Downloaded 13 times
![Magnificent Seven
vs S&P 493: Key
Financial Metrics
(Nov-Dec 2025)
Metric Magnificent Seven S&P 493 Sources
Market Cap $21.5-22T(37% of S&P 500) $36T(63% of S&P 500)
[Yahoo Finance] ,
[MacroMicro] ,
[MarketWatch]
TTM Revenue (Q4
2024 est)
$2.0T+($509B quarterly,
+12.8% YoY)
$3.5T+(+4.2-5.2% YoY) [LSEG] , [First Trust]
TTM
Profits/Margins
25.8% net margins($131B Q4
earnings, +31.7% YoY)
13.4% net margins(S&P
ex-Mag7 +9.4% Q2)
[LSEG] , [FactSet via
Reddit]
S&P 500
Weightage
37%(up from 12% in 2015) 63%
[Yahoo] ,
[MacroMicro]
2025 EPS Growth
(Forecast)
+17.1%(33% of total S&P
growth contribution)
+9.2%
[LSEG] , [Columbia
Threadneedle]
Revenue Growth
(Recent Q)
+12.8% Q4(Nvidia/Amazon
57% contribution)
+4.2-5.2% [LSEG]
Earnings Growth
Contribution
52% of S&P total(Q4 2024),
44% past year
48% of S&P total [First Trust] , [LSEG]
US GDP
Contribution (Est)
10-12%(via tech/AI infra
dominance)
88-90%(rest of economy)
[Accio est] ,
[WisdomTree]
YTD Return (2025
thru Q3)
Contributes 41.8%to S&P
14.8%
58.2% contribution [Accio]
Mag7 = 37% S&P but 52% earnings
growth
• Concentration Risk: Mag7 =
37% S&P weight (highest since
dot-com era); Nvidia alone = 21%
of group cap
• Earnings Power: Mag7 drove 3x
earnings growth vs rest (31.7% vs
9.4% Q2); margins 2x higher
• Growth Divergence: Mag7 EPS
expectations +4% in 2025; S&P
493 -1.48%
• AI Dependency: Mag7
performance tied to AI
capex/revenue; rotation risks if
growth slows.](https://image.slidesharecdn.com/genaiin2026-251206192417-b7be8fe6/85/Generative-AI-in-2026-Hype-Bubble-Winter-or-The-Real-Deal-4-320.jpg)
Uploaded byDr. Tathagat Varma
Generative AI in 2026: Hype, Bubble, Winter or The Real Deal?
In this presentation, I take a critical look at the state of AI/GenAI/AgenticAI as we go into 2026.
More Related Content
Similar to Generative AI in 2026: Hype, Bubble, Winter or The Real Deal?
![[DSC DACH 24] Harnessing AI for Sustainable Results: Strategies for Long-Term...](https://cdn.slidesharecdn.com/ss_thumbnails/johannesschauer-240921154533-335db038-thumbnail.jpg?width=640&height=640&fit=bounds)
![[DSC Europe 25] Mathias Halkjær Petersen - The AI workforce revolution.pptx](https://cdn.slidesharecdn.com/ss_thumbnails/3xviexv7q5gojhdsyvat-the-ai-workforce-revolution-251218084820-f3c286ed-thumbnail.jpg?width=640&height=640&fit=bounds)
![Conviction LP Letter - May 2024 [Redacted]](https://cdn.slidesharecdn.com/ss_thumbnails/convictionlpletter-may2024sentredacted-250215033505-a84e3cae-thumbnail.jpg?width=640&height=640&fit=bounds)
Generative AI in 2026: Hype, Bubble, Winter or The Real Deal?
- 1. Generative AI in2026: Hype, Bubble, Winter or The Real Deal? Dr. Tathagat Varma, PhD (GenAI) Executive Fellow, Indian School of Business, Co ’25 “Cognitive Chasms: A Grounded Theory of GenAI Adoption©”
- 2. Disclaimer! This presentation isbased on my doctoral research titled “Cognitive Chasms: A Grounded Theory of GenAI Adoption”. You are free to refer to my work and use as appropriate, so long as you cite it ☺
- 3.
