Tech & Business Innovation
Bridging deep-tech research with real-world business strategy.
Exploring AI, clean energy, space systems, and data analytics.
Toward a Human-Centric Artificial Intelligence: Diagnosis, Case Studies, and a Moral Grounding
Yoonhyuk Cho and Emmanuelle Ha
Published on August 2026
(Volume 9, Issue I), p.392- 400
Journal of Data Science for Applications (JDSA)
Summary
Corporate Optimization vs. Human Flourishing: The paper argues that AI development, tied to market forces, deepens socioeconomic inequality and homogenizes human thought under narrow metrics. Fixation on distant superintelligence risks obscures present harms. Because corporate self-regulation tends toward “ethics washing,” both democratic regulation and a robust moral compass grounded in human identity are needed.
- Case Studies (CHIA): Cambridge’s Centre for Human-Inspired AI fuses machine learning with the human sciences. EQUATE diagnoses language-AI inequality and builds community-led, low-resource language models with partners in Nepal, Uganda, and Peru. PRISM predicts damp and mold risks in social housing using vulnerability indicators, while trained officers make the final welfare decisions.
- Conceptual and Moral Grounding: The Harvard imago Dei conference and the Faraday Institute’s initiatives ground dignity in relationality and finitude, and view learning as formation of knowledge, virtue, and character. The Oxford Oath for AI Practitioners is a vocational pledge rather than a rulebook. It casts practitioners as morally disposed agents, offers religious and secular formulations, and is developed openly on GitHub.
- Conclusion: Market-led AI’s structural deficits cannot be fixed by markets alone. Democratic regulation combined with a deep understanding of dignity, virtue, and relationality can keep AI an instrument of human agency that serves human flourishing.


Impact of ChatGPT on High School Grade Point Average: A
Comparative Study by Subject, School Type, and Socioeconomic
Factors
Yoonhyuk Cho
EIET 2025, Vol 5, Issue 2, 22–28,
https://www.iikii.com.sg/journal/EIET
Educational Innovations and Emerging Technologies
Summary
This review (Cho, 2025, Educational Innovations and Emerging Technologies) synthesizes prior studies on how ChatGPT affects high school GPA, focusing on essay writing and mathematics, and how the effect differs by school type, socioeconomic background, and social context.
- Subject-level effects
-
Essay writing: Students who used ChatGPT for at least two writing assignments per week raised their GPA by about 0.31 points in one semester (Kim et al., 2024). Gains were largest for weaker writers. Risks included slightly reduced originality, formulaic essays, and plagiarism detectors failing to catch AI-written text.
-
Mathematics: A randomized trial of 1,200 students found an average gain of 0.18 points, strongest in algebra and geometry and weaker in AP-level courses (Micabalo et al., 2024). ChatGPT sometimes gives incorrect steps, so pairing it with tools like Wolfram Alpha or Desmos helps.
- School type and readiness
-
Private school students gained more than public school students (writing: 0.38 vs. 0.28; math: 0.23 vs. 0.15), attributed to better infrastructure, AI coordinators, and teacher training.
-
Well-run public school programs, such as Texas’s “AI for All,” narrowed the gap somewhat.
-
Where teachers received regular AI workshops and incentives, GPA gains averaged 0.29 points versus 0.12 where AI use was optional or unregulated.
- Conclusion:
The paper concludes that ChatGPT can meaningfully improve academic performance, but its most pressing consequence is that it may deepen existing educational inequalities. It identifies a “triple divide”:
-
Institutional: schools with infrastructure, AI coordinators, and trained teachers see larger gains.
-
Socioeconomic: higher-income, better-educated households are better at interpreting and applying AI feedback.
-
Human and pedagogical: prepared teachers and supportive peers are essential for responsible use. In under-resourced schools, AI adoption brings minimal gains and a higher risk of cheating.

🏆 GOLD AWARD @ AI-JAM US / Silicon Valley International AI Competition
When ESG is Real: Good-Practice Benchmarks, the Business Case, and the Substance Gap in ESG Management

