The AI Valuation Paradox

I was reading an article that was published recently and couldn’t help but ponder whether the AI bubble will burst sooner or later. There are two aspects to the bubble. The AI “value proposition” and the AI “valuation”. I am clear that the AI value proposition is compelling (unlike the dot.com era where both value proposition and value were suspect) and this is probably the best Digital transformation journey anyone should embark on, failing which we will end up losing your organisation to digital disruption

The AI revolution is undeniable. Unlike the dot‑com era, the value proposition of AI is real and compelling. Organizations that ignore it risk being swept away by digital disruption.

But here’s the paradox: while AI adoption is skyrocketing, the valuation bubble is inflating far faster than the actual business impact.

• By 2025, although 72% of organizations use AI, only 39% report any EBIT impact.

• Fewer than 5% of profits are attributed to AI.

• And according to McKinsey, RAND, and Gartner, 80% of AI projects fail to scale beyond pilots.

AI is everywhere, except in the P&L.

The AI Stack: Layers, Players, and Bubble Risks

1. Hardware Layer (GPUs & Chips)

• NVIDIA: $220 B (estimated) revenue, $4.32T market cap, trading at 23x sales → Extreme risk.

• AMD: $34B (estimated) revenue, $335B valuation, 10x sales → High risk.

• Intel: $53B (estimated) revenue, $170B valuation, 2.9x sales → Moderate risk.

• Google TPUs are now near‑parity with NVIDIA, threatening its dominance.

TECH UPDATE: Google has launched its seventh-generation AI accelerator chip, Ironwood (TPU v7 / TPU7x), optimized for large-scale AI inference and available exclusively through Google Cloud.

2. Cloud Layer (Hyperscalers)

• AWS: $127B revenue (estimated), 3x sales → Moderate risk.

• Microsoft Cloud (including Azure): $168B (estimated), 12x sales → Moderate risk.

• Google Cloud: $55B (estimated), 10x sales → Moderate risk.

• Alibaba (Cloud intelligence group): $14B (estimated), 2.6x sales → High risk.

3. LLM Layer (Foundation Models)

• OpenAI: $20B revenue (estimated), $500B valuation → High risk.

• Anthropic: $9B revenue (estimated), $350B valuation → Extreme risk.

• xAI : $500M revenue (estimated), $230B valuation → Extreme risk.

4. Application Layer (Vertical & Agentic AI)

• Vertical AI (e.g., Harvey): $9B+ revenue (estimated), $10.2B valuation → Extreme risk.

• Enterprise AI: $23B+ revenue (estimated), $98B+ valuation → Extreme risk.

• Agentic AI Startups: $2.5B revenue (estimated), $5.4B valuation → Extreme risk.

Why the Bubble Will Burst

• Overvaluation: 54% of investors see a bubble. AI firms trade at twice the dot‑com peak.

• ROI drought: 80% of projects fail  

Beyond the Bubble: Emerging Frontiers

🔹 Quantum Computing

• Revenues projected to hit $1B in 2025 and $72B by 2035.

• Breakthroughs like Google’s 105‑qubit Willow processor show promise.

🔹 Small Language Models (SLMs)

• Market projected to grow from $0.93B (2025) to $5.45B (2032) at 28.7% CAGR.

• Edge‑friendly, privacy‑first, energy‑efficient — SLMs may deliver more practical value than oversized LLMs.

🔹 Agentic AI

• Market expected to grow from $7.38B (2025) to $47B (2030) at 44.8% CAGR.

Conclusion: Surviving and Thriving After the Burst

The AI valuation bubble is real — and its burst will be painful. But it will also be necessary. This isn’t the end of AI; it’s the beginning of a more mature, value‑driven phase.

Advice for Leaders:

• Ensure data quality before starting your AI journey.

• Treat every AI project as digital transformation, not a side experiment.

• Prioritize workflow redesign and measurable pilots.

• Embrace SLMs for targeted, cost‑effective solutions.

• Stay skeptical of hype , especially in quantum and agentic AI.

• Invest in people, not just technology.

The bubble will prick the weak, but the AI stack will endure. The future belongs to those who can separate signal from noise.

Final Thought: AI is not a fad. It’s a revolution. But revolutions always go through corrections. The winners will be those who build real business value while others chase speculative multiples.

‍