How AI is (not) Considered a Real Threat to the Academic World

How AI is (not) Considered a Real Threat to the Academic World

AI shifting 3D drawing software applications?
Artificial Intelligence (AI) is currently starting to shift and change the paradigm of 3D drawing software applications (such as Blender, SolidWorks, AutoCAD, or AutoCAD 3D) through several key mechanisms:

Generative Text-to-3D (From Text Directly to 3D Models)
Previously, to create 3D objects such as rotor rig components or shafts, designers had to create 2D sketches, extrude, revolve, and fillet them manually.
Now, generative AI allows the creation of initial 3D models simply by entering text commands (prompts), such as "3D model of a rotating shaft with pillow block bearings." The AI instantly generates a basic 3D mesh in seconds.

Image-to-3D (Converting 2D Images to 3D)
AI can analyse 2D images or photos (even 2D engineering drawings) and automatically reconstruct them into 3D models with depth.
This significantly saves time on initial drafting because users don't need to build geometry from scratch.

AI-Driven Parametric & Generative Design (Shape Optimization)
In the mechanical engineering industry, AI doesn't simply "draw," but determines the best shape based on mechanical loads.
Designers simply input parameters (e.g., vibration loads, material limits, and bearing locations). The AI then simulates hundreds of variations of the most optimal, strong, and lightweight 3D structural shape without the need for manual drawing each one.

CAD Command Automation (Co-pilot in 3D Software)
Modern CAD software is starting to integrate AI assistants. Instead of navigating complex button menus, users can simply type or speak a command: "Make an M8 bolt hole at the center of this surface", and the AI will execute the modelling steps.

Will Manual 3D Software Disappear Completely?
Not completely. AI currently serves as a co-pilot, not a full replacement. Rapid prototyping, generative shape ideation, and automation of repetitive tasks. Manual control is required for: High-precision tolerances (e.g., shaft size tolerances of 0.01 mm), highly specific finite element simulations (FEA), and manufacturing engineering drawing standards (ISO/ANSI). Designers are still needed for finishing and technical verification.

Architects under threat?
The architectural profession is not threatened with extinction, but rather is undergoing a significant role shift. AI is not replacing architects as decision-makers, but rather replacing repetitive and time-consuming technical tasks.

The AI is transforming the architectural profession and which parts cannot be replaced.
Lightning Design Exploration (Text-to-Image / 3D): The conceptual stage, which once took weeks, can now be completed in a matter of hours. AI can generate dozens of visualization options for facades, floor plans, and room ambiance based on descriptive commands.
Layout & Building Optimization (Generative Design): AI can automatically analyze floor plan layouts that maximize natural lighting, air circulation, energy efficiency, and compliance with the basic building coefficient (KDB).
Technical Document Automation: The previously manual creation of working drawings, construction details, and material calculations (Bill of Quantities) is now increasingly automated by AI integrated into BIM (Building Information Modeling) systems.

Why Can't Architects Be Replaced by AI?
Legal and Safety Liability: AI is not professionally licensed and cannot be held legally responsible if a building structure fails or collapses.
Client Empathy and Context: Understanding the client's emotional desires, local cultural dynamics, and the occupant's sense of humor or lifestyle requires human empathy that algorithms lack.
Field Supervision & Construction Reality: Real-world site conditions, negotiations with contractors, and solving emergent problems in the field still require the presence and intuition of a physical architect.
The profession that is actually threatened is not "architects," but rather architects who refuse to use AI. Architects who utilize AI will work much faster, more effectively, and focus on aesthetic value and high-level concepts.

The fields are substituted by AI
The impact of AI doesn't eliminate any field entirely, but rather reduces the workforce requirement for jobs that are repetitive, rule-based, or focused on basic data processing.

Most Affected Professions (Experiencing a Decrease in Demand)
Drafter & Junior CAD Operator: Basic 2D/3D working drawings can now be rapidly automated by AI. The industry needs senior architects/engineers who verify results more than draftsmen who start from scratch.
Plain Text Translator & Simple Editor: Translation of general documents, articles, or standard business documents is now dominated by Neural Machine Translation and LLM. Human translators are now needed more only for literary works, complex legal documents, or cultural localization.
Data Entry & Basic Customer Service: Manual data entry tasks and answering general chat support are now fully shifted to interactive AI agents.
Generic Copywriter & Content Writer: Writing basic SEO articles, e-commerce product descriptions, or drafting standard marketing copy has seen a drastic decline.
Basic Programmer / Junior Coder: Basic coding (boilerplate code), simple bug fixes, and automation scripts are now handled by AI. Human developers are required to focus on complex system architecture and business logic.

