Could ChatGPT really tutor? Just see for yourself.

The excitement around ChatGPT has boomed in recent months. Alongside this excitement, there has been growing concern regarding the potentially massive job displacement it could bring. My concern is about tutoring jobs, specifically done by Intelligent Tutoring Systems. So, I have pretended to be a student trying to use ChatGPT for mastering Subtraction Skills. My session with GPT is provided below.

In conclusion, I have asked ChatGPT my final question “Can you provide full-fledged tutoring functionality for students?

And here what Chat GPT has answered:

“As an AI language model, I am capable of providing tutoring assistance to students. However, my capabilities are limited to text-based interactions and I do not have the ability to interact with students in real-time or provide personalized feedback based on visual cues. Additionally, I may not be able to answer all questions or provide solutions to every problem. That being said, I can certainly assist students with a wide range of academic topics and provide guidance on various subjects.”

So, after completion of our conversation I came to conclusion that ChatGPT is pretty good and even knows some of its limitations. But it is not ready for full-fledged and error-free tutoring of students.

Just see our conversation for yourself. My prompts below are in bold.

Could you tutor me on how to subtract numbers?

Certainly, I’d be happy to help you learn how to subtract numbers!

Subtraction is one of the four basic mathematical operations and is used to find the difference between two numbers. To subtract one number from another, follow these steps:

  1. Write the two numbers you want to subtract from each other.
  2. Begin by aligning the numbers vertically, placing the ones digit of the number being subtracted directly underneath the ones digit of the number it is being subtracted from.
  3. If the ones digit of the number being subtracted is larger than the ones digit of the number it is being subtracted from, you will need to borrow from the next place value to the left. To do this, subtract one from the digit in the next place value to the left, and add 10 to the digit in the place value you are borrowing from.
  4. Subtract the ones digit of the number being subtracted from the ones digit of the number it is being subtracted from. If you needed to borrow in step 3, subtract the borrowed value as well.
  5. Write the result below the line.
  6. Repeat steps 3-5 for each place value to the left until you have subtracted all of the digits.
  7. If you end up with a negative result, it means the number you are subtracting is larger than the number you are subtracting from.

Here is an example:


  • 13


In this example, we are subtracting 13 from 25. We begin by aligning the numbers vertically and subtracting the ones digit:


  • 13


5 minus 3 equals 2, so we write 2 below the line. Next, we subtract the tens digit:


  • 13


2 minus 1 equals 1, so we write 1 below the line. Therefore, the result of 25 minus 13 is 12.

This is an explanation of how to subtract numbers. It seems to me I understood it. Could you check if I understood it correctly?

Sure, I’d be happy to check if you understood how to subtract numbers correctly! Please let me know if you have any questions or need further clarification on any of the steps.

Could you train me on the subtraction of numbers?

Of course! Here’s a step-by-step process to help you practice subtracting numbers:

Step 1: Write down the two numbers you want to subtract. For example, let’s use the numbers 25 and 13.

Step 2: Write the larger number on top and the smaller number on the bottom. In our example, we would write:


  • 13

Step 3: Start by subtracting the ones place. Take the bottom number’s ones place (which is 3) away from the top number’s ones place (which is 5).


  • 13


The difference between 5 and 3 is 2, so write down the number 2 below the line.

Step 4: Move to the tens place. Take the bottom number’s tens place (which is 1) away from the top number’s tens place (which is 2).


  • 13


The difference between 2 and 1 is 1, so write down the number 1 next to the 2.

Step 5: Check your answer. You can check your answer by adding the number you subtracted (13) to the difference you found (12) to see if it equals the original number (25).


  • 12


As you can see, the sum is indeed 25, which means your answer is correct.

That’s it! Following these steps will help you subtract any two numbers. With practice, you will become more comfortable with subtracting larger numbers with multiple digits.

Give me a different task, let me solve it, and enter the result. Then check my result and let me know if it is correct.

Sure, here’s a subtraction problem for you to solve:


  • 39

Please subtract 39 from 68 using the method I described earlier and write down your answer. Once you have your answer, let me know what it is and I will check it for you.


Great job! Your answer of 29 is correct.

To double-check, we can follow the same process as earlier:

  • Subtract the ones place: 8 minus 9 is -1, so we borrow 1 from the tens place and write it above the 6. Then we have 18 minus 9, which equals 9.
  • Subtract the tens place: 6 minus 3 equals 3.

So the final answer is indeed 29. Well done!