1. Award & Recognition
-
Event: 2026 AI-JAM (International AI Competition)
-
Honor: GOLD Award
-
Location & Date: Silicon Valley, California (US) / September 2026
2. Research Topic
Whether ESG (Environmental, Social, Governance) management creates genuine financial value, or whether it functions merely as reputational window-dressing — examined through the case of South Korea's SPC Group against global ESG evidence and 2024-2026 benchmark leaders.
3. Problem Statement
ESG investing has grown to over $40 trillion in assets under management, with mandatory disclosure regimes (EU's CSRD, ISSB standards) now in force. Yet companies can publish polished ESG reports while committing serious environmental, social, and governance failures underneath — as SPC Group did (unhygienic factories, a worker's death, coerced pregnant-worker contracts, and unlawful trade-union coercion). The paper asks: does ESG only matter as disclosure, or does substantive ESG practice actually pay off financially — and how can the "substance gap" between genuine ESG leaders and greenwashers be identified?
4. Methodology
-
Case study: Mapped SPC Group's incidents onto the three ESG pillars using investigative reporting.
-
Primary survey: A 50-respondent survey measuring public awareness of and participation in the SPC boycott, and consumer knowledge of SPC's subsidiary brands.
-
Financial event analysis: Tracked SPC Samlip's share price and annual net profit around two 2022 crisis events.
-
Secondary meta-analysis: Drew on the NYU Stern Center for Sustainable Business meta-analysis of ~1,400 studies on the ESG–financial performance relationship.
-
Benchmarking: Compiled updated 2024-2026 data on ESG leaders (Microsoft, Orsted, Schneider Electric, NVIDIA, Patagonia, Best Buy) to characterize what substantive ESG practice looks like.
5. Key Conclusions
-
Awareness of SPC's boycott was high (80%) but active participation was low (20%), and the financial market signal was weak/noisy — share price dipped modestly after each crisis, but annual net profit still rose, showing that short-run market signals can mask real human and reputational costs.
-
Across ~1,400 studies, ESG shows a positive financial relationship in 58% of cases (negative in only 8%), but this benefit is concentrated in long-term horizons and crisis-time downside protection, and depends on ESG being an operating strategy rather than mere disclosure.
-
Genuine leaders (Microsoft, Orsted, Schneider Electric, NVIDIA, Patagonia) demonstrate specific, time-bound, externally verified commitments matched by real operational change — the "substance gap" is what separates them from reporting-heavy laggards like SPC.
-
Substantive ESG creates value through four channels: talent/employee engagement, brand/customer loyalty, risk reduction/resilience, and access to cheaper capital.
-
Recommendation: companies should treat ESG as genuine operating strategy — with verified, pillar-specific targets and governance alignment — rather than a communications exercise, since real market and regulatory scrutiny (and human cost) will eventually expose the gap.
🏆 GOLD MEDAL @ I CANADA / The Invention Innovation Entitled
Tackling Dying Alone in a Super-Aged Nation

1. Award & Recognition
-
Event: 11th Annual Edition of the IIIC in CANADA
-
Honor: GOLD MEDAL
-
Location & Date: Toronto, Canada / 29 August 2026
2. Research Topic
The causes of elderly “lonely death” (godoksa) in South Korea’s rapidly aging society, and the welfare-policy response needed to prevent it — updated with 2024-2025 national data and a regional risk index.
3. Problem Statement
South Korea became a “super-aged” society in December 2024 (20.3% of its population aged 65+), and elderly lonely deaths have risen sharply — 3,924 cases in 2024, up 7.2% year-over-year and roughly 33% since 2019, with 81.7% of victims being men. Korea simultaneously has the OECD’s highest elderly relative poverty rate (39.8% in 2023, about 2.7x the OECD average) and highest elderly suicide rate (46.6 per 100,000). The paper asks: what underlying causes drive this phenomenon, and what welfare-society policies could prevent it?
4. Methodology
The study updates a prior studentbuilt Loneliness Danger Rating (LDR) index — a standardized weighted sum of three regional indicators: the elderly living-alone rate and elderly suicide rate (both risk-raising), and the inverse of the welfare-facility rate (protective). The index was applied across 243 municipalities (using 2021 data) and rescaled to a common mean-100/SD-10 scale for regional comparison. The authors supplemented this with a correlation analysis linking individual-level suicide risk to poverty, living alone, age band (65-74), and regional residence, and synthesized the quantitative findings with national statistics and policy document review (Korea’s Lonely-Death Prevention Act, Basic Pension, Long-Term Care Insurance) and international case comparisons (UK and Japan’s loneliness ministries, Seoul’s “Without Loneliness” plan).
5. Key Conclusions
Mountainous, rural Gangwon Province remains the highest-risk region under the updated LDR index, confirming that lonely-death risk is measurable, regionally patterned, and driven by identifiable, overlapping factors — demographic (rapid aging and rising one-person elderly households), economic (poverty from inadequate pensions), social (eroding family-care norms), health (depression and suicide risk), and policy/regional (uneven welfare-facility access). The authors conclude that elderly lonely death is not an individual misfortune but a preventable, systemic outcome, and propose a five-pillar welfare-society agenda: (1) strengthening income security via an expanded Basic Pension, (2) scaling social-connection programs (AI/IoT safety-check calls, community outreach), (3) elderly-specific mental health and suicide prevention, (4) regional equity using LDR-style risk mapping to target underserved rural areas, and (5) stronger governance — funding and enforcing the 2021 Lonely-Death Prevention Act with a dedicated coordinating body, following the UK, Japan, and Seoul models.
🏆 GOLD AWARD @ Advanced Innovation Global Competition 2026 (Singapore)
The calculation study of the ball possession between two teams using the Markov chain and PCA Analysis

