TL;DR
Forrester principal analyst J.P. Gownder says current data offer no clear evidence that AI has boosted productivity. His firm estimates AI could displace about 6% of jobs by 2030, but many corporate AI projects so far show limited measurable returns.
What happened
In an interview with The Register, Forrester vice president and principal analyst J.P. Gownder argued that contemporary productivity statistics do not show a measurable boost from AI adoption. He pointed to long-running productivity trends compiled by the US Bureau of Labor Statistics — including higher growth rates in earlier decades and slower growth in recent periods — as context for what he described as a weak link between information technology and productivity gains. Forrester’s research, which assessed roughly 800 job categories and 34 skills defined by the BLS and incorporated input from about 200 companies, estimates that automation technologies could eliminate roughly 6% of jobs by 2030 (about 10.4 million roles) through tools such as robotic process automation, business process automation, physical robotics and generative AI. Gownder also said many generative AI initiatives have not produced clear profit-and-loss benefits, and that some recent workforce reductions are cost-cutting moves rather than direct results of deployed AI.
Why it matters
- If AI is not yet raising measured productivity, economic expectations for rapid growth or wage gains tied to AI may be premature.
- Structural job losses from automation could persist even if productivity gains lag, affecting employment composition and worker retraining needs.
- Companies freezing hiring to test AI may temporarily reduce labor demand without delivering the intended automation benefits.
- Weak ROI on many AI projects suggests firms may face lost investments or delayed returns, influencing future adoption and corporate strategy.
Key facts
- Forrester estimates AI and related automation could displace about 6% of jobs by 2030, roughly 10.4 million positions.
- The research mapped roughly 800 job types and 34 skills as defined by the US Bureau of Labor Statistics and involved conversations with about 200 companies.
- Productivity growth cited from US BLS: about 2.7% per year (1947–1973), 2.1% (1990–2001), and 1.5% (2007–2019); an exception of ~2.8% annual growth occurred 2001–2007.
- Gownder referenced the Solow Paradox — the observation that technological advances (like PCs) did not show expected productivity gains in statistics — as still relevant today.
- Forrester used a methodology similar to the 2013 Frey and Osborne study to estimate automation susceptibility across job categories.
- Gownder cited studies suggesting many generative AI projects have not delivered measurable P&L benefits; he referenced an MIT study and McKinsey findings reported by participants.
- Researchers found some recent corporate layoffs were financial belt-tightening rather than immediate replacements by functioning AI systems.
- Gownder noted a pattern where jobs are cut with the intent to replace them later, sometimes followed by outsourcing to lower-cost locations.
What to watch next
- Upcoming US productivity releases and whether they show any sustained improvement tied to AI adoption — not confirmed in the source
- Trends in enterprise AI project ROI and whether the share of projects delivering measurable P&L benefit increases — not confirmed in the source
- Changes in hiring behavior (restarts of frozen roles or replacements) if AI pilots fail to meet expectations — not confirmed in the source
Quick glossary
- Productivity: A measure of output produced per unit of input, commonly used to assess economic performance over time.
- Robotic Process Automation (RPA): Software tools that automate routine, rule-based digital tasks previously performed by humans.
- Generative AI: AI systems that create new content—text, images, code or other media—based on learned patterns from training data.
- Automation potential: An assessment of how susceptible a job or task is to being performed by automated technologies.
- Solow Paradox: The observation that technological advances may be visible everywhere except in productivity statistics, named after economist Robert Solow.
Reader FAQ
Is AI already boosting productivity according to Forrester?
Forrester’s J.P. Gownder said current productivity statistics do not show a clear boost from AI.
How many jobs could AI displace by 2030?
Forrester estimates around 6% of jobs, about 10.4 million roles, could be uprooted by 2030.
Are recent corporate layoffs directly caused by AI?
Gownder said many cuts appear to be financial belt-tightening or hiring freezes to test AI, not always direct replacements by functioning AI systems.
Will jobs lost to AI return after economic rebounds?
Gownder argued that jobs replaced by automation tend to be lost structurally and do not typically come back, according to the interview.
Is outsourcing related to AI-driven job loss?
Gownder reported firms sometimes fire domestically citing AI and then hire lower-cost teams abroad; further details are not confirmed in the source.
AI + ML AI may be everywhere, but it's nowhere in recent productivity statistics Forrester principal analyst JP Gownder says jobs eaten by bots don't come back O'Ryan Johnson Thu 15 Jan 2026…
Sources
- AI may be everywhere, but it's nowhere in recent productivity statistics
- New Study Says AI Isn't Boosting Productivity (Yet)
- AI is everywhere except in worker productivity stats
- "AI is everywhere but in the productivity statistics…"
Related posts
- Under pressure to remove X, Apple hears xAI say Grok won’t undress people
- X says Grok can no longer undress people in images — tests suggest otherwise
- Anthropic Appears to Block ‘OpenCode’ in Claude OAuth System Prompts