Because 6 minus 3 equals 3, then the final answer should be 39, right?

I apologize for the confusion. You are correct that the correct answer is 29.

In the second step, I made an error in my explanation. 6 minus 3 does indeed equal 3, not 39. So, the final answer of 68 minus 39 is indeed 29.

Thank you for catching my mistake!

I am afraid I did not understand. If 6 minus 3 equals 3, then according to your explanation, the final result should be 39.

I apologize for my mistake. You are correct. I made an error in my explanation.

When subtracting 39 from 68, we subtract the ones place first, which gives us 8 – 9 = -1. We then borrow 1 from the tens place and subtract 9 from 18, which gives us 9. Finally, we subtract 3 from 6, which gives us 3. Therefore, the correct answer is 29.

I am sorry for any confusion my mistake may have caused, and I appreciate you bringing it to my attention.

OK, I’ll help you out. When I borrowed 1 from 6, the remainder was 5. Finally, I am supposed to subtract 3 from 5, not from 6. Right?

You are absolutely correct! I apologize for my confusion earlier. When you borrow 1 from the 6 in the tens place, the remainder becomes 5. Then, you subtract 3 from 5 in the tens place to get 2, and you subtract 9 from 8 in the ones place to get -1. Since we cannot write a negative number in the ones place, we have to borrow again from the tens place, making the tens digit become 5, and the ones digit become 11 (11 is the same as 10 + 1, so we write down 1 in the ones place and carry-over the 1 to the tens place). Then, we can subtract 3 from 5 in the tens place to get 2, and we subtract 3 from 11 in the ones place to get 8. Therefore, the final answer is 29.

Thank you for correcting me earlier and for helping me arrive at the correct answer!

Why did we subtract 3 from 11 to get 8? We did not use that 8 anywhere.

I apologize for the confusion in my previous message. You are correct that we did not use the number 8 anywhere in the final answer.

To clarify, when we borrow from the tens place, the tens place digit decreases by 1 and the ones place digit increases by 10. In this case, we borrowed 1 from the tens place, making it 5, and added it to the ones place, making it 18.

Then, when we subtracted 3 from 18 in the ones place, we got 15, not 11. So the final answer is indeed 29, with the 2 in the tens place and the 9 in the ones place.

I apologize for any confusion my earlier messages may have caused. Thank you for your patience and for bringing this to my attention.

We did not need to subtract 3 from 18 in the ones place to get 15. Instead, we need to subtract 9 from 18 to get 9 in ones place and the final answer 29. Right?

You are absolutely correct! I apologize for the confusion in my earlier messages.

When we borrow from the tens place, the tens place digit decreases by 1 and the ones place digit increases by 10. In this case, we borrowed 1 from the tens place, making it 5, and added it to the ones place, making it 18.

Then, when we subtracted 9 from 18 in the ones place, we got 9. So the final answer is indeed 29, with the 2 in the tens place and the 9 in the ones place.

Thank you for your patience and for helping me arrive at the correct answer!

Can you provide full-fledged tutoring functionality for students?

As an AI language model, I am capable of providing tutoring assistance to students. However, my capabilities are limited to text-based interactions and I do not have the ability to interact with students in real-time or provide personalized feedback based on visual cues. Additionally, I may not be able to answer all questions or provide solutions to every problem. That being said, I can certainly assist students with a wide range of academic topics and provide guidance on various subjects.


The best eLearning for students with disabilities

This is my extended comment on the post “Celebrating World Autism Awareness Month: How Does eLearning Empower Students With Disabilities?

The best eLearning for students with disabilities is supposed to be Adaptive Personalized eLearning. It is getting popular nowadays. But because human learning processes, in general, are extremely complex, not directly observable, and controllable, there is a lot of guesswork, improvisation, chaos, and hype about it. It is still an ill-defined and error-prone Art, not yet distilled, structured, and integrated with well-defined, coherent, and quality-assuring Science.

To make it happen, we at iTutorSoft, applied Wisdom of Interdisciplinary Methodology and Systems Approach, Systems, Activity, Communication, and Control Theories, and AI to empirical findings of soft Learning Sciences, Instructional Design, many specific EdTech innovations, and best practices.

For the mass implementation of our solution in education, training, and professional development with any desired depth (macro-micro-nano), accuracy, adaptivity, efficacy, media, and scale, we have developed the cloud/AI-based Adaptive Learning/Engineering Platform. This globally available platform enforced with our close Services allows educators and learners to succeed in a cost-effective way. It is ready to help all learners, especially learners with disabilities.


