Abstract
Artificial intelligence (AI) is transforming the world rapidly; many tasks that humans perform can be performed by robots and machines using AI. Generative Pre-Trained Transformer (GPT) is available for generating responses to user inputs (prompts). The user can chat and interact with the machines using GPT. It is known as prompt engineering. The prompt engineering is the art and science of communicating effectively with AI systems. To fetch relevant and exact information from the GPT engines is also an art, and it depends on the prompt (input) that the user provides to the GPT engine. In this work, we used one of the GPT engines, known as DeepSeek, for prompt engineering to generate simulations for the disabled students (students with hearing and speech impairments) because seeing is believing. The simulation will be beneficial for the students with disabilities and will also be useful for the students without disabilities. We present two examples of simulation from physics: (i) Projectile motion, and (ii) collision of two bodies, which were generated by prompt engineering using the DeepSeek GPT engine. The proper use of prompt engineering will make teaching and learning easier. Similar simulations can be made using prompt engineering as a teaching aid for teaching disabled students.
Keywords
Introduction
Artificial intelligence has been rapidly growing in its significance and use in daily routine. It looks like the world is in a state of transformation. Perhaps it would be the most significant technological transformation in human history. The AI tools like ChatGPT, DeepSeek, and other AI engines have created ease for human beings. AI is helping companies, businesses, and in personal assistance, content writing, analyzing research data, predicting, and forecasting in various fields, etc. Students and teachers are also benefiting from these tools. One can say AI is technology for everyone, just as it is the right of every student to learn. Generative Pre-Trained Transformer (GPT) may help a lot in teaching students with disabilities. Various engines are now available on the web for GPT. One of the engines that was used for prompt engineering in this work was DeepSeek. The communication with the AI system is prompt, Engineering. 1 The usefulness of output from AI systems depends on how precisely the prompt is crafted. 2 The prompt engineering is a recent field and rapidly growing, causing a technological transformation.3–10
Disability is an unfortunate reality; children with disabilities need special attention and supportive responses from everyone without disabilities. A compassionate and supportive response from families, communities, and government is essential for the education and healthcare of the disabled. The disabled are a part of our society; proper education and guidance can enable them to contribute positively to society. They can communicate with their peers with and without disabilities. The way of communication is different, but they know how to convey their messages. The students with disabilities are a special class of students who need special education services and tools in learning lessons and concepts. The students can be classified in various groups depending upon their disabilities, for example, blind students need different services and tools to learn than students who are Deaf and Dumb.
The researchers have also made significant contributions to make the lives of the disabled easier through engineering tools, artificial organs, and various educational accessories and systems. Mastropieri et al. carried out a literature review for science for students with disabilities. 11 Glimour et al. conducted a study to investigate the educational gap in students’ academic outcomes between students with and without disabilities. They also discussed the reason for the large achievement gap. 12 Glimour and Joseph used logistic regression on a dataset to study the association between the percentage of SWDs and teacher turnover. They found that there is a need for more training and support for the teachers. 13 Nina Yssel et al. worked on whether students with disabilities still face hindrances in their success in higher education; however, the support services and other factors have encouraged the student ratio in higher education. They interviewed college students about the student-teacher relationship and found that the students reported a positive environment. 14 In recent years, researchers focused on various issues related to disabled students, for example, mathematical proficiency of students with disabilities (Students with Disabilities), SWD, dropout intent of SWDs, mental health of SWDs, etc.15–18
Simulations are an educational aid that makes learning faster. Through simulations, not only does concept learning become easier, but it leaves a long-lasting impression on the mind, and students keep the record of learning longer than those that are taught conventionally. 19 Simulations have been used in teaching medical, engineering, natural science, and social science.20–23 Prompt engineering can also be used in constructing effective simulations. In this study, we also generated two Hypertext Markup Language (HTML) files for two simulations using prompt engineering. As mentioned in the Abstract, the simulations were made specially for the students with hearing and speech impairments, which includes both the groups of students with partial & complete hearing loss and speech impairments. The simulation will be beneficial for the students at higher secondary levels.
Method and material
Prompt engineering means effective communication with AI systems. Effective communication depends on the design, selection of accurate prompts (input), and the optimization of inputs. Prompt engineering can perform various tasks, e. g., code generation, problem solving, language translation, content writing, assistance in writing, research assistance, and decision support. The effectiveness of the prompt engineering depends on the instruction and crafting of questions. The proper information in the form of statements and precise queries fed to the AI system would produce more relevant, useful, and accurate output. The following algorithm was employed in generating the most accurate and relevant output. (1) Make a brief sketch of your Task. (2) Craft your question precisely and feed proper prompts (input). (3) Generate output. (4) Check if you get the desired results. (5) If not, modify prompts according to the requirements and iterate the process to refine the output.
