Weekly Cheonlian: Week 3 of November
Pokute
Category
Empty
This week, the tech industry was filled with conflicting signals about the explosive growth of AI infrastructure investment, along with concerns about operational costs and social structural risks. NVIDIA's record earnings and the massive capex of Big Tech demonstrate the intensifying competition for AI platforms. AI is fundamentally transforming business models across coding, commerce, and B2B markets. Simultaneously, warnings about the potential for AI-induced job losses and social disruption are growing.
1. AI Platform Wars: Distribution is the Key
As the gap in AI model quality rapidly narrows, the competitive focus is now shifting from model performance itself to platform strategy, deployment, and integration.
With the release of Gemini 3.0, Google is leveraging its distribution channels by bundling it deeply across the Google ecosystem, including Android, Chrome, Search, and Workspace. As Gemini evolves to OS-level capabilities and becomes integrated into the user experience, model quality may become a secondary concern.
OpenAI has taken a different path. Rather than relying on existing platforms, they're pursuing a high-risk, high-reward strategy of building ChatGPT into a superapp. They're attempting to attract users to a new center of gravity through ChatGPT's app store/SDK, note-taking functionality, and transition to a persistent agent.
Anthropic focuses on the API layer, emphasizing enterprise, security, and reliability over consumer use. Like AWS's initial cloud strategy, it clearly aims to secure developer trust and spread AI deep into enterprise systems.
Meta is deploying another strategy: aggressively releasing robust open-source models to level the playing field and force competitive differentiation into higher-end products and services.
The winners of the platform wars will not be those who create better models, but those who seamlessly integrate their offerings into the daily lives of more users.
2. AI Coding: From Optional to Essential
AI coding tools are no longer optional; they've become essential (Table Stakes).
Anysphere's Cursor's impressive valuation of $29.3 billion signals a market shift from experimental to essential in AI-based development. While Copilot enabled faster code creation, agent-based coding models like Cursor's Composer are transforming the very process of writing code.
In response to this competition, OpenAI is preparing the GPT-5.1-Codex-MAX model to handle large-scale software projects and long-term development tasks at the scale of repositories. This is an attempt to address the limitations of current coding assistants, which struggle to maintain understanding of code in large repositories.
In the future, Cursor will evolve beyond its agent capabilities to include CLI-based multi-agent orchestration and programmable development workflows. Organizations that adopt this first will gain a competitive advantage of 12 to 18 months.
As enterprises accelerate the adoption of AI coding tools, they must also update their security and governance frameworks for AI tools that require different system access privileges than traditional development tools.
3. Visual Innovation in Generative AI: Nano Banana Pro
Google has unveiled the Nano Banana Pro, based on the Gemini 3 Pro Image processor, showcasing new capabilities in image creation and editing.
The model can process up to 14 images simultaneously as input, maintain visual consistency across five characters, and render accurate multilingual text within images.
A particularly significant innovation is the integration of the real-time Google Search API. By refining generated images based on up-to-date, real-world data (such as weather and sports scores), it eliminates the risk of AI model hallucination in visualization tasks where data changes over time.
This model empowers data professionals to transform complex datasets into publication-quality infographics and dashboards using only natural language. Early users estimate it will reduce design iteration cycles by 80% and save millions of dollars in creative labor costs.
4. Nvidia: The Pickaxe and Shovel of AI Infrastructure
Nvidia is solidifying its unique position as the "pick and shovel" provider for the AI infrastructure boom.
The third quarter of fiscal year 2026 results were impressive:
•
Total sales of $57 billion (up 62% year-over-year)
•
Net income of $31.9 billion (up 65% year-over-year)
•
11 consecutive quarters of sales growth
Data Center segment revenue reached $51.2 billion, up 66% year-over-year, accounting for nearly 90% of total revenue. Increased shipments of Blackwell Ultra architecture-based products (GB300) drove this growth.
Networking revenue grew 162% year-over-year, growing into the world's largest networking business. Demand for NVLink Compute Fabric and XDR InfiniBand products for Grace Blackwell systems exploded.
