Artificial intelligence is having a major impact on the global economy by transforming how businesses operate, how people work, and how products and services are developed. AI is increasing automation, productivity, data processing, and demand for computing infrastructure, while also creating new industries and investment opportunities.

• Accelerates business automation and productivity.

• Changes the demand for labour and technical skills.

• Creates new markets around AI software, hardware, and services.

• Increases demand for semiconductors, GPUs, memory, and data centres.

• Attracts significant corporate and investor capital.

• Creates opportunities for entrepreneurs to build AI-powered businesses.

• Disrupts traditional industries and existing business models.

• Increases the importance of technology infrastructure in the economy.

The rise of Sam Altman and OpenAI has been one of the biggest stories in the AI economy. However, as OpenAI's valuation and spending have grown to extraordinary levels, some investors have begun questioning whether the company's future returns justify the risks.

This does not mean investors have abandoned OpenAI. In March 2026, OpenAI closed a $122 billion funding round at an $852 billion post-money valuation, demonstrating enormous continued investor support.

The issue is increasingly about whether confidence in the valuation and strategy is weakening at the margin.

💰 The Valuation Problem

OpenAI's enormous valuation creates a difficult question for potential investors:

"How much more valuable can this company realistically become from here?"

• OpenAI was valued at approximately $852 billion in its latest funding round.

• Some investors have questioned whether that valuation is justified.

• One investor cited by the Financial Times said underwriting the latest round required assuming an eventual IPO valuation of $1.2 trillion or more.

• The higher the entry valuation, the harder it becomes to generate attractive future returns.

• Investors therefore aren't simply evaluating whether OpenAI will succeed.

• They are evaluating whether OpenAI will succeed enough to justify the price they're paying.

📉 The Anthropic Problem

OpenAI's biggest issue isn't operating in isolation.

Investors have another major frontier-AI company to compare it against: Anthropic.

• Anthropic has gained significant traction in enterprise AI and coding.

• Some investors have reportedly begun viewing Anthropic as having a more attractive risk-reward profile.

• Secondary-market demand has reportedly shifted toward Anthropic while OpenAI shares have traded at discounts in some markets.

• This creates a simple investment question:

"Why pay an enormous premium for OpenAI when another AI company may offer better growth relative to its valuation?"

Competition changes investor psychology.

🏗️ Massive Capital Requirements

OpenAI's strategy requires enormous infrastructure investment.

AI isn't just software.

It requires:

GPUs → Data Centres → Electricity → Cooling → Networking → Capital

OpenAI has been committing huge amounts of capital toward computing infrastructure while simultaneously trying to expand revenue.

That creates a fundamental investor concern:

Can revenue and margins eventually grow fast enough to justify the infrastructure spending?

Investors therefore have to consider:

• Compute costs

• Data-centre commitments

• Energy requirements

• Employee costs

• Model-development costs

• Revenue growth

• Gross margins

• Future capital requirements

• Competition

🎯 Strategic Focus

Another source of investor concern has been strategic direction.

OpenAI has pursued numerous areas across:

• Consumer AI

• Enterprise AI

• Coding

• AI agents

• Video generation

• Hardware

• Search

• Advertising

• Data centres

• Scientific applications

Some investors have questioned whether the company is spreading itself too widely while facing intense competition.

Reports in 2026 described investors questioning OpenAI's strategic shifts as Anthropic and Google strengthened their positions.

The investor perspective is essentially:

"You have limited capital and compute. Where should you deploy it to generate the highest return?"

🧑‍💼 Sam Altman's Personal Reputation

This is where the issue becomes more personal.

Altman is effectively one of the most important public faces of the AI industry. Therefore, confidence in OpenAI's leadership can influence confidence in the company itself.

In May 2026, Reuters reported that testimony during Elon Musk's lawsuit against OpenAI included repeated accusations from former colleagues questioning Altman's trustworthiness. The jury ultimately rejected Musk's case, but Reuters noted that the testimony could still affect investor perceptions ahead of a potential IPO.

This creates an interesting distinction:

Legal victory ≠ reputational victory.

A company can win a legal battle while investors still ask:

"Do I trust this leadership team with hundreds of billions of dollars?"

