Observed Signal · Aug 18, 2026 · Other · Source: UX Collective · Impact: 2/5 · Sentiment: Neutral
Purpose of Software in the Age of AI
Michael Burnett argues that despite enormous increases in computing power and visible efficiency gains, software has not produced sustained macroeconomic productivity growth. Reviewing historical economic research and productivity data (citing Solow, Krugman, Gordon and others), he contends the purpose of software is to increase the density of information in digital outputs rather than to raise value output per hour. Burnett revisits the "productivity paradox," discusses Baumol's cost disease, and hypothesizes a "disengagement curve" driven by search/chat workflows. He warns that AI and LLMs will act as force multipliers for software production but are unlikely to change software's fundamental relationship to measured productivity, even as they create more digital output and new efficiencies.
Analytical long-form perspective linking AI/LLMs to macro productivity debates; informs strategic thinking but does not announce platform policy, product releases, or market-moving transactions.
Track Medium Signals & Market Shifts in Real-Time
Polaris7 autonomous intelligence agents track regulatory filings, primary sources, executive changes, and deal flow 24/7. Create your Polaris7 workspace to monitor these entities.
Key Takeaways & Evidence Grounding
- Author Michael Burnett published a long-form essay arguing software has not increased macroeconomic productivity.
- Computing power at the frontier has increased by roughly 10^12 (a trillion-fold) since 1971, with transistor counts and FLOPS rising dramatically.
- Burnett cites economists and studies including Robert Solow, Paul Krugman, Robert J. Gordon, Federal Reserve analyses (Oliner & Sichel), Brookings, NBER, and others to support his analysis.
- He proposes the purpose of software is to increase the density of information per unit of digital output, not to increase productivity (value output per hour).
- The essay introduces the author's hypothesis of a 'disengagement curve' and revisits the 'Productivity Paradox' and Baumol's cost disease in the context of IT and AI.
Connected Companies & Entities
“Get Michael Burnett's stories in your inbox. Join Medium for free to get updates from this writer.”
“Paul Krugman, another Nobel winner, wrote in his NYT column in March 2023, Like previous leaps in technology, [AI] will make the economy mor...”
“ChatGPT generated this newspaper style cartoon in seconds.”
“Chart generated by Claude based on publicly available data points.”
“2023 (Apple M2 Ultra): 134 billion transistors”
“2018 (NVIDIA V100, as the frontier moved from CPU to GPU): ~125 TFLOPS. 2023 (NVIDIA H100): ~2,000 TFLOPS.”
“2018 (AMD Ryzen 7 2700X): 4.8 billion transistors”
“1971 (Intel 4004): 2,300 transistors. 1993 (Intel Pentium): 3.1 million transistors.”
“Has quality of life increased since the semiconductor was invented? We have Spotify, Netflix, LLMs, and very large TVs.”
“Has quality of life increased since the semiconductor was invented? We have Spotify, Netflix, LLMs, and very large TVs.”
“Gallup found that globally 41% of employees experienced a lot of stress the previous day in their 2024 report.”
“22% of US employees rate their level of burnout as high or very high according to Statista.”
Ontology Mapping & Concepts
Related Market Signals & Shifts
Recent verified developments and strategic activity across this market segment.
What Will More AI Intelligence Do?
The essay argues that although AI has achieved superhuman ability in narrow tasks (solving open math and cryptography problems), the broader economic and social impact has been more incremental than some expected. One hypothesis is that intelligence faces diminishing returns because the information extractable from data is bounded or costly to obtain; critics propose governance and frictions also slow change. The author highlights three mechanisms by which AI could still drive large productivity gains: replicability (running many agents in parallel), roboticization combined with energy/battery improvements, and AI’s ability to extract and diffuse distributed tacit knowledge or discover “cloud laws” — complex regularities humans cannot easily formalize. The piece cites examples (Zeiss/ASML mirrors, rare-earth refining) and surveys views from researchers including Francois Chollet and Arvind Narayanan.
Why AI Isn't Boosting Firm-Level Productivity
This essay argues that while AI and LLM tools (e.g., Claude Code, ChatGPT) have clearly increased individual and team-level productivity, those gains have not yet translated into proportional firm-level ROI. The author maps the phenomenon to a three-stage adoption ladder inspired by electrification: Stage 1 (lightbulb) improves individual tasks, Stage 2 (group drive) speeds workflows without changing organisational decision logic, and Stage 3 (unit drive) requires reorganising firms around faster decision-making and autonomous AI agents. Empirical signals cited include widespread internal use of Claude Code, fast-growing Anthropic enterprise spend, and surveys showing only 27% of executives report AI meeting ROI expectations. The piece concludes firms must redesign decision pipelines and allow AI agents to take certain decisions to realize full productivity gains.
AI Won't Replace Software, But Will Disrupt SaaS
The article examines claims that AI will replace traditional software, arguing that while AI agents and local personal systems (e.g., Clawbot) and platform releases (e.g., Anthropic's Cowork) increase automation and lower development friction, AI is unlikely to fully replace complex, large-scale software in the near term. The author contends AI will simplify development, push modular callable functions, and disrupt existing SaaS pricing (monthly and per-user models), user interfaces, and distribution (possible App Store‑like gatekeeping). Utility and plug‑in software are most at risk, while industry-specific and complex workflow software will retain value. The conclusion: AI changes profit models, use cases, and software architecture, but does not eliminate the need for sophisticated software systems.
Receive all future signals for Medium & competitors in your briefing
Never miss material M&A deals, leadership changes, account wins, or product rollouts. Polaris7 monitors 20,000+ primary sources 24/7, delivering verified telemetry directly into your executive briefing.
