusions, the values for each option that Socrates Digital will use to calculate the logic of a conclusion have been displayed for this example dialog. These values would not be typically displayed to users.
Very Logical (0.75)
Probably Logical (0.5)
Unknown How Logical (0)
Probably Not Logical (-0.5)
Illogical (-0.75)
UP1: I would say that it is probably logical.
SOCRATES DIGITAL: What are the implications of this conclusion? (Questions that Target the Parts of Thinking -- Questioning Implications and Consequences)
UP1: Increased marketing can raise the occupancy rate and total revenue for motels in the Rosebud Motel chain.
SOCRATES DIGITAL: How would you rate your confidence in the implications of this conclusion? (Questions that Target the Quality of Reasoning -- Questioning Depth)
Again, the values for each option that Socrates Digital™ will use to calculate the confidence in the implications of a conclusion have been displayed for this example dialog. These values would not be typically displayed to users.
Highly Justified (0.75)
Probably Justified (0.5)
Insufficient Evidence to Judge (0)
Not Justified by the Evidence (-0.5)
Refuted by the Evidence (-0.75)
UP1: Insufficient evidence to judge.
SOCRATES DIGITAL: Can we get more confirming evidence about this? (Questions that Target the Parts of Thinking -- Questioning Information, Data, and Experience)
UP1: Yes, I believe we can.
/* Socrates Digital continues to work with the user persona on information, concepts, assumptions, conclusions, and implications. We pick up the conversation below. */
UP2: Always explains. Let’s look at information.
Since the User Persona answered the following question ahead of time, the natural language system skips over the question and applies the answer to advance the program.
SOCRATES DIGITAL: OK, of the following, what would you like to work on? (Questions that Target the Parts of Thinking -- Questioning Questions)
Information
Assumptions
UP2: Let’s look at Information.
SOCRATES DIGITAL: Do you have an example data item that illustrates this? (Questions that Target the Quality of Reasoning -- Questioning Clarity)
UP2: Rosebud Motels put $145,000 into Saint Ola Inn and raised the sentiment analysis score for comments from 0.3 (average for Rosebud Motels) to 0.9 by making repairs, painting the interior and exterior, staff training on customer service, and creating promotional programs with local restaurants and stores to improve customer experience and raise total revenue in the Rosebud Motel chain. Saint Ola raised its total revenue by 4.6%.
SOCRATES DIGITAL: Any of these assumptions underlie this observation? (Questions that Target the Parts of Thinking -- Questioning Assumptions)
Occupancy Rate is positively correlated with total revenue.
Average Daily Rate is positively correlated with total revenue.
Revenue per Available Room is positively correlated with total revenue.
Customer Experience is positively correlated with total revenue.
UP2: Customer experience is positively correlated with total revenue. Let’s definitely apply it.
Since the User Persona answered the following question ahead of time, the natural language system skips over the question and applies the answer to advance the program.
SOCRATES DIGITAL: Do you want to apply this assumption in your reasoning? (Questions that Target the Quality of Reasoning -- Questioning Significance)
Definitely Apply It (0.75)
Probably Apply It (0.5)
Maybe Apply it (0)
Disregard It for Now (-0.5)
Delete It (-0.75)
UP2: Let’s definitely apply it.
SOCRATES DIGITAL: Is this assumption true for the larger dataset? (Questions that Target the Parts of Thinking -- Questioning Information, Data, and Experience)
UP2: Yes. In the Rosebud Motel chain, an analysis of key influencers shows that customer experience influences total revenue. A more detailed analysis of key influencers shows that total revenue is more likely to increase when sentiment score is higher than 0.5 (neutral). I think that it is probably accurate.
Since the User Persona answered the following question ahead of time, the natural language system skips over the question and applies the answer to advance the program.
SOCRATES DIGITAL: How would you rate the accuracy of this Information? Which following statement more closely represents your confidence in its accuracy? (Questions that Target the Quality of Reasoning -- Questioning Accuracy)
Accurate (0.75)
Probably Accurate (0.5)
Unknown Accuracy (0)
Probably Not Accurate (-0.5)
Inaccurate (-0.75)
UP2: I think that it is probably accurate.
SOCRATES DIGITAL: With these concepts, assumptions, and information, what is the best conclusion? (Questions that Target the Parts of Thinking -- Questioning Inferences and Conclusions)
UP2: This is more evidence that the Rosebud Motel chain can increase revenue by improving customer experience, as measured by sentiment analysis scores of comments about staying at their motels. I would say that it is very logical.
Since the User Persona answered the following question ahead of time, the natural language system skips over the question and applies the answer to advance the program.
SOCRATES DIGITAL: Given this is all the evidence that you currently have, how would you rate the logic of this conclusion? (Questions that Target the Quality of Reasoning -- Questioning Logic)
Very Logical (0.75)
Probably Logical (0.5)
Unknown How Logical (0)
Probably Not Logical (-0.5)
Illogical (-0.75)
UP2: I would say that it is very logical.