🧠 The Trust Equation

For major investors, the equation isn't simply:

AI capability = investment

It becomes:

Technology + Revenue + Margins + Leadership + Governance + Valuation = Investment Thesis

If one component becomes questionable, the entire thesis needs to be reassessed.

⚠️ The Bigger Risk

The biggest danger for Altman isn't necessarily that investors suddenly stop believing in AI.

It is that investors continue believing in AI but become less convinced that OpenAI is the best vehicle for capturing its economic value.

That's a much more serious competitive problem.

Investors could believe:

"AI will be enormous."

while simultaneously believing:

"OpenAI might not capture enough of that value to justify an $852B+ valuation."

That's the distinction between believing in the technology and believing in the investment.

🏆 What Could Restore Confidence?

OpenAI can strengthen investor confidence by demonstrating:

• Sustained revenue growth

• Improving gross margins

• More efficient compute utilization

• Strong enterprise adoption

• Defensible technological advantages

• Clear strategic priorities

• Disciplined capital allocation

• Strong corporate governance

• Transparent financial reporting

• Evidence that AI products can become high-margin businesses

Ultimately, investors want to see the transition:

AI hype → AI adoption → AI revenue → AI profits → AI economic moat

Final Lesson

Sam Altman's challenge isn't simply building the world's leading AI company.

It is convincing investors that the company can become valuable enough to justify the enormous expectations already embedded in its valuation.

And that's a brutal game.

At a $10 billion valuation, investors can imagine enormous upside.

At an $852 billion valuation, investors start asking:

"Show me the trillion-dollar business."

Technology creates potential.
Revenue creates evidence.
Profitability creates durability.
Trust creates capital.
Valuation determines whether the opportunity is actually worth buying.

The higher the valuation rises, the less investors pay for the dream—and the more they demand proof.

The rapid expansion of AI infrastructure has increased demand for high-performance memory and related semiconductor components. Memory markets can experience significant price movements when demand rises faster than manufacturing capacity.

• Increased demand for DRAM and high-performance memory.

• AI data centres require substantial amounts of memory and computing infrastructure.

• Manufacturers may prioritize higher-margin AI-oriented memory products.

• Supply constraints can contribute to higher memory prices.

• Rising component costs can affect PCs, servers, and electronics.

• Demonstrates how AI demand can create second-order effects across supply chains.

• Highlights the relationship between technology demand, manufacturing capacity, and pricing.

AI has become a major theme in financial markets, influencing investor sentiment, corporate valuations, capital expenditure, and sector performance. Companies associated with AI infrastructure and applications have attracted substantial attention as investors attempt to identify the businesses positioned to benefit from the technological shift.

• AI has become a major investment and market theme.

• Investors evaluate companies based on their AI exposure and growth potential.

• AI-related companies can experience significant changes in market valuations.

• Large technology companies are investing heavily in AI infrastructure.

• AI expectations can influence stock prices and investor sentiment.

• Capital is flowing toward areas such as semiconductors, cloud computing, data centres, and AI software.

• Creates both investment opportunities and valuation risks.

• Demonstrates how technological innovation can influence financial markets and capital allocation.

The rise of AI-generated courses has made it possible for almost anyone to turn an idea into a structured educational product within minutes. AI can generate lessons, scripts, quizzes, presentations, images, and course materials at a fraction of the traditional cost and time.

However, this convenience creates a major challenge:

People may question whether an AI-generated course is actually worth learning.

The Rise of AI-Generated Education

• AI can rapidly generate course outlines and lesson plans.

• It can turn a topic into structured modules and lessons.

• AI can create quizzes, exercises, summaries, and study materials.

• Individuals can produce courses without traditional educational infrastructure.

• Course creation becomes cheaper and faster.

• More people can become educational content creators.

⚠️ The Trust Problem

The biggest issue isn't necessarily whether AI can create a course.

It's whether people trust the person selling it.

Students may ask:

• "Did you actually learn this yourself?"

• "Does the instructor have real-world experience?"

• "Is this just AI-generated information?"

• "Has anyone actually tested these lessons?"

• "Why should I pay for information I could ask ChatGPT for?"

• "Is there any evidence that this course works?"

This creates a major credibility gap.

🧠 Information ≠ Expertise

AI can generate information extremely quickly.

But information alone doesn't automatically create expertise.

There is a difference between:

Knowing information

and

Knowing how to apply information in the real world.

