AI & Technology - The Emerging Investment Theme

By EC Assets Research Team, Technology Themes · Published 19 May 2026 · Updated 15 June 2026

AI & Technology — Artificial intelligence and broader technology themes have become a primary institutional investment focus since 2022. The category spans direct AI infrastructure (chips, data centres, models) and AI-affected sectors across the economy.

Definition

Artificial intelligence and broader technology themes have moved from peripheral to central in institutional investment thinking since 2022. The category encompasses direct AI infrastructure (semiconductors, data centres, networking, power generation), AI model developers (OpenAI, Anthropic, Google DeepMind, others), AI applications across software and services, and the multi-sector productivity implications of AI deployment.

The thematic acceleration began with the November 2022 launch of ChatGPT, which demonstrated capabilities that had not been publicly available before and triggered an unprecedented surge in AI-related capital allocation. By 2024, AI-related capital expenditure exceeded $300 billion annually globally, with the largest cloud providers (Microsoft, Google, Amazon, Meta) committing tens of billions each to AI infrastructure buildout.

The investment thesis spans the spectrum from bull (AI is the most consequential general-purpose technology in decades, justifying current investment) to bear (current spending is a speculative bubble that will produce overcapacity and capital destruction). The institutional question is portfolio positioning under genuine uncertainty about which thesis proves more accurate.

The AI Investment Universe

The category includes both pure-play AI companies and existing companies whose value is materially affected by AI exposure.

Nvidia and the Infrastructure Layer

Nvidia's position in AI infrastructure is the defining story of the current cycle:

[!key] Nvidia's revenue grew from $27B in 2022 to over $130B in 2024, driven almost entirely by AI training and inference workloads. The company's market capitalisation expanded from approximately $400B to over $3T over the same period, briefly making it the world's most valuable company. The concentration of AI compute on Nvidia hardware creates both substantial economic opportunity for Nvidia and substantial concentration risk for the broader AI investment thesis - if Nvidia's pricing power compresses or alternatives gain meaningful market share, the broader thesis is materially affected.

Nvidia faces growing competition: AMD's MI300 GPUs, Google's TPU (used internally and for cloud customers), Amazon's Trainium and Inferentia chips, and custom silicon from Meta and Microsoft. The competitive dynamic over the next 3-5 years will determine whether Nvidia's current pricing power persists or compresses toward more normal semiconductor industry economics.

The Bull and Bear Cases

The institutional investment debate splits into recognisable bull and bear cases:

Bull case.

• AI is a genuine general-purpose technology comparable to electricity or the internet
• Productivity gains across the economy will be substantial and persistent
• Current capital expenditure builds infrastructure that supports decades of subsequent value creation
• Concentrated winners (Nvidia, hyperscale cloud providers, frontier model developers) have durable moats
• Even if current valuations overshoot, the long-run direction is up and to the right

Bear case.

• Current capital expenditure exceeds plausible near-term return-on-investment
• Foundation model commoditisation will compress the margins that justify current valuations
• Productivity gains will materialise more slowly and unevenly than current expectations suggest
• Open-source models will erode the moats of proprietary developers
• Regulatory frameworks may constrain deployment, particularly in Europe and certain sectors
• The pattern resembles previous technology cycles (internet, mobile) where the technology proved real but investor returns were uneven

Most sophisticated institutional positioning acknowledges both cases and sizes exposure to be operationally durable across multiple scenarios rather than concentrated on either pure bull or pure bear outcomes.

The Multi-Sector Implications

Beyond direct AI infrastructure, every major economic sector faces AI-driven changes. The investment question is which companies in each sector benefit, which are disrupted, and which are unaffected:

Software. Existing software companies face both opportunity (AI features increase software value) and threat (AI-native competitors challenge incumbents). Microsoft has integrated OpenAI capabilities; ServiceNow and Salesforce have launched AI products; Adobe has integrated AI into creative workflows. The competitive dynamics within software are particularly active.

Professional services. Consulting, accounting, legal, and financial services face material productivity gains. The same productivity gains may compress billable-hours business models. Firms have generally embraced AI for efficiency while struggling with what it means for fees.

Healthcare. AI applications in drug discovery (DeepMind's AlphaFold and others), medical imaging (radiology assistance), and patient care show genuine clinical and economic promise. Specific companies and specific therapeutic areas remain selective bets.

Manufacturing. AI-enabled optimisation, robotics, and quality control offer productivity improvements. The category is operationally more difficult to invest in than software.

Financial services. Banking, insurance, and asset management face both opportunity (AI for risk assessment, fraud detection, customer service) and disruption (algorithmic competition with traditional advisory models).

Implementation: How Institutions Are Allocating

Institutional approaches to AI exposure vary by mandate:

Public equity allocators typically maintain overweight positions in large-cap technology (the Magnificent Seven dominate AI exposure) with specialised satellite allocations to AI-focused ETFs and individual stocks across the value chain.

Venture capital allocators have substantially increased commitments to AI-focused venture funds. Multiple new managers (Andreessen Horowitz's a16z American Dynamism, Lightspeed AI Fund, Sequoia's continued AI focus) have raised dedicated AI funds in 2023-2024.

Private equity allocators are seeing more AI-related opportunities in late-stage growth equity, with valuations recovering from 2022-2023 compression as AI thesis solidified.

Infrastructure allocators are focusing on data centres, electricity transmission, and power generation - the physical infrastructure that AI deployment requires.

Common Misconceptions

"AI investing is just tech investing." Partial truth. Major tech companies represent a large share of AI value capture, but the theme extends to power generation, real estate (data centres), semiconductors specifically, professional services, and sector-specific applications. Treating AI as synonymous with megacap technology misses substantial parts of the opportunity set.

"The AI bubble will pop and AI investing will end." Most likely partly correct. Specific valuations may correct sharply at some point - historical technology cycles have always included corrections. But the underlying technology and its productivity implications are real. Even after a hypothetical correction, AI as an investment theme will persist; the question is just which specific companies and at what valuations.

"You can replace human judgment in investing with AI." Premature. Current AI systems can process information and identify patterns at scale, but high-stakes investment decisions involving genuine uncertainty, novel situations, and qualitative judgment still require human input. The deployment is in augmentation rather than replacement, and the productivity benefit is real but specific.

References

1. Lopez de Prado, M. (2018). Advances in Financial Machine Learning. Wiley.
2. CFA Institute. Machine Learning and Big Data in Investment Management. CFA Program Curriculum.