Nvidia forecasts revenue of $65 billion for the fourth quarter of fiscal 2026, suggesting continued strong Blackwell sales.
5. AI Economics: Capital Beats Algorithms
The AI industry is economically proving Richard Sutton's "Bitter Lesson": that large-scale computing and capital triumph over clever algorithms.
Global spending on AI systems is expected to exceed $300 billion by 2026, with big tech companies planning to spend a total of more than $1.15 trillion on capital expenditures (capex) between 2025 and 2027.
AI projects differ from traditional software projects. Even with a smaller team, they can invest significantly more capital and achieve historic growth rates. This is because AI is best suited to problems that are difficult to specify but can be verified, such as creativity and language reasoning. This reduces technical issues to mere economic considerations.
Investment is driven by the constant demand for infrastructure and proven efficiency improvements:
•
Logistics company: GenAI pilot improves efficiency by 20-30%.
•
UK hospitals: 40% reduction in diagnosis times
6. OpenAI's Profitability Puzzle
OpenAI is seeing rapid revenue growth, but the astronomical costs of running its models are raising fundamental questions about its ability to remain profitable.
According to leaked financial data:
•
Microsoft's net revenue allocation for Q1-Q3 2025: $865.8 million
•
Based on this, OpenAI's minimum revenue is calculated to be over $4.3 billion.
•
CEO claims annual run rate: over $20 billion
But here's a shocking figure: the core computing and operational costs (inference costs) required to run OpenAI's models will reach $8.67 billion in the first three quarters of 2025.
Operating expenses are exceeding revenues, or at least increasing linearly with them, and are eating into profits.
OpenAI may be the most cash-intensive startup in history, raising concerns that the costs of running its large-scale language models may be unsustainable.
7. The AI Bubble: Between Fear and Opportunity
While astronomical investments and rising stock prices in the AI sector have sparked speculation about a "bubble," there are structural factors at play that are different from those of the past.
While expectations for Nvidia's stock (P/E ratio of 54-55) are high, the current AI boom is different from the dot-com bubble, as it is being funded by Big Tech's own operating cash flow rather than issuing new debt.
Some warn:
•
AI investment is excessive, exceeding the inflation-adjusted cost of the Apollo program ($300 billion).
•
Potential bubble formation as returns fail to keep pace with investments
However, there are also positive signs. The AI sector is currently experiencing a capacity shortage, with demand exceeding supply. This positively impacts the future outlook, despite short-term price declines.
Investors' outlook is also changing. Due to concerns about a tech crash driven by AI:
•
Probability of a bullish scenario (meltup) for the S&P 500: 25% → 15%
•
Bearish Scenario Probability: 20% → 30%
8. 12 Mega Trends in the B2B Market
The 12 trends that will drive the B2B market in 2026, analyzed by market research firm StartUs Insights, reveal sweeping changes across AI, commerce, regulation, and global supply chains.
AI Copilot and Agentic Workflow
The value of AI comes from contributing to the P&L through "proper deployment," not "demo." However, there are caveats: high costs, unclear ROI, and uncontrollable risks can lead to the termination of more than 40% of AI projects.
B2B marketplaces and digital self-service
With B2B ecommerce sales growing to $2.64 trillion, cloud marketplaces are driving this growth.
A notable change: 39% of B2B buyers are willing to complete orders exceeding $500,000 without a salesperson (rep-free). This highlights the potential of a "digital-first + sales support" strategy.
RevOps Automation and Subscription Models
To address inaccurate sales forecasting and tool proliferation, RevOps and sales enablement are integrating. Subscription and usage-based models are becoming the norm, with 61% of SaaS companies adopting usage-based pricing (UBP) elements.
ABM 2.0 and Data Privacy
Regardless of the discontinuation of third-party cookies, the industry has already shifted to first-party data and data clean rooms (DCRs). Sixty-four percent of companies adopting privacy technologies are already using DCRs.
Omnichannel and content strategy
77% of B2B buyers experience complexity in the purchasing process. Sellers must shift their role from "selling" to "facilitating."