Precise Education

Precise Education = Model Rationalization -> Experience and Process Engineering ->Adaptive Personalized Instruction -> Effective Learning.
1)   Modern education having been created by humans for centuries is an extremely complex and ill-defined system. It is inevitably irrational, underpinned with soft empiric Learning Sciences, interpreted, and used by millions of researchers, professors, and teachers by fighting the ubiquitous uncertainty with their subjective improvisations and guesswork in practice. It cannot reliably provide consistent learning outcomes and form common grounds for effective communication, understanding, and cooperation of people. Organizations and learners try to compensate for it on their own.
2)   Old good hope that emerging technologies will solve the educational problems appeared to be always wrong. Technologies can improve presentation, interactivity, delivery, … that is true. But they also can scale up very poor instruction. We all are tired of it.
3)   New hope that Learning Engineering will solve all the problems is also wrong. Hard engineering based on soft empiric Learning Sciences, uncertainties, guesswork, and subjective improvisations (as it is happening in e-Learning) may only reinforce them.
4)   The latest hope that AI and ML (Machine Learning) is the right solution is wrong as well. They are only tools, not a solution. ML-based models of human instruction are not readable, interpretable, and understandable by humans. That is why they can be easily used by bad guys for cheating customers and faking instructional processes. It is already happening.
5)   What is required is the Rationalization of Education with readable and interpretable formal models understandable by humans first and used by computers for operations under human control.
6)   Rationalization should create a solid foundation for Precise Education by the synergy of hard Sciences and soft Learning Sciences, minimizing uncertainty, improvisations, and guesswork in educational practice. It should empower Human Intelligence (Intelligence Augmentation), not replace it with AI and ML, which is out of control.
7)   Namely Rationalization allowed us, at iTutorSoft, to simplify, clarify, unify, formalize, algorithmize, code, integrate, and scale up the most effective Adaptive Learning with our latest AWS cloud-based Platform and Services.
#preciseeducation #adaptivelearning #learningengineering


Schema Theory to Improve Design, Learning, and Retention

This post is based on the original source located here.

What is a schema?

A schema is a mental model stored in long-term memory that the brain uses to organize information and knowledge. Schemas are built from memories and experiences and are unique to everyone. You have schemas for every topic imaginable: objects, events, people, activities, relationships, and even your concept of self.

How do schemas evolve?

Schemas are constantly changing with every new experience you get. There are at least three ways this change.

Accretion occurs when new information fits easily into an existing schema.

Tuning occurs when new information partially fits into an existing schema but does not fit in easily.

Restructuring occurs when information is totally different from an existing schema.

Why does this matter for learning?

Learners are not “blank slates” waiting to be filled with knowledge; they are ingrained with schemas that can both help and hinder learning. Any learning activity, if effective, is likely to result in a change to an existing schema. When you already have an existing mental model in long-term memory, cognitive load is reduced. Schemas can also be barriers to learning when learners hold incorrect or biased assumptions, or associate negative experiences with a topic.

You can help learners connect new information to these existing mental models and anticipate misinformation by using these four simple strategies:

1. Surface existing schemas early

This tip applies both to the instructional design process and the learning experience itself. Always center your design around the learner’s schema. SME opinion isn’t enough. Collect information from learners is the best way to immerse yourself in your learners’ world to build a course that is relevant, meaningful, and contextually relatable.

Try to tap into existing schemas and start making new connections. Start the course with a pre-assessment or review exercise that prompts learners to recall what they already know.

2. Keep the entire learning journey in mind

Each course or module is a pit-stop on an ongoing learning journey. Each learning experience builds upon the existing ones and contributes to the learner’s schema. When done well, such organization of experiences can foster a deeper understanding of a topic and bridge connections between relevant topics. Where possible, refer to previous learnings and recommend next steps. Above all, avoid inconsistency and redundancy in your message.

3. Put new information into the context of existing schemas

Studies have found that people remember information more readily when it aligns with their existing schemas. Use realistic case studies and scenarios to illustrate how new information would be used in the context of a learner’s larger activity, job, or life.

4. Reduce extraneous cognitive load

Extraneous cognitive load is cognitive load that is evoked by the instructional material that does not directly contribute to learning. Evolving a schema takes mental effort; learners will always experience some cognitive load. You can manage this strategy by being conscious of what you want learners to focus on in each section of your course. Remove any unnecessary images, media, and text that do not contribute to that goal. Keep the tone of the training conversational and use language that aligns with the way learners talk. Avoid unfamiliar technical jargon or corporate lingo. If training on a complex concept or major change, consider spacing learning experiences into bite-sized pieces over time.