- 4. Magnificent Seven vs S&P493: Key Financial Metrics (Nov-Dec 2025) Metric Magnificent Seven S&P 493 Sources Market Cap $21.5-22T(37% of S&P 500) $36T(63% of S&P 500) [Yahoo Finance] , [MacroMicro] , [MarketWatch] TTM Revenue (Q4 2024 est) $2.0T+($509B quarterly, +12.8% YoY) $3.5T+(+4.2-5.2% YoY) [LSEG] , [First Trust] TTM Profits/Margins 25.8% net margins($131B Q4 earnings, +31.7% YoY) 13.4% net margins(S&P ex-Mag7 +9.4% Q2) [LSEG] , [FactSet via Reddit] S&P 500 Weightage 37%(up from 12% in 2015) 63% [Yahoo] , [MacroMicro] 2025 EPS Growth (Forecast) +17.1%(33% of total S&P growth contribution) +9.2% [LSEG] , [Columbia Threadneedle] Revenue Growth (Recent Q) +12.8% Q4(Nvidia/Amazon 57% contribution) +4.2-5.2% [LSEG] Earnings Growth Contribution 52% of S&P total(Q4 2024), 44% past year 48% of S&P total [First Trust] , [LSEG] US GDP Contribution (Est) 10-12%(via tech/AI infra dominance) 88-90%(rest of economy) [Accio est] , [WisdomTree] YTD Return (2025 thru Q3) Contributes 41.8%to S&P 14.8% 58.2% contribution [Accio] Mag7 = 37% S&P but 52% earnings growth • Concentration Risk: Mag7 = 37% S&P weight (highest since dot-com era); Nvidia alone = 21% of group cap • Earnings Power: Mag7 drove 3x earnings growth vs rest (31.7% vs 9.4% Q2); margins 2x higher • Growth Divergence: Mag7 EPS expectations +4% in 2025; S&P 493 -1.48% • AI Dependency: Mag7 performance tied to AI capex/revenue; rotation risks if growth slows.
- 5. GenAI Report Card Metric2025 Estimate Key Sources & Notes Total Infrastructure Capex $350-611B globally (hyperscalers: Amazon, Google, Microsoft, Meta, Oracle) • KKR: Top 4 hyperscalers >$350B (mid-30% YoY growth) • BofA: Global hyperscale $611B (+67% YoY) • Morgan Stanley: $342B (top hyperscalers) • IO Fund: $405B total AI infra AI-specific portion: 60-80% of total capex Total VC Investment $192.7-366.8B globally (AI = 53-63% of all VC) • PitchBook: $192.7B YTD (1st time >50% of global VC); Q3 alone $97B • CB Insights: Q2 $47.3B (H1 $116.1B, 2nd highest ever) • US dominates: $250B total VC, 63% AI Total Revenues $16-71B (GenAI market); $20-40B (leading AI labs) • S&P Global: GenAI $16B (2024) → $85B (2029), 2025 ~$22-37B • Epoch AI: OpenAI ~$10-13B ARR; Anthropic $1.4B; DeepMind single- digit B • Grandview: GenAI $22.2B (2025) • MarketsandMarkets: $71.36B total GenAI Total Profits Net losses dominant (~$50-100B+ infra losses offset rev) • No aggregate profits; hyperscalers capex > rev growth • OpenAI: $13B rev but massive losses (CFO: "well more than $13B") • IBM CEO: "No way" hyperscalers profit on data centers soon • DeepMind: £1.5B rev (2023) but R&D losses
- 6.
- 7. AI/GenAI/AgenticAI Failure Rates CategoryFailure Rate Details & Examples GenAI Pilots 95% fail • MIT: 95% enterprise GenAI pilots yield zero measurable business return despite $30-40B spend. Core issue: "GenAI Divide" (no workflow redesign). Enterprise AI Overall 42-80% abandoned • S&P: 42% companies scrapped most AI initiatives (up from 17% in 2024). • RAND: 80%+ never reach production. • Gartner: 40% agentic canceled by 2027. Agentic AI 90-95% fail • Beam.ai: 90% implementations fail (edge cases, no metrics). • MIT: Agentic subset of 95% GenAI failures. • NTT: 70-85% miss ROI. AI Companies Financial Success <5-26% profitable • MIT: Only 5% pilots succeed. • BCG: 26% move beyond PoC, 4% generate "significant value". • Epoch: Leading labs (OpenAI/Anthropic) unprofitable ($5-15B losses despite $10-20B rev).