College Majors Experiencing Curriculum & Focus Shifts

Majors will not close, but their basic curricula have shifted and are required to adapt to ensure their graduates remain relevant:

Literature & Foreign Languages:
Shift: The focus of studies is no longer simply memorizing grammar or literal translation techniques, but rather on Computational Linguistics, Cultural Discourse Analysis, and Prompt Engineering.

Information Technology / Computer Science:
Shift: The emphasis on simply being able to code is starting to diminish. The focus is shifting to Software Architecture, Cybersecurity, Data Management, and AI Engineering Systems.

Fine Arts & Visual Communication Design (DKV):
Shift: Technical skills in operating design software (such as Adobe Photoshop/Illustrator) are no longer the primary selling point. The focus is shifting to Art Direction, Creative Concepts, Intellectual Property Rights (IPR), and User Experience (UX).

Accounting & Business Administration:
Shift: Manual bookkeeping and transaction recording have been automated. Graduates are expected to become Strategic Financial Analysts, Forensic Auditors, and Business Consultants.

Human Skills That Cannot Be Replaced by AI

In this era, any field you pursue will remain secure if you possess this combination of skills:
Critical Thinking & Problem Solving: The ability to validate whether AI output is correct, safe, and ethical (such as checking the mathematical equations of machine vibrations).
Interpersonal Communication & Empathy: Ability to negotiate, lead a team, and understand the real needs of clients.
Legal/Professional Responsibility: Official license (Registered Engineer, Integrated Architect, Professional Certification) that can hold legal responsibility for an engineering work.

But are there any professions or majors at universities that will be displaced by AI?
Yes, there are. The impact of AI doesn't eliminate any field entirely, but rather reduces the workforce requirement for jobs that are repetitive, rule-based, or focused on basic data processing.

Most Affected Professions (Experiencing a Decrease in Demand)
Drafter & Junior CAD Operator: The creation of basic 2D/3D working drawings can now be rapidly automated by AI. The industry needs senior architects/engineers who verify results more than draftsmen who start from scratch.
Plain Text Translator & Simple Editor: The translation of general documents, articles, or standard business documents is now dominated by Neural Machine Translation and LLM. Human translators are now needed more only for literary works, complex legal documents, or cultural localization.

Data Entry & Basic Customer Service: Manual data entry tasks and responding to general support chats have now shifted entirely to interactive AI agents.
Generic Copywriter & Content Writer: Writing basic SEO articles, e-commerce product descriptions, or drafting standard marketing copy has seen a drastic decline in demand.

Entry-Level Programmer / Junior Coder: Basic coding (boilerplate code), simple bug fixes, and automation scripts are now handled by AI. Human developers are required to focus on complex system architecture and business logic.
College Majors Experiencing Curriculum Shifts & Focus
The following majors will not close, but their basic curriculum has shifted and is required to adapt to ensure graduates remain relevant:

Literature & Foreign Languages:
Shift: The focus of courses is no longer simply memorizing grammar or literal translation techniques, but rather on Computational Linguistics, Cultural Discourse Analysis, and Prompt Engineering.

Information Technology / Computer Science:
Shift: The emphasis on simply "coding" is decreasing. The focus is shifting to Software Architecture, Cybersecurity, Data Management, and AI Engineering Systems.

Fine Arts & Visual Communication Design (DKV):
Shift: Technical skills in operating design software (such as Adobe Photoshop/Illustrator) are no longer the primary selling point. The focus has shifted to Art Direction, Creative Concepts, Intellectual Property Rights (IPR), and User Experience (UX).

Accounting & Business Administration:
Shift: Manual bookkeeping and transaction recording have been automated. Graduates are in demand as Strategic Financial Analysts, Forensic Auditors, and Business Consultants.
Human Skills That Cannot Be Replaced by AI
In this era, any field you pursue will remain secure if you possess this combination of skills:
Critical Thinking & Problem Solving: The ability to validate whether AI output is correct, safe, and ethical (such as checking mathematical equations for machine vibrations).
Interpersonal Communication & Empathy: The ability to negotiate, lead teams, and understand the client's real needs.
Legal/Professional Liability: Official licenses (Registered Engineer, Integrated Architect, Professional Certification) that can hold you legally accountable for your engineering work.

