Impact of Artifical Intelligence on various industries.
1. Software & Knowledge Work
Why #1: AI directly replaces or augments human cognitive labor across every industry.
What changes
- Documents, spreadsheets, presentations → become interfaces, not artifacts
- Work shifts from “doing” to directing agents
- Massive productivity compression (1 person → 5–10x output)
Examples
- Financial models built via prompts instead of Excel formulas
- Sales, HR, legal, IT ops handled by autonomous agents
- “Post-document workplace” where outputs are dynamic, not static files
2. Healthcare & Life Sciences
Why #2: AI attacks the cost + discovery bottleneck simultaneously.
What changes
- Drug discovery timelines collapse (years → months)
- Diagnostics become earlier, cheaper, more accurate
- Personalized medicine at scale
Examples
- AI-designed proteins & molecules
- Radiology & pathology triage by models
- AI copilots for doctors, not replacements
3. Finance & Capital Markets
Why #3: Finance is already digital + data-heavy → AI plugs in fast.
What changes
- Research, modeling, compliance, and risk become automated
- Personalized financial advice at scale
- Faster capital allocation with fewer humans
Examples
- AI-built earnings models & scenario analysis
- Fraud detection in real time
- Autonomous portfolio rebalancing
4. Manufacturing, Robotics & Industrial Systems
Why #4: AI + robotics = physical world leverage.
What changes
- Self-optimizing factories
- Predictive maintenance replaces downtime
- Labor-light production at scale
Examples
- Robots trained via simulation, not hand-coding
- AI-managed supply chains
- Autonomous inspection & quality control
5. Media, Marketing & Advertising
Why #5: AI collapses the cost of content + targeting.
What changes
- Infinite personalized creatives
- Real-time campaign optimization
- Agencies become smaller but more powerful
Examples
- Ads generated per-user, per-context
- AI-driven customer journey orchestration
- Automated A/B testing at massive scale
Education & Training
Why #6: Massive disruption, but slower institutional adoption.
What changes
- Personalized tutors at near-zero cost
- Skill-based learning replaces credential-based systems
- Continuous reskilling becomes normal
Examples
- AI tutors adapting to each student
- Corporate training agents
- Automated assessment & feedback
Main players
- Duolingo
- Khan Academy
- Coursera
- OpenAI
Energy, Infrastructure & Climate
Why #7: Huge upside, but capital- and regulation-heavy.
What changes
- AI-optimized grids & energy markets
- Faster materials discovery
- Predictive infrastructure management
Examples
- Load balancing for AI data centers
- Nuclear + renewables optimization
- AI-driven energy trading
Main players
- Schneider Electric
- GE Vernova
- NextEra Energy
- Siemens Energy
TAM / CAGR
Estimated Total AI Infrastructure Spend (2025)
- According to Jon Peddie Research, global spending on AI infrastructure (covering hardware and software) is estimated at around $375 billion in 2025, with projections rising to $500 billion in 2026.
- Nvidia forecasts that this investment will expand significantly and expects cumulative global AI infrastructure spending to reach $3–4 trillion by 2030, with annual spending around $600 billion in 2025
Top Entities Driving the Spending
| Entity | Overall | 2025 | 2026 |
|---|---|---|---|
| Open AI | ~550 Bill$ for 10 GW of AI datacenter with NVDA. NVDA will contribute 10Bill$ per GW | 1 GW of AI datacenter with NVDA Vera Rubin | |
| Amazon | ~$100 billion | ||
| Microsoft | ~$80 billion | ||
| Alphabet | ~$85 billion | ||
| Meta | ~$64–72 billion | ||
| Apple* | Part of a $500 billion multi-year investment |
Hardware Compute
Cost to build a 1 GW datacenter
Total : 50 – 60 Bill $
Nvidia chips and systems: 35 Bill $ ( 35 : Rubin, 30: Blackwell, 25 : Hopper )
Infrastructure provider : 15 – 20 Bill $
Compute
1 NVL72 GB300 = 36 Grace CPUs and 72 Blackwell GPUs
1NVL72 provides around 150KW of compute
1GW of compute = ~7500 Nvidia NVL72 racks.
Invest Alpha 