SOCRATES DIGITAL: Can you explain the implications of this conclusion? (Questions that Target the Parts of Thinking -- Questioning Implications and Consequences)
UP2: This is more evidence that the viewpoint for making repairs, painting the interior and exterior, staff training on customer service, and creating promotional programs with local restaurants and stores that improves customer experience and increases total revenue in the Rosebud Motel chain. I think it is highly justified.
Since the User Persona answered the following question ahead of time, the natural language system skips over the question and applies the answer to advance the program.
SOCRATES DIGITAL: How would you rate your confidence in the implications of this conclusion? (Questions that Target the Quality of Reasoning -- Questioning Depth)
Highly Justified (0.75)
Probably Justified (0.5)
Insufficient Evidence to Judge (0)
Not Justified by the Evidence (-0.5)
Refuted by the Evidence (-0.75)
UP2: I think it is highly justified.
SOCRATES DIGITAL: Can you rephrase these current implications, listed below, into a viewpoint? (Questions that Target the Parts of Thinking -- Questioning Viewpoints)
A viewpoint is needed for making repairs, painting the interior and exterior, staff training on customer service, and creating promotional programs with local restaurants and stores that improves customer experience and increases total revenue in the Rosebud Motel chain.
This is more evidence that the viewpoint for making repairs, painting the interior and exterior, staff training on customer service, and creating promotional programs with local restaurants and stores that improves customer experience and increases total revenue in the Rosebud Motel chain.
UP2: The viewpoint is focused on making repairs, painting the interior and exterior, staff training on customer service, and creating promotional programs with local restaurants and stores that improves customer experience to increase total revenue in the Rosebud Motel chain.
SOCRATES DIGITAL: Can you provide more data that backs this up? (Questions that Target the Parts of Thinking -- Questioning Information, Data, and Experience)
UP2: Not at this time.
SOCRATES DIGITAL: You have expressed an overall confidence of .81 (good confidence) for this viewpoint.
1Any similarity to the Rosebud Motel in the situation comedy “Schitt's Creek" (TV Series 2015–2020) is merely coincidental.
When and Where is Socrates.Digital™ Used?
Michael LaFargue describes Socratic reasoning as what philosophers call "inductive" reasoning. Inductive reasoning is generalizing from observations in specific concrete cases. LaFargue argues that Isaac Newton was using inductive reasoning when he derived the general law of gravity from many specific concrete observations about falling apples, and the movement of the planets.
When we contrast inductive reasoning with deductive reasoning, we note that deductive reasoning deduces conclusions from a general principle. Deductive reasoning assumes that we can be certain about general abstract principles. Euclidean geometry is built on deductive reasoning, since it starts from general abstract axioms assumed to be self-evident, such as "The shortest distance between two points is a straight line."
LaFargue points out that inductive Socratic reasoning assumes that we can be more certain about our perceptions in clear concrete cases than we can ever be about general principles. As in all inductive reasoning, when some general principle conflicts with a specific concrete observation (a "counterexample"), this is assumed to show a weakness in the general principle, which will need revising in the light of the concrete observation.
This makes inductive Socratic reasoning a better approach than deductive reasoning for solving complex problems that do not have a single “right answer.” That’s because in solving many problems, we are not certain what general principles we should start with. For example, if we are trying to determine if investing in a motel chain is a good investment, we might start by answering the question, “Which concept do you want to use to explain how to answer the question at hand?” We might answer that we wish to start with the concept of occupancy rate to determine if the investment will boost total revenue in the motel chain enough to produce the rate of return offered.
The next question we could be asked is, “What information, data, or experience do you have that supports or informs this analysis?” To answer this question, we might find that the Dundalk Inn & Suites motel has low occupancy rates and low total revenue and note this finding. This might prompt the question, “Do any assumptions underlie this observation?” After some examination, we find that we assume occupancy rate is positively correlated with total revenue. This brings us to this question, “Given these concepts, assumptions, and information, what is the most reasonable conclusion?” Using the occupancy rate concept, identifying data and information to support it, and examining the assumptions that underlie using it, we could conclude that raising the occupancy rate for the motels would raise the total revenue for the chain – possibly enough to produce the offered rate of return.
Note that we have decided that the concept, information, and assumptions naturally imply the conclusion. This is the basis for inductive thinking. We are moving from specific statements to a general conclusion. We can take this a step further by seeking the implications of the conclusion by answering this question, “What are the implications of this conclusion?” We may decide that this conclusion implies that if the occupancy rate can be raised, say, through increased marketing, then total revenue can be raised enough to meet the projected rate of return for the investment in the motel chain.
After this step, the question at hand can be asked. That is, whether to invest in the Rosebud Motel chain. In this inductive thinking process, each step is evaluated to see if it implies the next following step. These steps, called the “elements of thought” by Paul and Elder form a chain of inductive reasoning that represents a point of view. That is, the concepts, information, assumptions, conclusions, and implications form a line of reasoning that answers the question at hand -- whether to invest in the Rosebud Motel chain. Note that a different set of concepts, information, assumptions, conclusions, and implications will describe a different point of view that may form the basis for a different answer for the question at hand.