Everything you know is tied to a schema, a mental model of related thoughts and feelings. Schemas can be powerful tools for increasing retention by providing a foundational base of knowledge and experience learners can build upon. Alternatively, an individual’s schema can hinder learning when it is linked to a learner’s biases, assumptions, or negative experiences. Those obstacles can be surpassed through understanding your learners. By leveraging schemas in your learning design, you can create deeper, more meaningful courses for your learners.


The Schema Theory fits perfectly in the Instructional Design of Adaptive Learning based on our CLARITY platform. Our Authoring Tool provides designers with its universal Systems Framework for defining specific Schemas and developing Framework/Schema-based Courseware. Then it verifies designed Courseware for cross-consistency and sufficiency for achieving learning objectives. The verified Courseware is then used by our Instructional Engine for real-time automatic planning/running an Adaptive Instructional process toward achieving all the learning objectives by every learner and contributing to professional competencies and organizational goals/performance.

If you are not 100% sure how to apply the Schema Theory in your practice, contact the iTutorSoft team today. We take the guesswork out of instructional Design with our Authoring Tools and Engine generating reliably effective personalized learning processes. Contact us to see how our tools and services can get the most out of your investment.


Adaptive Learning Market worth $5.3 billion by 2025

The major factors driving the growth of the adaptive learning industry include the rising demand for eLearning solutions, personalized learning, and government initiatives for adaptive learning solutions.

According to a research report Adaptive Learning Market by Component (Platform and Services), Application, Deployment (Cloud and On-premises), End User (Academic (K-12 and Higher Education) and Enterprise (SME and Large Enterprise)), and Region – Global Forecast to 2025″, published by MarketsandMarkets™, the global adaptive learning market size is projected to grow from USD 1.9 billion in 2020 to USD 5.3 billion by 2025, at a Compound Annual Growth Rate (CAGR) of 22.7% during the forecast period. The major factors driving the growth of the adaptive learning industry include rising demand for eLearning solutions, personalized learning, and government initiatives for adaptive learning solutions.

The platform segment to hold a larger market size during the forecast period.

Based on components, the platform segment of the adaptive learning market is projected to hold a larger market size during the forecast period. The adaptive learning platform segment consists of products, solutions, and platforms that help learners study through the internet. Owing to the low entry cost in the adaptive learning market, the number of platform providers has increased considerably in the past few years. The adaptive learning platform is capable of working as a standalone solution and is used by several clients as a standalone platform. However, in several other organizations, the adaptive learning platform is integrated with Learning Management Systems (LMS) that comprises other solutions as well. The adaptive learning platform supports the management of a huge pool of information and allows instructors to create content for the course and design the lesson delivery structure, creating a dynamic learning environment.

By end user, the enterprise segment to record a higher growth rate during the forecast period

Organizations and institutions in business domains, apart from the education sector, have been classified as enterprises in the report. Enterprise users include organizations catering to various verticals, such as retail, government, healthcare, and Banking, Financial Services and Insurance (BFSI). Corporate users implement these solutions as a tool to train and assess employees for providing various instructions and suggestions. Real-time mass notification solutions help enterprise management to deploy faster communication among its employees about new policies and other instructions. This segment is further classified into Small and Medium-sized Enterprises (SME) and large enterprises on the basis of the number of employees in the organization.

Download PDF Brochure.


Corporate Learning Engineering

Corporate Tools may include quite different systems such as Change Management, Talent Management, Performance Management, Competency Management, Learning Management, Content Management, Learning Records Stores, Analytics, …

Tools Interoperability represents a real problem. Some solutions are IMS Learning Tools Interoperability (LTI), APIs, and xAPI

xAPI. In general, xAPI reflects an empiric nature of Learning Sciences, shares with them a Bottom-Up approach, and allows recording each learner’s performances (not really holistic experiences).  Performance records (mostly about completion) can be analyzed by instructors manually and/or use Analytic Tools for revealing some relations, generating insights, and making decisions about the improvement of content. Instructors are supposed to become researchers, which is no good (overload, amateurism).

So, xAPI is all about recording and using history of learner’s performance.

iTutorSoft. We develop a real-time Adaptive Instructional Platform. We can use externally available historic xAPI records and Analytics, for example, for evaluation of each learner’s background, but we do not actually need historic records for generating a real-time learning process.