Someone can use AI to generate a course about entrepreneurship without ever having built a business.

Someone can generate a course about investing without having experienced actual market cycles.

Someone can generate a programming course without having built and maintained real software.

The content might sound professional while lacking genuine practical experience.

🤖 The "AI Slop" Problem

When AI makes content extremely cheap to produce, the internet can become flooded with:

• Generic courses

• Recycled information

• AI-written ebooks

• AI-generated tutorials

• Low-effort certifications

• Repackaged public information

• Courses created primarily to make money rather than teach

This creates a paradox:

AI makes education easier to produce → more educational content appears → quality becomes harder to distinguish → trust becomes more valuable.

🏆 What Makes an AI-Assisted Course Trustworthy?

AI itself isn't necessarily the problem.

The difference is how it is used.

Weak Model

AI → Generate course → Sell course

The creator contributes little original knowledge or validation.

Strong Model

Experience → Research → AI Assistance → Human Verification → Practical Testing → Course

Here, AI becomes a tool, rather than the source of authority.

A credible course should ideally contain:

• Real-world experience

• Original frameworks and perspectives

• Demonstrated expertise

• Verified information

• Practical examples

• Case studies

• Exercises and projects

• Evidence of application

• Honest disclosure of AI usage

• Continuous updates and corrections

🔍 Proof Becomes More Valuable

As AI-generated content becomes more common, proof of competence becomes a competitive advantage.

Instead of saying:

"I am an expert."

A stronger approach is:

"Here is what I built."

Instead of:

"This strategy works."

Show:

"Here is how I applied it, what happened, and what I learned."

Instead of:

"Take my course."

Demonstrate:

"Here is the framework, here is the practical exercise, and here are the results."

💡 The New Education Economy

AI may eventually create an interesting shift in education.

When information becomes abundant, the scarce resource becomes:

Trust.

And trust comes from:

Experience + Evidence + Reputation + Transparency + Results

AI can generate the content.

Humans still need to provide the credibility.

Final Lesson

The future probably isn't:

"AI replaces teachers."

It is closer to:

"AI makes creating educational content incredibly easy, while making genuine expertise incredibly valuable."

The challenge won't be creating another course.

The challenge will be proving that your course deserves to be trusted.

AI can generate a lesson.
Experience can validate it.
Evidence can strengthen it.
Results can build trust.

In an internet flooded with AI-generated knowledge, credibility becomes the real product.

AI development requires enormous amounts of computing power, electricity, cooling, networking, and physical infrastructure. As AI models become more sophisticated, demand for specialized data centres capable of supporting large-scale computation continues to increase.

• Require large quantities of GPUs and computing hardware.

• Consume significant amounts of electricity.

• Create demand for advanced cooling systems.

• Increase demand for semiconductors and networking equipment.

• Require substantial capital investment and infrastructure development.

• Create opportunities in energy, construction, real estate, and telecommunications.

• Become increasingly important to the AI supply chain.

• Represent the physical infrastructure behind the AI economy.

Data centres consume water primarily for cooling the massive amounts of heat generated by servers, GPUs, and other computing equipment. As AI workloads become more computationally intensive, water use has become an increasingly important environmental and infrastructure consideration.

Why Do Data Centres Need Water?

• Cooling servers — High-performance processors generate substantial heat.

• Evaporative cooling — Water can be evaporated to remove heat efficiently.

• Cooling towers — Many facilities use cooling towers as part of their heat-rejection systems.

• Humidity control — Some facilities use water-related systems to maintain suitable environmental conditions.

• Electricity generation — Water can also be consumed indirectly because some power-generation methods require water for cooling.

🤖 AI's Impact

AI data centres can create particularly high cooling demands because AI workloads often use large numbers of GPUs and other high-performance processors.

• More computing → more heat

• More heat → greater cooling requirements

• Greater cooling requirements → potentially greater water consumption

Large AI training and inference operations can therefore increase the resource requirements of data-centre infrastructure.

🌊 Direct vs. Indirect Water Consumption

Direct Water Consumption

• Water used inside the data centre for cooling.

• Evaporation from cooling towers.

• Water used for maintaining cooling systems.

Indirect Water Consumption

• Water used to generate the electricity powering the facility.

• Water associated with manufacturing semiconductors and computing hardware.

