Awesome Ventures Logo

AI's Unprecedented Acceleration: A VC's Smart Brevity Breakdown of BOND Capital's 2025 AI Report

By YekSoon LOK
AI Artificial Intelligence Tech Trends Venture Capital Investment Strategy BOND Report Geopolitics Future of Work

BOND Capital’s May 2025 “Trends – Artificial Intelligence” report paints a vivid picture of an industry in hyper-acceleration, creating both immense disruption and unparalleled opportunities. This isn’t just about new features; it’s about new market creation and the redefinition of existing ones.

As a VC, these are my key takeaways for our community of investors and founders:

The 1 Big Thing: Artificial Intelligence is reshaping the world at an unprecedented velocity. This transformation is driven by exponential growth in users, data, compute power, and investment, compelling a rapid re-evaluation of business strategies, the nature of work, and global geopolitical landscapes.

Why It Matters: The current AI wave transcends typical technology cycles. Its sheer speed and breadth are simultaneously generating immense opportunities and acute challenges. We’re witnessing the redefinition of entire industries and the fundamental nature of work, alongside an intensification of global power competition, particularly between the USA and China. Leadership in AI is poised to dictate economic and geopolitical dominance for decades to come.

By the Numbers: AI’s Exponential Trajectory

The report highlights several staggering metrics:

  • 800MM: Weekly active users for a leading USA-based LLM (ChatGPT) were reached in just 17 months—an 8x growth (p5, p55).
  • $212 Billion: Projected CapEx by “Big Six” USA tech companies in 2024, up +63% Y/Y, largely fueled by AI infrastructure demands (p5, p102).
  • +448%: Growth in USA AI job postings over 7 years, while non-AI IT job postings experienced a -9% decline (p7, p332).
  • 99.7%: Drop in cost per 1 million tokens for leading AI model inference over a mere two years (p6, p137), complemented by a >100,000x decline in energy per token for NVIDIA GPUs over the past decade (p136).
  • 25% & Rising: NVIDIA’s commanding share of global data center CapEx revenue (p109).
  • +1,150%: Surge in the release of large-scale multimodal AI models over two years, with China increasingly taking the lead in open-source contributions (p252, p264-266).
  • 122 Days: The remarkable time taken for xAI to construct its “Colossus” data center, compared to an average of 234 days for a US home (p122).

The Big Picture: A Foundational Technology Shift

The BOND report compellingly frames AI as a foundational technology shift, occurring at a pace and scale that outstrips previous revolutions like the internet or mobile. Key emerging themes include:

  • Change Happening Faster Than Ever: AI adoption, model development, and infrastructure build-out are progressing at breakneck speed.
  • Unprecedented User, Usage & CapEx Growth: Massive user uptake (notably ChatGPT) is mirrored by skyrocketing capital expenditure on compute power and data centers.
  • Compute Cost Dynamics: While training sophisticated AI models remains incredibly expensive (and costs are rising), the cost of using these models (inference) is plummeting. This dynamic is fueling wider adoption and unlocking a plethora of new applications.
  • Monetization Threats & Opportunities: The landscape is complex, characterized by intense competition, the proliferation of open-source models (especially from China), and high operational burn rates for AI companies. Incumbents are leveraging vast user bases and substantial cash reserves, while agile new entrants are innovating at a furious pace.
  • Physical World Integration: AI is rapidly moving beyond purely digital applications and into robotics, autonomous vehicles, scientific discovery (drug development, materials science), and defense systems.
  • Work Evolution: AI is profoundly transforming workflows, creating entirely new job categories, and demanding a rapid upskilling of the workforce, with significant implications for productivity.
  • Geopolitical Race: The USA and China are locked in an acute race for AI supremacy. This competition impacts everything from semiconductor manufacturing and model development to data sovereignty and ethical guidelines.

Dive Deeper: Key Areas of Focus from the Report

For those looking to explore further, the report offers rich detail in several areas:

  • The AI Arms Race (USA vs. China): The report dedicates significant analysis to this escalating competition, covering AI development, model releases (especially open-source), compute capabilities, and strategic applications like industrial robotics. China’s rapid catch-up and potential leadership in certain domains is a recurring and critical theme. (See “AI Monetization Threats,” p248-298 in the BOND report)
  • The Cost-Benefit Equation: Explore the inherent tension between soaring AI model training costs and CapEx on one hand, versus the dramatically falling inference costs on the other. This paradox is democratizing access to AI capabilities but simultaneously creating intense pressure on the business models and profitability of AI providers. (See “AI Model Compute Costs High / Rising…”, p129-152 & “AI Usage + Cost + Loss Growth,” p153-182)
  • Incumbents vs. New Entrants: Analyze how established tech giants are leveraging their scale, cash flow, and distribution channels to integrate AI, versus the rapid growth and high valuations of specialized AI startups and foundation model builders aiming to disrupt them. (See “AI User + Usage + CapEx Growth” sections on tech incumbents, and “AI Monetization” sections on new entrants, p62-77, p187-228)
  • Data Centers: The New Factories: The report underscores the critical role and explosive growth of data centers as the physical backbone of the AI revolution, detailing construction speed, capacity expansion, and enormous energy consumption. (See “CapEx Spend – Big Technology Companies = Inflected With AI’s Rise,” p99-128)
  • The Future of Work with AI: The profound impact on job creation, potential displacement, and overall productivity is explored, highlighting a surge in AI-related job postings and early evidence of AI boosting worker output across various sectors. (See “AI & Work Evolution,” p323-336)

Why This Matters to Me

The unprecedented speed (“Change Happening Faster Than Ever”) and the “AI & Physical World Ramps” detailed in the report are particularly compelling from an investment standpoint. While the CapEx figures involved in building foundational AI are indeed eye-watering, the plummeting inference costs are crucial—they are democratizing access to powerful AI tools and will undoubtedly unlock a wave of innovation in applications and industries we haven’t even conceived of yet.

The core challenge, and therefore the opportunity for astute VCs and founders, lies in identifying and building sustainable business models amidst this AI gold rush.

What specific AI-driven disruption are you most excited (or concerned) about in your sector?

About the author
YekSoon LOK is the Founding Partner of Awesome Ventures, backing visionary early-stage teams in ClimateTech, FinTech, and frontier markets. His portfolio includes Twilio, RedHat, Cloudflare, and early positions in BTC, ETH, and XLM. He writes about venture strategy, emerging tech, and founder insight. Connect at LinkedIn.