But is AI truly capable of completely replacing software applications like Onshape, Inventor, and SolidWorks?

No, AI will not completely replace or replace CAD software (Onshape, Inventor, and SolidWorks).
Instead, AI is integrated into these software as an additional feature.
The main reason is the fundamental difference between AI drawings and 3D CAD models for engineering.

1. AI Drawing vs. Precision Geometry (CAD)

Generative AI (Image/Mesh): Generates shapes based on visual approximations (pixels or a mesh of surface polygons). AI doesn't care whether the bolt holes are ∅20 mm precise or ∅20.04 mm.
CAD Software (SolidWorks/Inventor): Works based on absolute parametric geometry. Dimensional tolerances, axes of symmetry, and center distances require 100% mathematical certainty for machine components (such as bearings and shafts) to be assembled and manufactured on a CNC machine/lathe.

1. CAD Functions Aren't Just "Drawing" CAD software is the foundation of the entire manufacturing and engineering analysis ecosystem: Dynamics Testing & Simulation (FEA/CFD): Calculates residual stresses, vibration resonances, and structural loads. Manufacturing (CAM): Converts 3D shapes into machine instruction code (G-code) for real-life metal cutting. BOM & Bill of Materials: Generates material lists, bolt specifications, and production costs in a structured manner.
2. Big CAD Brands Are Adopting AI Dassault Systèmes (SolidWorks) and Autodesk (Inventor) are not threatened, but are instead incorporating AI into their products: Generative Design: You input load requirements (e.g., "withstand shaft vibration at 1200 RPM"), then the AI in Inventor/SolidWorks designs the strongest structural shape and places it directly in the CAD workspace. Text-to-CAD Copilot: In the future, you can simply type in SolidWorks: "Make a pillow block bearing seat"

What areas cannot be solved by AI?
Despite the rapid development of AI, there are fundamental human domains that AI cannot solve or replace independently:

1. Legal and Moral Accountability AI is simply software without legal status. AI cannot be sued, detained, or have its professional license revoked if a system failure occurs. Example: Medical decisions involving life-threatening risks, the signing of structural feasibility documents for buildings (architectural/civil engineering), or the approval of machine designs operating at high speeds still require the stamp of human legal responsibility (Certificate of Expertise/SIP).
2. Real-Time Empathy and Deep Emotional Understanding AI can mimic empathetic words through language processing, but it does not experience or feel emotions.

Example: In-depth psychotherapy, trauma crisis management, high-level diplomatic negotiations, and relationship counseling require intuitive awareness and mutual trust between humans that cannot be fostered by mathematical algorithms.

1. Field Intuition and Handling Unexpected Situations (Unstructured Physical Environment) AI and robotics operate based on historical data or predictable environments. However, the dynamics of the real physical world often present unpredictable variables (chaos). Example: A mechanic/vibration technician in the field identifying a strange noise in a turbine based on a burning smell, a sudden temperature change, or a vibration felt through his foot — then making an instant, improvised decision without any data in the database.
2. Discovery of New Theories and Scientific Leaps (True Scientific Breakthrough) AI excels at analyzing massive amounts of data (pattern recognition), but it lacks the curiosity or imagination to question existing laws of physics. Example: Formulating the theory of relativity like Einstein or creating a new physics paradigm that has no historical data points on the internet. AI simply extrapolates existing data.
3. Art Based on Lived Experience and Indigenous Culture Generative AI can imitate a painting style or compose a song, but the value of high-class artwork often lies in the creator's backstory (life struggles, suffering, and the socio-political context of the era). AI lacks lived experience.

Main Pattern:
AI can solve problems based on calculation, pattern analysis, and the automation of technical tasks. However, problems requiring moral responsibility, real empathy, free will, physical intuition, and creation from nothing remain the absolute domain of humans.

Do institutions such as universities or the educational world restrict AI?