• Water consumed throughout the broader technology supply chain.

This distinction matters because a data centre can have relatively low direct water consumption while still having a significant overall water footprint.

⚠️ Why It Matters

Water consumption becomes a bigger concern when data centres are located in water-stressed regions.

Potential challenges include:

• Competition with local communities and agriculture.

• Increased pressure on municipal water infrastructure.

• Environmental concerns surrounding water availability.

• Greater scrutiny of new AI data-centre projects.

• Rising operating and infrastructure costs.

• Difficulties balancing energy efficiency with water efficiency.

♻️ Reducing Water Consumption

Data-centre operators can reduce their water footprint through:

• Air cooling

• Closed-loop cooling systems

• Liquid cooling with water reuse

• Recycled or reclaimed water

• More efficient cooling infrastructure

• Better server utilization

• Locating facilities where water and energy resources are suitable

• Using electricity sources with lower associated water consumption

🔗 The Bigger Picture

The AI economy isn't just about chips and software.

It requires:

AI → GPUs → Data Centres → Electricity → Cooling → Water

This creates a broader infrastructure chain where AI growth can affect energy demand, water resources, construction, utilities, and local communities.

Final Lesson

AI may run in the cloud, but the cloud still needs physical resources.

Every AI query ultimately depends on physical infrastructure—servers, electricity, cooling systems, buildings, and increasingly, water.

The challenge for the next generation of data centres is therefore not simply:

"How do we build more computing power?"

It is:

"How do we build more computing power while using energy, water, and infrastructure sustainably?"

An AI worm is a type of malicious software designed to use artificial intelligence systems as a mechanism for spreading, adapting, or carrying out attacks. Unlike traditional computer worms that primarily exploit software vulnerabilities, AI-powered worms could potentially exploit the way AI agents process information, communicate, access tools, and interact with other systems.

How AI Worms Could Work

• Exploit vulnerabilities in AI-powered applications or agents.

• Manipulate AI systems through malicious instructions or injected content.

• Automatically identify potential targets or attack paths.

• Spread malicious instructions between connected AI systems.

• Use AI to adapt its behavior based on the environment.

• Potentially interact with email, messaging, databases, APIs, or cloud services.

• Exploit excessive AI permissions and system access.

• Automate parts of an attack that previously required human intervention.

Why It Is Different

Traditional malware generally follows pre-programmed instructions.

An AI-enabled worm could potentially use an AI model to:

Observe → Interpret → Adapt → Act

This creates a different cybersecurity challenge because the malicious system may be capable of responding to changing conditions rather than following one rigid sequence.

Potential Risks

• Rapid propagation across interconnected systems.

• Automated discovery of vulnerable systems.

• Data theft or manipulation.

• Unauthorized access to connected services.

• Generation of convincing phishing or social-engineering content.

• Abuse of AI agents with excessive permissions.

• Difficulty determining whether an action was intentional or manipulated.

• Increased scale of cyberattacks.

The AI Attack Surface

As businesses connect AI to more systems, the potential attack surface expands:

AI Model → AI Agent → API → Database → Cloud → Enterprise Systems

The AI itself may not necessarily be the weakest point.

Sometimes the bigger vulnerability is what the AI is allowed to access and control.

The Bigger Concern

The most important shift is from:

"Malware that executes instructions."

to:

"Malware that can potentially interpret its environment and dynamically respond."

This doesn't mean AI worms are automatically autonomous super-malware. Their capabilities depend heavily on the AI model, permissions, system architecture, vulnerabilities, and security controls surrounding them.

🛡️ Defensive Perspective

Organizations can reduce these risks through:

• Least-privilege access

• Strong authentication

• Sandboxing AI agents

• Input and output validation

• Network segmentation

• Continuous monitoring

• Human approval for high-impact actions

• Secure API permissions

• Malware and endpoint detection

• Regular security testing

• Clear separation between AI reasoning and system execution

Final Lesson

AI creates enormous opportunities—but connecting AI to real-world systems also creates new cybersecurity risks.

The fundamental principle is:

The more autonomy you give an AI system, the more important security boundaries become.

AI → Capability
Connectivity → Reach
Autonomy → Action
Security → Control

The future of cybersecurity won't simply be about protecting computers from humans.

It may increasingly involve protecting AI systems from being manipulated into attacking other systems.