Welcome to the 75th edition of our Newsletter EdGenie's 📜 Sunday Scroll...

Every Sunday I send out actionable tips, tricks and real-world application insights from my 15 year experience coaching students to achieve As and A*s in their Economics A Levels via EdGenie.

Hey Genies, 👋

You won’t get marks just for telling the examiner what might happen.

It’s not enough to say, “An increase in indirect taxes will reduce consumption of alcohol.”

That’s just a statement.

Where’s the proof? Where’s the specifics? Where’s the evidence for evaluation?

You need to show them.

You need to prove it.

You need to be specific.

Here’s how:

📉 Use Diagrams – Show the tax burden on consumers and producers. If demand is inelastic, most of the tax is passed on to consumers (like in the diagram above).

🖥️ Efficient AI Training:  DeepSeek trained its R1 model using just 2,048 Nvidia H800 GPUs and $5.6 million, a fraction of the cost incurred by OpenAI and Google for similar models.

🏆 Breaking AI Barriers:  Despite limited resources, DeepSeek developed a 671 billion parameter AI model, proving that innovation is possible without extensive funding.

🚀 No AI Monopoly: AI researcher Ritwik Gupta suggests that AI capabilities are not exclusive to tech giants, as new entrants can rapidly catch up.

⏳ Second-Mover Advantage:  Gupta explains that while early AI developers make costly investments, later entrants like DeepSeek can replicate success more efficiently.

🇨🇳 A Proudly Chinese AI Lab:  DeepSeek prides itself on being a 100% local Chinese company, staffed with PhDs from Peking, Tsinghua, and Beihang universities rather than overseas-trained talent.

📈 Self-Funded Success:  Founder Liang Wenfeng funds DeepSeek using profits from his quant trading hedge fund, High-Flyer, allowing the company to focus on research rather than commercial pressures.

🏛️ A Research-First Approach:  Unlike traditional AI firms, DeepSeek operates like a research lab, prioritising breakthroughs over monetisation.

🎯 Maximising Limited Resources:  DeepSeek engineers have found ways to optimise China’s restricted AI chips, ensuring high performance despite US export bans.

💰 Top AI Salaries:  DeepSeek and TikTok-owner ByteDance offer some of China’s highest AI salaries, attracting top-tier engineers.

🏢 University-Like Culture:  DeepSeek’s offices in Hangzhou and Beijing resemble academic research centres, fostering innovation and deep AI exploration.

Q. Discuss the extent to which barriers to entry remain high in the AI industry, despite DeepSeek’s ability to enter the market with limited resources.
A. While DeepSeek’s success suggests that barriers are lowering, AI remains capital-intensive with high sunk costs and economies of scale. The presence of patents, data access restrictions, and regulatory constraints means true contestability is still limited. However, technological advancements and second-mover advantages may reduce these barriers over time, increasing market accessibility.

Q. Evaluate whether DeepSeek’s emergence supports Schumpeter’s concept of creative destruction in AI markets.
A. DeepSeek challenges the dominance of OpenAI and Google by demonstrating that AI innovation is not exclusive to major firms. This supports Schumpeter’s theory that new, innovative firms can replace outdated market leaders. However, major players still control key resources like cloud computing, proprietary data, and capital, meaning DeepSeek’s disruption may be limited in scale.

Q. To what extent will DeepSeek’s entry increase competition and reduce the pricing power of established AI firms?
A. Increased competition may drive down AI model prices, forcing dominant firms to differentiate through exclusive features or premium services. However, network effects and brand loyalty may allow firms like OpenAI and Google to maintain market power. Unless more firms replicate DeepSeek’s cost-effective approach, AI may remain oligopolistic rather than fully competitive.

Q. Justify whether the AI industry can be considered a contestable market following DeepSeek’s breakthrough.
A. A market is contestable if barriers to entry and exit are low. DeepSeek’s entry suggests increasing contestability, yet access to computing power, regulatory barriers, and intellectual property rights still restrict new entrants. While firms can develop AI models at lower costs, scaling them without access to proprietary data remains a challenge, making true contestability debatable.

Q. To what extent does DeepSeek’s second-mover advantage outweigh the benefits of being a first-mover in AI development?
A. Second-movers like DeepSeek can learn from first-movers’ mistakes, replicate technology at lower costs, and bypass expensive R&D phases. However, first-movers maintain brand recognition, data control, and intellectual property protections, which can limit second-mover advantages. The degree to which DeepSeek can sustain its edge depends on continued innovation and access to talent.

Q. Discuss whether the AI industry is likely to remain an oligopoly despite new entrants like DeepSeek.
​A. AI is likely to remain oligopolistic due to high fixed costs, limited access to computing power, and strong network effects. However, DeepSeek’s approach suggests that new firms can challenge dominance by leveraging cost efficiencies and open-source development. The extent to which AI becomes more competitive depends on technological progress, government regulation, and industry consolidation.

If you or your child are looking to Boost your A level Economics Grades in under 30 days, I'd recommend starting with an all-in-one support network where you get 24/7 access to a SuperTutor:

→ Join EdGenie 🧞‍♂️: Transform your A-Level Economics essays and exam marks (genuinely) with our comprehensive on-demand learning platform. This carefully curated course blends engaging content with effective exam techniques, the same ones that have empowered over 1,000 of my students to achieve an A or A* over the last 13 years.