AI: The Final Act of the Digital Revolution
The current surge in AI may represent the final, maturing phase of the digital revolution rather than the beginning of a new technological era. This 'late-cycle' stage is marked by big tech dominance, platform saturation, and the optimization of existing systems through efficiency gains. As a result, the future of AI may favor pragmatic, industrial applications over the speculative 'AGI' goals currently pursued by US firms.
Key Points
- AI may be the final stage of the ICT surge that began in 1971, acting as an efficiency breakthrough rather than a new paradigm.
- The dominance of well-funded incumbents over garage startups indicates that the 'installation' phase of infrastructure is over and 'deployment' has matured.
- AI is being used to conquer the final resistant sectors of the physical economy, such as healthcare delivery and government services.
- Economic returns in this late stage are likely to be 'normal' rather than the 'increasing returns' seen during the early growth of the digital era.
- China's focus on lean, industrial AI applications may align more closely with late-cycle realities than the American pursuit of AGI.
Sentiment
The community is notably anxious and divided. A significant plurality sympathizes with the article's skepticism about AI as a revolutionary new force, but often for personal economic reasons (job displacement fears) rather than agreement with the Perez framework specifically. There is strong pushback from those who see AI as genuinely transformative, particularly in areas like robotics and biotech. The overall tone tilts pessimistic, with recurring themes of skill erosion, enshittification of digital spaces, and corporate cost-cutting masquerading as innovation. However, a vocal minority maintains that the discussion conflates LLMs with all of AI and underestimates the broader transformation underway.
In Agreement
- AI investment is being driven by incumbents deploying massive capital rather than scrappy startups, consistent with a mature optimization phase rather than a new technological revolution
- Consumer willingness to pay for AI products remains low — the primary revenue model is corporate cost-cutting and layoffs, not new consumer demand
- AI will cause margin compression across the software industry: if everyone adopts it, competitive pressure erodes the productivity gains, leading to consolidation rather than growth
- AI is destroying Silicon Valley's traditional capital-light, moat-protected business model by shifting to capital-heavy infrastructure vulnerable to open-source and Chinese alternatives
- The proliferation of AI-generated content is already degrading online experiences (shopping, search, content quality), driving people back to physical alternatives
- Renewable energy and the solar-battery-EV stack may be the actual transformative technology surge of this era, with AI being overhyped by comparison
Opposed
- US AI investment patterns actually resemble Perez's installation phase (speculative, financialized bets) rather than late-cycle deployment optimization, suggesting AI may be a new surge after all
- The article's biggest omission is robotics, which is experiencing explosive growth using the same AI compute infrastructure and could represent genuinely new capabilities
- AI has enabled genuinely novel capabilities like protein folding solutions that were previously impossible, not just faster versions of existing processes
- Humanity has industrialized the production of intelligence — this is a fundamental change that transcends the digital wave framework entirely
- AI's rapid adoption rate by non-technical users far exceeds that of personal computers, suggesting something more transformative than late-stage optimization
- Previous systems helped humans work; AI systems can actually do the work — this is a qualitative difference that may require entirely new economic frameworks