Buyers now trust existing user reviews (77%) instead of analyst reports. The key to content is to provide "helpful content without being intrusive."
Industrial Cloud Platform and Edge
The cloud is moving beyond simple infrastructure and entering the era of industry-specific, customized finished products (ICPs), with adoption expected to exceed 70% by 2027.
Edge intelligence is proliferating, pushing analytics and control capabilities to the edge to reduce latency and data transmission costs.
Supply Chain Risk and Nearshoring
Geopolitical risks (Red Sea bypass) are driving a return to in-house production (nearshoring) in North America, with Mexico becoming the United States' largest trading partner.
Strengthening regulations
With the EU AI Act and the EU Cyber Resilience Act (CRA) looming, AI governance and software supply chain security (Zero Trust, SBOM) are becoming essential requirements for businesses.
9. AI Marketplace: Reinventing Failure
AI is reviving previously failed marketplace categories, simplifying complex purchasing processes, and changing the paradigm of commerce.
Marketplace Reactivation
AI is improving profitability by reducing operating costs and solving matching complexities in skilled labor, real estate, and home services markets that have previously struggled due to high customer acquisition costs (CAC) and low lifetime value (LTV).
The role of AI:
•
Automated intake/screening via artificial voice agents
•
Minimize manual reconciliation through transaction coordination and communication.
•
Encourage repeat transactions through personalized notifications
AI can provide transparent, fixed pricing for complex services (legal, roof repair), enabling a compelling value proposition for buyers.
The Rise of Conversational/Agentic Commerce
•
Conversational Commerce: AI Conducts Research to Inform Purchase Decisions
•
Agentic Commerce: AI executes the entire purchase process except payment confirmation.
When purchasing complex items (like long-distance hiking tents), AI provides more accurate and faster recommendations by referencing specialized websites rather than traditional search engines (which focus on advertising and conversion rates).
Recommended shopping traffic via ChatGPT is growing, but AI-powered traffic is still in its infancy, accounting for less than 1% of total traffic.
Trust and Personalization
Customers trust a retailer's own onsite agents 3x more than third-party agents (ChatGPT).
Amazon points out that third-party agents have many limitations, including lack of personalization, shopping history, and delivery/price accuracy. Amazon's own AI, Rufus, has contributed to a 60% increase in customer purchase completion rates and $10 billion in annual incremental sales.
10. Human capabilities enhanced by AI
Efficiency in the medical field
According to a Northwestern University clinical study, generative AI can reduce radiology documentation time by 15.5%, contributing to overall healthcare cost savings.
Public Sector Automation
According to research from the Alan Turing Institute, 84% of the UK government's routine processes could be automated, reducing email processing time by up to 70%.
Robotics and Physical World AI
Jeff Bezos' Project Prometheus focuses on applying AI to core areas of the physical world, such as engineering, manufacturing, and aerospace.
Robots like Physical Intelligence's π*0.6 model and Sunday's Act-1 model are improving their ability to handle household chores like washing dishes and laundry through reinforcement learning and high-precision human demonstration data.
Biotechnology and Climate Technology
General Control is using CRISPR to develop epigenetic editing therapies for aging, targeting chronic diseases like obesity and muscle loss.
Rainmaker has launched the largest cloud seeding project in U.S. history in Utah and Idaho using drones to spray silver iodine.
11. The Great Job Transformation: Opportunity or Crisis?
There are warnings that the speed and scale at which AI replaces existing jobs could be much faster than previous industrial revolutions, which could lead to serious social disruption.
Widespread job replacement
AI already threatens to replace hundreds of thousands of jobs for middle managers and knowledge workers, including performance marketing directors, auditors, consultants, financial analysts, and copywriters.
Deepening inequality
High-productivity workers could see their wages soar thanks to AI tools, but low-productivity workers and new graduates who lose out to robots will struggle to find jobs, exacerbating class inequality.
Social Collapse Scenario
Mass unemployment and rising inequality could lead to votes favoring socialist policies (higher taxes). This could trigger an exodus of capital and entrepreneurial talent, leading to a vicious cycle of government fiscal deterioration, inflation, and national debt defaults.