1. Award & Recognition
-
Event: 2026 AIGC (Advanced Innovation Global Competition)
-
Honor: GOLD Award
-
Location & Date: Singapore / January 2026
2. Paper & Research Summary
-
This study applies Markov chain analysis to model and predict soccer ball possession patterns.
-
Basic model: Teams are treated as states in a Markov chain, with passing accuracy determining transition probabilities. The steady-state vector of the resulting stochastic matrix yields each team's long-run ball possession percentage.
-
Positional analysis: The model was extended to four sectors (forward, midfielder, defense, goalkeeper) using passing data from FC Barcelona's 38 games in the 2020-2021 season. A stochastic matrix was built for each game, and a custom Java program (determinant and eigenvalue calculators) computed the steady-state vectors.
-
Key finding: Games Barcelona won showed notably higher midfielder ball possession (44.9%) compared to games lost (34.9%), with midfield showing the largest gap among all positions between wins and losses — suggesting a relationship between midfield possession and match outcomes.
-
PCA validation: Principal Component Analysis was used to assess how well various statistics differentiate winning games. Midfield ball possession had an eigenvalue of 7.2 — lower than major stats like total passes but higher than minor stats like corner kicks or yellow cards, indicating a meaningful (if not primary) link to winning.
-
Formation optimization: Using the concept of "ball distance" (passes needed to move the ball between two players), four formations were tested. A 3-2-3-2 formation maximized midfield possession and best matched the calculated ideal ball distribution.
-
Visualization and application: A video visualized how possession shifts sector-by-sector over the course of a match (starting 100% with forwards at kickoff). The "ideal formation" was also tested in a soccer video game, though the short match length limited observable results.
🏆 GOLD MEDAL @ EUROINVENT 2026 (Iași, Romania)
The Future trends in Digital Journalism and Traditional Journalism

1. Award & Recognition
-
Event: 2026 Euroinvent (European Exhibition of Creativity and Innovation)
-
Honor: GOLD MEDAL
-
Location & Date: Iași, Romania / May 2026
2. Paper & Research Summary
-
The paper traces journalism's evolution from print-based reporting to digital/online journalism, driven by internet expansion since the 1990s. Key points:
-
Digital journalism's advantages: broader perspectives, direct journalist-reader interaction, and reader freedom to choose sources; unlike traditional media, it offers on-demand access and eliminates print/broadcast distribution barriers.
-
Pros and cons: Digital journalism offers unlimited access, interactivity, and low distribution costs, but struggles with monetization, portability, and information overload — while traditional journalism retains advantages in analytical depth and credibility.
-
Skills required: Future journalists need technical competence (editing systems, databases, internet research), strong writing/communication skills, subject expertise, collaborative ability, and critical self-reflection.
-
Future trends: AI and machine learning will automate routine reporting and personalize content; data privacy will demand greater transparency; VR/AR will enable immersive storytelling; and audience participation will grow.
-
Legal/ethical concerns with AI: potential copyright infringement liability for journalists using AI-generated content, plus concerns about transparency, algorithmic bias, and preserving human judgment in reporting.
-
Conclusion: Despite ongoing technical and financial challenges, digital journalism is expected to keep growing, requiring journalism education to adapt by teaching digital skills alongside traditional fundamentals, while emphasizing trustworthiness amid the rise of AI-generated and synthetic media.
🏆 Uber Excellence Award 2025
Uber Teens Summer Marketing Strategy Competition : Excellence Award


Team Group 5 — Taeyun Nam, Yoonhyuk Cho, Jitae Kim, Daniel Chung
Awarded by Jinwoo Song, Country Manager, Uber Korea (Aug 9, 2025)
Project: "Uber Teen Marketing — The F1 Concept"
-
Designed a marketing strategy for Uber's new 'Teen' ride feature by leveraging the F1 movie's cultural momentum (1.5M+ Korean viewers; 800M+ global fanbase)
-
Segmented three target audiences: students (primary — convenience, trend, safety), parents (secondary — child safety, budget control), and drivers (tertiary — clear protocols, professional recognition)
-
Defined core brand values as Convenience, Safety, and Comfort, and storyboarded concrete safety features (PIN verification, GPS tracking, arrival notifications)
-
Proposed a phased go-to-market plan: F1-themed Reels/TikTok challenge → pilot in Seoul-area schools → KPI measurement → national rollout
-
Recognized by Uber Korea for "an excellent, highly feasible marketing strategy grounded in a deep understanding of teen culture and logical analysis"
🏆 GOLD MEDAL @ EUROINVENT 2025 (Iași, Romania)
AI-Driven Predictive Modeling
& Medication Support for Suicide Prevention