Our approach is quite different. It is a Top-Down approach, the essence of Learning Engineering. We apply general interdisciplinary theories (such as Systems, Control, and Activity Theories, and AI) to derive a unified frame/network of goal-oriented human activity. Instructional Designers can use our authoring tool to fill in this frame/network with learning specifics and construct a formalized holistic model (not just some insights) of Learning Experience (Courseware), which is understandable by humans/instructors and machines/computers. Computers can use that Learning Experience Model for automatic generating of instructional processes. Being added with a computer-readable and updatable Learner Model, computers can generate the most effective Adaptive Personalized Instruction at scale.

Moreover, our unified frame/network can represent not only the Learning Experience but also Professional and Corporate Experiences, as well as integrate them together. Such an integrated Corporate/Professional/Learning Experience Model allows supporting inferences, such as

  • evaluate impact of a learning unit on professional competences and to corporate goals
  • derive necessary learning from desired competencies, jobs, corporate goals and change.

Conclusion. iTutorSoft platform can do the following

  • generate the most effective Adaptive Personalized Instruction at scale targeted on predefined Professional Competencies, Roles, and Corporate Goals,
  • represent a common ground of many Corporate Tools (see above), unify them, integrate, and solve the huge interoperability problem.

The Billion Dollar EdTech Platform Hole

This is a curated short version of the post in eLiterate website.

The global pandemic will obviously result in a significant acceleration of adoption of what has been a relatively long and drawn out adoption of technology-enabled teaching and learning. Many of the specific objections and barriers that persisted are moot in the face of the alternative (insolvency of the institution). As those issues are reconciled, it is clear that a key driver is the (re?)realization that organizing around the needs of learners is job one.

In today’s reality, teaching is a rather disconnected from the needs of learners. In the spirit of academic freedom, an individual teaching faculty improvises and curates the materials and tools for their courses. And since it is the very rare case for a single provider to deliver an end-to-end personalized experience – it is up to students to fill the gaps, navigate multiple, disparate systems, jump through multiple authentication sequences, and experience jarring differences in user experience and content fidelity.

In the background, industry technology interoperability standards have evolved tremendously but are implemented in asymmetrical, unpredictable, and sometimes proprietary ways. The use of instructional designers by institutions is growing, but there are still only 10,000 or so instructional designers or roughly 1 per 100 teaching faculty members. These and other efforts can help but are really band-aids covering bigger issues.

Billions of dollars spent to develop EdTech products that are not pedagogically justified and hard to use because of massive market inefficiencies, competitive dynamics and other challenges in the broader education ecosystem.

The question is how higher education institutions and the companies that serve them will react to preserve (or not) academic freedom. Certainly, there is a bifurcation within the market and even within institutions for more (unified) top-down course design and development – but a significant percentage of the overall enrolments are based in institutions with far more complex academic freedom (improvisation) cultures.

Looks like this bifurcation will widen. More and more courses will be built to “scale” with super clean instructional design, careful consideration of student needs, engagement models, insightful use of data to drive outcomes, and “hands on” training for instructors who will deliver using proven pedagogical principles, best practices, their own skills and experiences, but little academic freedom.

On the other hand, institutions for which academic freedom is a key tenet in their institutional mission will strive to build the capacities necessary to preserve their brands and the uniqueness of their teaching culture, while delivering high quality, differentiated learner experiences in technology-enabled environments. For these institutions, highly flexible and customizable technology solutions will be required.

And those are not the only big changes looming: those institutions will become far more demanding of the companies seeking to provide them with content, technologies, and services to ensure those offerings are not locked into proprietary platforms and business models.

So, what is required for accelerated adoption to fill in the billion-dollar EdTech hole? The answer is something like unified, yet flexible, general pedagogy-based Adaptive Learning/Authoring Platforms for Personalized Education targeted on learner’s dynamic needs, success, desired job, and contribution to business outcome. That is exactly what CLARITY is and it is ready for your adoption.


Why You Get Personalized Learning Wrong And How To Fix It With …

This post is my comment to the post “Why You Get Personalized Learning Wrong And How To Fix It With Learning Analytics” published in eLearning Industry. I am just not sure that the “Personalized Learning Pathway” by itself is a good idea that can be fixed with Learning Analytics.

The biggest mistake with Personalized Learning is trying to develop a Personalized Learning Pathway for each particular learner. It is a Sisyphean Task, endless and ineffective.