- 8. Productivity Gains ByArea Area Productivity Gain Specific Stats & Caveats Coding (General) 12-21% net gain • Developers write 12-15% more code (+21% self-reported productivity). • GitHub Copilot: 126% more projects/week. • But: METR RCT (experienced devs): +19%slower completion time despite 24% expected gain. AI-Assisted Software Dev 21-36% time savings • Google: 21% of code AI-assisted. Enterprises: 33-36% less dev time. • M365 Copilot: 30 min/week saved, 12% faster docs. • Security risk: 48% AI code has vulnerabilities (needs review). • Small firms: 50% faster testing/debugging. Overall Enterprise <10% typical • McKinsey/Stanford: Most report <10% cost savings, <5% revenue gains (marketing/sales: 71% see gains, but low magnitude). • Leaders: 1.5x revenue growth.
- 9. Reality Check • Hypevs. Reality: Devs expect 24% speed-up but get slowdowns in RCTs (over-reliance). 41% code AI-generated (2025), but 80% needs human fixes. • Success Factors: Winners (5%) define metrics upfront, build for failure, iterate like "employee onboarding" (93% ROI possible). • Financial Reality: AI labs burn $5-20B/year despite rev growth; hyperscalers profitable overall but AI is capex drag.
- 10. So, what are thebig questions? Why: Why does AI fail every so often? What: What does adoption mean? How: How do we measure success?
- 11. Cognitive Chasms: AGrounded Theory of GenAI Adoption”©
- 12. Multistage Scaling Strategy(§7.1) Stage 1: Explore Stage 2: Experiment Stage 3: Improve Stage 4: Transform Stage 5: Innovate
- 13. Cognitive Chasms (§7.2) GenAIAdoption Failure Modes Represent the potential discontinuities in adoption 4 distinct chasms represent exponential risks and rewards
- 14. Hype / Reality (H/R)Chasm • "Expectations Gap" • Right technology must work Pilot / Production (P/P) Chasm • “Scaling Gap” • Pilots must be scalable Technology / Business (T/B) Chasm • “Value Gap” • Business need must exist Business / Social (B/S) Chasm • “Safety Gap” • Societal interests sacrosanct Cognitive Chasms (§7.2)
- 15. Multistage Scaling andConsequents: Stage 1: Explore Stage 2: Experiment Stage 3: Improve Stage 4: Transform Stage 5: Innovate
- 16. Multistage Scaling andCognitive Chasms Hype / Reality (H/R) Chasm Pilot / Production (P/P) Chasm Technology / Business (T/B) Chasm Business / Societal (B/S) Chasm
- 17. Human Mindset Decision Mindset PurposeMindset Cognitive Mindsets are sequential, accretive, and cumulative Learning Mindset
- 18. Bridging The Chasm:Multistage Scaling, Cognitive Chasms and Cognitive Mindsets Human Mindset Decision Mindset Purpose Mindset Learning Mindset
- 19. More details • Theoryof Cognitive Chasms: A Grounded Theory of GenAI Adoption: My doctoral dissertation at Indian School of Business (ISB): https://eprints.exchange.isb.edu/id/eprint/2340/ • GenAI in Business: Invited talk for Emerging Leaders Program at ISB, May 2025:- https://www.slideshare.net/slideshow/generative-artificial-intelligence-genai-in-business/278708699 • Theory of Cognitive Chasms: Failure Modes of GenAI Adoption: Lecture for Executive MBA students at Walton School of Business, Apr 2025: - https://www.slideshare.net/slideshow/theory-of-cognitive-chasms-failure-modes-of-genai- adoption/278463456 • PhD Colloquium at Indian Institute of Management, Bangalore (IIMB), Dec 2024: https://www.youtube.com/live/UBmkl7Vpr0Q?si=lqw2JJUVhUI8cWL7 • GenAI Value Spectrum – delivering the “true value” from GenAI initiatives: Keynote at IEEE Bangalore Technology Conference (BTC) 2024: https://www.slideshare.net/slideshow/genai-value-spectrum-delivering-the-true-value-from-genai- initiatives/273734671 • The role of Cognitive Mindset in GenAI Adoption: Invited talk at Project Management Institute (PMI) Bangalore Chapter Footprints: https://www.slideshare.net/slideshow/the-role-of-cognitive-mindset-in-genai-adoption-pdf/274055306 • Managing Fast-Evolving GenAI Adoptions: Invited Keynote at Siemens Agile Conference 2024: https://www.slideshare.net/slideshow/managing-fast-evolving-genai-technology-adoptions/273470070
- 21.