This is similar to the Luddite movement of the past or the European Revolutions of 1848. It is comparable to situations where new political systems (e.g., communism) emerged amidst the economic turmoil brought about by technological innovation.
These social problems are more likely to arise first in wealthier countries like Europe, Canada, and Australia than in countries with strong, diversified economies (with AI value creation and low tax pressure) like the United States and China.
12. The Recruitment Arms Race: Resumes Are Now for Machines
As the rise of AI intensifies the "arms race" in the hiring process, resume (CV) formats are being optimized for machines, not humans.
Accelerating Recruitment Automation
AI is increasing the number of applicants, and recruiters are increasingly relying on AI-powered applicant tracking systems (ATS) to handle the growing number of applicants.
ATS-friendly CV
The visual design-centric CVs of the past are disappearing. Machine-readable CVs are becoming:
•
Single column layout
•
Conventional titles (Experience, Skills, etc.)
•
Text-centered
These 'ATS-friendly' formats are becoming standardized.
The ease of a customized CV
Candidates can now easily generate customized resumes for each job posting in minutes using AI tools like ChatGPT. Even if your CV exceeds two pages, it's not a problem, as the ATS reads the entire content.
The primary purpose of a CV is now to get through the automated first hurdle.
13. Regulation and Ethics in the AI Era
AI regulatory bombshell
The EU AI law is set to come into full effect in August 2026. Violations could result in stiff fines of up to €35 million or 7% of global turnover.
Intensifying security threats
AI is being used in sophisticated fraud, such as deepfake video conferencing scams, making AI's "trust" and "security" key.
Strengthening Software Supply Chain Security
With the formalization of the Secure Software Development Attestation Form (SSDF) by the US CISA and the enactment of the EU Cyber Resilience Act (CRA), zero trust and software bill of materials (SBOM) submissions are becoming the default for delivery and procurement.
However, our research shows that 48% of security leaders are not meeting regulatory SBOM standards.
Legal Issues in AI and Social Relations
In the United States, a new social issue has emerged: "AI infidelity." Cases of emotional dependence on AI chatbots to alleviate loneliness are on the rise, and questions are being raised about how to legally define this as adultery and how to handle it in custody cases.
Digital Platform Responsibility
The case of French authorities imposing a travel ban on Telegram founder Pavel Durov on charges related to the platform's users' activities highlights a significant shift.
It's an attempt by the government to regulate digital communication in a way that targets those who create tools that provide online freedom.
References
1.
12 Megatrends That Will Drive the B2B Market in 2026 | Today's IT
2.
AI might help doctors be more efficient
3.
Divorces in the US due to AI infidelity... A new social problem is emerging.
4.
Bitter Economics - by Martin Casado - a16z
5.
Clouded Judgment 11.20.25 - The AI Platform Wars
6.
Cursor's $29 Billion Valuation - by Robert Matsuoka
7.
Data-Dependent
8.
France's Censors Release Their Favorite Captive
9.
Google Unveils Nano Banana Pro
10.
How AI is reshaping CVs
11.
Is the AI sector currently a bubble? - MARGINAL REVOLUTION
12.
Leaked finances hint that OpenAI's inference may be swallowing its revenue
13.
Marketplaces in the Age of AI - by Olivia Moore - a16z
14.
NVIDIA 2026 Q3 Financial Results - by Ryan Smith
15.
OpenAI prepares GPT-5.1-Codex-MAX for large-scale projects
16.
The $300 Billion Question: Why AI Investments Are Skyrocketing in 2025
17.
The AI-Led Pullback: Recalibrating The Odds Of Three Scenarios
18.
The appeal of conversational commerce
19.
Weekly Dose of Optimism #171
20.
When AI Takes Our Jobs - by Tomas Pueyo
21.
When will OpenAI start making money?
Do
Subscribe to 'Dogma13'
If you subscribe to the site, you can be the first to receive the latest updates, including new posts, via notifications and email.
Join Slashpage and subscribe to 'Dogma13'!
Join Slashpage and subscribe to 'Dogma13'!