Simulations and animations are additional tools to teach students beside conventional teaching. Websites like https://phet.colorado.edu/ are great resources to make physics understanding easy. The simulations developed in this work are in addition to these resources. Our simulations are a bit different from these sources in the sense that they can be run without any platform or website. The simulations cover topics that are taught at higher secondary schools or college levels. The simulations were made to keep the students with disabilities in mind.
Result and discussion
Prompt Engineering has been employed to generate simulations using HyperText Markup Language (HTML). Precise prompts were developed as input for AI systems to yield the best response. Two different physics simulations have been generated by DeepSeek GPT using specific prompts (input). One of them describes projectile motion and also compares projectile motions on Earth and the moon. The second simulation describes the collision of two bodies, both cases (i) head-on collision, and (ii) oblique collision have been covered.
Simulation of projectile motion
A simulation of the motion of a projectile is generated using DeepSeek GPT. To get the best result in the form of a simulation, two different prompts were given to the GPT.
Prompt I: “HTML file: simulation of projectile motion for various angles and velocities.”
This prompt generated a projectile simulation where the motion of the projectile can be controlled with two sliders managing velocity and angle of projection. The Fire Projectile button fires a projectile (spherical object), with the velocity and angle of projection selected through sliders. The projectile trajectory is generated, and the projectile follows it by moving along the curve. In Figure 1, the simulation generated by the first prompt is shown. There are two different trajectories shown in blue (low trajectory) and red (high trajectory). The Fire projectile button was pressed for two launching angles (30° and 60°). The simulator also generated values of the range, max, height, time of flight, and impact velocity after the projectile strikes the ground. Figure 1 shows the values of the projectile projected at an angle of 30°. This example also shows that the range is the same at two complementary angles. Larger angles of projections have higher trajectories and longer times of flight. Simulation of projectile motion generated by GPT using prompt I.
Prompt II. Do the same for projectile motion on the moon, and show both in the same window, but with their vertical axes separated on the left and right sides.
The second prompt was given to generate the motion of the same projectile on the moon. The students can learn the difference in the motion of two projectiles, projected from the surfaces of Earth and the Moon. They can easily understand the role of gravity at the two surfaces. Figure 2 shows a screenshot of the simulation, two projectiles: blue (on the left) and yellow (on the right) are projected on Earth and the moon, respectively. After completing the trajectories by both the projectile, the simulator also calculates and shows the range, max. height, time of flight, and impact velocity for both the projectiles. The comparison of the results would give students an insight into the motion under different gravities. Simulation of projectile motion generated by GPT using prompt II.
Simulation of the collision of two bodies
In the second example of prompt engineering, a simulation of the collision of two bodies is constructed using two prompts (III and IV as inputs) by GPT.
Prompt III. HTML file: Simulate the oblique collision of two objects
The above prompt was given to simulate the collision of two spherical objects having specific velocities. The GPT constructed a simulator that showed the head-on collision of two bodies.
Prompt IV. Both masses should have mass and velocity control.
The velocity of one of the objects could be controlled by sliders, but the initial simulator generated by GPT had no control over the velocity of the second object. Therefore, another prompt (IV) was given to include a slider for the second object. In Figure 3, the screenshot of the final simulator is shown. There are three sliders for each of the objects, which allow the user to change the mass, velocity, and angle with respect to the horizontal axis. For the head-on collision, both objects should move along the same line; therefore, one of the angles is 0° and the other has an angle of 180°. For oblique collisions, the user can set specific values of the angles for the two objects. Various cases of head-on and oblique collisions between objects of various masses can be demonstrated through these simulations. Simulation of the collision of two bodies generated by GPT using prompt III and IV.
Conclusion
Artificial Intelligence has been transforming the world; it is becoming inevitable to keep oneself isolated from this transforming world. AI systems are replacing humans and also increasing the efficiency and speed of the tasks assigned to them. To receive the best from AI systems, one should be well-versed in communicating effectively with these systems. It is done by prompt engineering; proper prompts (input) to the AI system improve accuracy, quality, usefulness, and relevance of the AI outputs. We have demonstrated in two examples a communication with the AI systems. The prompts were given and reiterated in both examples to construct the desired task. In both examples, the AI system generated a simulation in HTML programming, with a single and well-crafted prompt; the second prompt was used to refine the output. The simulations were generated to keep in mind that these would help students with disabilities, especially Deaf and Dumb students. In the first example, an AI system (GPT, DeepSeek) generated simulated motion of a projectile on Earth and the Moon to compare the effect of gravity on the motion. In the second example, an AI system generated simulated motion of two bodies in head-on and oblique collisions. Both simulations are impressive and work well according to their mathematical relations, and would be beneficial in teaching students with disabilities as well as ordinary students. Prompt engineering can also be used in other subjects to construct teaching tools to help learning by students with disabilities.
Footnotes
Funding
The authors received no financial support for the research, authorship, and/or publication of this article.
Declaration of conflicting interests
The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.