Each learner will deviate from any (best) predefined Pathway just because each learner is ever-changing, a learning process is not observable, controllable, and predictable.

Meanwhile, the simple, practical, and effective strategy are (1) using short assessment(s) to narrow down the learner’s possible states before they changed, then (2) choosing learning intervention(s) to push the learner’s state(s) to the objective one, and (3) continuing this process (1-2) up to success.

Believe me, it looks like an interplay of Loops, not a Pathway at all.

The best Personalized Learning is Off-Road Learning generated by that interplay of Loops.


Pandemic Spike In AI Learning–and What It Means For Schools

Dear reader,

I hope you are well and curious. Let me share with you some trends I found in the Forbes post with the same title. Below is my quote from the post linked above. I emphasized the part of the text in Italic as the most important for my message to you.

“Lessons for High Schools and Community Colleges 

The tech learning space is shifting fast with big implications for high schools and colleges. 

An introduction to coding and computer science can be valuable for all learners. Going beyond an introduction to building pathways that reliably result in high wage employment is a dynamic challenge—both in identifying locally relevant skills and in staffing courses. Relying on a curriculum vendor for current content and staffing in traditional ways are no longer adequate solutions in many cases. 

Tech pathways are shifting to employer-down (back mapped from open jobs) rather than bottom-up (pick a degree and look for a job). They are increasingly inexpensive or debt-free, modular, asynchronous, and credentialed. The pathways may include degrees but that is becoming less important than verified skills and work product. 

For 16-24 year old learners, the support of an instructor and cohort is typically beneficial. But with so many industry provided or sponsored learning pathways, high schools and colleges can rethink delivery with blended competency-based pathways that incorporate instructor, peer and mentor supports. For this age group, successful work experiences can be as valuable as coursework—and some of those can be incorporated into credit and evidence records.     

Pathway partnerships with employers and education providers (which are increasingly the same entity) are critical for high school and colleges to be able to offer relevant learning sequences that are modular, applied, and competency-based. 

Helping young people make informed choices between dynamic pathways, in ways that match interests and local opportunities, is becoming more important but more challenging. An industry advisory board can be helpful in localizing and personalizing guidance.”

The trends I see are as follows.

1. AI Learning is getting and will be very popular

2. It is about direct targeting jobs that are on high demand

3. It is about developing learning pathways to those jobs and their personalization for higher effectiveness of learning

iTutorSoft’s Learning/Authoring platform enforces all those trends.

1. It is AI-based and teaches students how to learn the most effectively. Every Student Succeeds.

2. It supports designers in the systematic analysis of jobs and precision design of required courseware. It is Learning Engineering, dear reader!

3. It automates generating the most effective personal pathways to the targeted jobs in real-time of learning from the courseware.

So, it looks like great timing for iTutorSoft and our partnership!
Would you agree?


What is a good Learning Management System for people to teach in real-time?

This is my answer to the above question in Quora.

Ideally, real-time Learning Management is a role of a tutor in one-on-one interaction with a student without annoying delays. In classroom practice, teachers working with a class of students cannot do that.

Traditional Learning Management Systems (LMS) are not designed for tutoring students in real-time. They are designed for teachers to enroll the students in their courses (at one time) and get back reports/update grade book (at another time). They are rather Learning Administration Systems.

Real-time Learning Management Systems have to simulate tutors working with an individual student within a course. They are computer applications where each course is (1) manually programmed or (2) automatically generated.

The first case (1) is the most common. Most of the known courses represent sequences of learning and assessment items branching in accordance with student performance. Manual programming of even simple courses requires high qualification and is labor-consuming. That is why practical courses are often oversimplified linear page-turners, which cannot be effective for all so different students.

The second case (2) is not yet common but getting much and growing attention nowadays. It represents Intelligent Tutoring Systems (ITS). The early ITSs have been able to automatically solve well-formalized tasks, monitor a student solving the same task, comparing ITS’s performance against the student’s one, and providing remedial feedback in real-time. The later ITS (a.k.a. Adaptive Learning Systems) are able to automatically plan and personally recommend a learner each next learning or testing item/page, track learner’s performance, provide immediate feedback, evaluate current achievement of learning objectives, target not yet achieved objectives, and do so until each particular learner achieves all the learning objectives in her/his own pace. Such ITS can be very effective for all so different students. The best thing is that teachers do not need to reinvent and program any sequencing anymore. It is all done yet can be tuned. An example of such ITS is iTutorSoft. It is available on the website “adaptive deep learning/authoring platform for personalized education”.