Yes, educational institutions and universities restrict AI, but the forms of restrictions have changed drastically.
While at the beginning of ChatGPT's emergence (around 2022–2023), many universities attempted to completely ban the use of AI, now (2025–2026) the majority of universities worldwide and in Indonesia have shifted from total prohibitions to regulations and ethical restrictions.
Forms of AI Restrictions in Academic Environments
Restrictions in Final Projects & Scientific Articles (Theses/Dissertations/Journals)

Co-Authorship Prohibited: International publishers (such as Elsevier, IEEE, ASME) and universities prohibit the inclusion of AI as an author because AI has no legal responsibility for the accuracy of the content.
AI Use Must Be Declared: Students/researchers are required to include a detailed list of AI uses (e.g., "AI is used only to check English grammar in Section 2").
Prohibited for Primary Data Analysis: Numerical data processing, research conclusion generation, and primary analysis should not be left entirely to AI without manual validation.

Restrictions During Exams & Critical Evaluations
Colleges are returning some evaluation methods to oral examinations (trials/practicals), written exams in class without internet access (closed-book), or live case analysis in laboratories to ensure student understanding.

Automatic Similarity Detection (AI Detector & Plagiarism Tools)
Colleges are using software integrations such as Turnitin AI Detection to detect how much of the text is AI-generated. Manuscripts with too high a percentage of AI indications may be rejected or returned.

Why Don't Universities Ban AI Completely Anymore?
AI Has Become an Industry Standard: Industries (including engineering, architecture, and IT) need graduates who are proficient in using AI as a co-pilot. Banning AI completely will leave students unprepared for the workforce.
AI Is Absolutely Difficult to Detect: AI detection tools often produce false positives (false accusations against otherwise neat human writing), so universities are placing greater emphasis on academic integrity than simply relying on detectors.
Current Campus Policy: "AI-Assisted, Human-Driven"
In general, official regulations on current campuses categorize the use of AI into three categories:

🟢 Allowed: Grammar checking, brainstorming, or formatting mathematical codes/equations.
🟡 Conditionally Restricted: Using AI to summarize literature or translate languages (must be stated in the appendix/declaration).
🔴 Strictly Prohibited (Ethical Violation): Asking AI to write a thesis from scratch, falsifying simulation/experimental data with AI, or using AI during written exams.
But what about informatics majors, which rely heavily on coding? Now, people can easily acquire coding skills from AI?

In Informatics/Computer Science majors, the ease of obtaining code from AI has indeed drastically changed the educational landscape. However, this hasn't killed off Informatics majors, but rather completely changed the value of an Informatics graduate. Previously, the primary value of a programmer was memorizing syntax and writing code from scratch. Now, AI can do that in seconds. Shifting Focus in Informatics Education: Informatics education is currently shifting from simply "How to Code" (How to write code) to Systems Engineering & Critical Thinking (How to design and validate systems). From "Code Writer" to "Code Reviewer & Architect"AI often produces code that looks correct on the surface, but has security vulnerabilities, is memory-intensive, or is inefficient when accessed by thousands of users simultaneously. Informatics students are now required to become scientists who can distinguish AI code that simply "runs" from code that is safe, optimal, and scalable (easily developed). 2. Emphasis on Problem Solving & System LogicAI does not know what business problems actually need to be solved. Students are trained to analyze real-world needs, design database architectures, and determine complex system flows before the code is created. 3. Understanding Algorithms & FundamentalsThe code generated by AI comes from historical data. When new problems arise that have never been data on the internet, a deep understanding of computational mathematics, basic data structures, and algorithm complexity ($O(n)$) becomes a key that only educated computer scientists have. How Do Lecturers Evaluate Students Today? Because AI can easily code assignments, the way universities test Informatics students has changed: Oral & Defensive Exams (Code Defense): Students can use AI to code, but when submitted, they must be able to explain the logic line by line in front of the lecturer: "Why is this algorithm used?", "What if the website is accessed by 100,000 people simultaneously?" Live Coding Exams Without Internet: For basic courses, students are still tested on basic logic on a whiteboard or laptop without an internet connection. Advanced & Complex Projects: Assignments are no longer just "Create a simple calculator/CRUD application" (which is easy to create AI), but also create distributed systems, hardware/IoT integration, or network security. Conclusion The ease of getting code from AI has actually increased the standard for graduating Informatics. People who can only copy-paste AI code without understanding Their computing fundamentals will be easily displaced in industry. Conversely, Informatics graduates who understand in-depth theory and know how to control AI as a tool will become Super-Developers capable of completing projects five times faster than before.

Dr. -Ing. Salman, ST., MSc.
Teaching staff at the Department of Mechanical Engineering, Mataram University