1. Award & Recognition
-
Event: 2025 Euroinvent (European Exhibition of Creativity and Innovation)
-
Honor: GOLD MEDAL
-
Location & Date: Iași, Romania / May 2025
2. Paper & Research Summary
-
Research Topic: AI Deep-Tech Healthcare Model for High-Risk Patient Prediction & Medication Adherence.
-
Problem Statement: Existing psychiatric systems face challenges in early suicide risk detection and low patient adherence to prescribed medication protocols.
-
Methodology: Developed an AI predictive algorithm analyzing patient biological and behavioral data patterns, integrated with a modular healthcare tracking system to monitor real-time medication intake.
-
Key Conclusion: Demonstrated that quantitative, data-driven AI monitoring significantly reduces observation gaps for high-risk patients and improves medical adherence and prevention rates.
3. Market & Business Impact
-
Establishes B2B/B2G pipelines for Digital Therapeutics (DTx).
-
Validates SaaS subscription models for hospitals and psychiatric centers with potential medical insurance coverage.
🏆 GOLD AWARD @ WICO 2025 (Seoul, Korea)
Experimental Efficiency Analysis of Hydrogen Fuel Cell
vs. Electric Vehicles

1. Award & Recognition:
-
Event: 2025 WICO (World Invention Creativity Olympics)
-
Honor: GOLD AWARD
-
Location & Date: Seoul, South Korea / July 2025
2. Paper & Research Summary:
-
Research Topic: Quantitative Energy Efficiency Comparison of Next-Gen Mobility (EV vs. FCEV)
via Small Model Car Experiments. -
Problem Statement: Lack of objective, empirical comparative data regarding real-world energy
conversion efficiency between Battery Electric Vehicles (EV) and Hydrogen Fuel Cell Vehicles (FCEV). -
Methodology: Built identical small-scale vehicle testing platforms to measure power loss during energy conversion, power consumption per distance, and cell load variation under controlled conditions.
-
Key Conclusion: Proved that while Battery EVs excel in short-distance urban driving, Hydrogen Fuel Cells (FCEV) offer superior energy density and conversion efficiency under high-load, long-distance environments.
3. Market & Business Impact:
-
Evaluates commercial feasibility for hydrogen infrastructure in heavy-duty logistics (trucks, buses, aviation).
-
Provides data-driven insights for B2G policy decision-making in global hydrogen value chains.
🏆 GOLD MEDAL @ iCAN 2025 (Toronto, Canada)
Electromagnetic Processing System
for Space Debris Mitigation
1. Award & Recognition:
-
Event: 2025 iCAN (International Invention Innovation Competition in Canada)
-
Honor: GOLD MEDAL
-
Location & Date: Toronto, Canada / August 2025
2. Paper & Research Summary:
-
Research Topic: Electromagnetic Induction and Lorentz Force-Based Space Debris Capture System to Prevent Kessler Syndrome.
-
Problem Statement: Catastrophic threats posed to satellites by high-velocity orbital debris in LEO, and the physical limitations of mechanical capture methods (nets/robotic arms).
-
Methodology: Designed a non-contact orbit alteration mechanism leveraging Electromagnetic Induction and Lorentz Force to decelerate and redirect hyper-velocity orbital fragments.
-
Key Conclusion: Verified the physical feasibility of safely de-orbiting high-speed debris into the upper atmosphere for thermal disposal without risking direct collision.
3. Market & Business Impact:
-
Unlocks B2B orbital cleanup services for satellite operators in the New Space economy.
-
Introduces ESG-aligned space environmental solutions to reduce liability and insurance losses in space commercialization.

🏆 GOLD AWARD @ AI-JAM ASIA 2025
Advanced Data Analytics
& Sabermetrics in Sports

1. Award & Recognition:
-
Event: AI-JAM ASIA Online Competition
-
Honor: GOLD AWARD
-
Location & Date: Silicon Valley, USA / August 31st, 2025
2. Research & Project Summary:
-
Project Topic: Analyzing Performance Using Advanced Data Analytics in Sports.
-
Problem Statement: Traditional qualitative sports evaluations often fail to capture complex player efficiency and game dynamics under high-stress conditions.
-
Methodology: Applied advanced data analytics and sabermetric statistical modeling to evaluate player metrics, predict outcome probabilities, and optimize tactical decision-making in competitive sports.
-
Key Conclusion: Demonstrated that data-driven predictive models significantly outperform traditional metrics in forecasting athletic performance and team strategy optimization.
3. Market & Business Impact:
-
Unlocks B2B Sports Tech analytics opportunities for professional leagues, sports management agencies, and sports betting platforms.
-
Connects advanced data analytics directly with commercial sports media and performance consulting services.