📊 Full opportunity report: The Ghost Story Became a Forecast. on ThorstenMeyerAI.com — validation score, market gap, and execution plan.
TL;DR
Jack Clark’s latest essay presents a 60% probability of automated AI R&D by 2028, but also a 40% chance that current paradigms hit fundamental limits, requiring new invention. This reframes expectations for AI timelines and risks.
Jack Clark’s recent essay introduces a bivalent forecast for AI development, assigning a 60% probability to automated AI R&D by the end of 2028, and a 40% chance that current technological paradigms reveal fundamental limitations, requiring new approaches. This shift in outlook has significant implications for AI research, policy, and industry planning.
In his essay, Clark explicitly states a 60% probability that automated AI research will be achieved by 2028, with a central forecast of reaching this milestone within that timeframe. He also assigns a 40% probability that progress will stall due to unrecognized fundamental limitations in existing paradigms, such as compute, data, or architectural constraints, which would necessitate new invention to advance.
This bivalent forecast marks a departure from traditional linear projections, emphasizing a structural insight: if AI does not reach automation by 2028, it may indicate that the current paradigm is incomplete or flawed, rather than simply a matter of slower progress. Clark also provides a 30% probability that automation could arrive as early as 2027, based on corporate commitments and technological trajectories, highlighting the uncertainty and high stakes involved.
The ghost story
became a forecast.
Reading Clark’s closing — the bivalent 60%/40% credence. The 30% by 2027 alternative. What it means when a frontier-lab co-founder publicly says “I’m persuaded.”
Jack Clark’s closing section — “Staring into the black hole” — contains the most important sentence in the essay for the public discourse. Not the 60%/2028 number — though that’s the technical claim that gets quoted. The discourse-crossing sentence is the personal credence statement: “I have written this essay in an attempt to coldly and analytically wrestle with something that for decades has seemed like a science fiction ghost story. Upon looking at the publicly available data, I’ve found myself persuaded that what can seem to many like a fanciful story may instead be a real trend.”
The standard discourse reads 40% as benign — “slower AI.” Clark’s actual claim is stronger. The 40% reveals a fundamental deficiency within the current technological paradigm. Both outcomes are major findings. The franchise has read the 60% side. The coda reads the 40% side and the bivalence itself.
“For decades, it has seemed like a science fiction ghost story.“
The most important sentence in the essay is not the 60% number. The discourse-crossing sentence is the personal credence statement. When a frontier-lab co-founder publicly says “I am persuaded by the data that this is no longer science fiction,” the discourse changes.
“I have written this essay in an attempt to coldly and analytically wrestle with something that for decades has seemed like a science fiction ghost story. Upon looking at the publicly available data, I’ve found myself persuaded that what can seem to many like a fanciful story may instead be a real trend.”
Scoping Reviews Unlocked: A Six-Step Cookbook for Beginners: Mastering AI-Enhanced Literature Analysis and Writing Review Papers (Research & Development Methods)
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Nine pieces. One structural finding.
Six different forms of evidence aggregating to one structural finding: the labs are building what they say they’re building; the forecast is the plan; the institutional response window is the only variable that remains unfixed.
Six different forms of evidence. One structural finding. The labs are building what they say they’re building. The institutional response window is the only variable that remains unfixed.
The AI Agent Starter Kit: Build, Deploy, Automate: A No-Code Guide to Creating Smart AI Assistants for Your Business, Brand, or Life
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Three paths. All major. All need capacity.
Three structural possibilities for what the next 32 months produce. Asymmetric cost-of-being-wrong points toward building response capacity now. There is no scenario where the capacity goes unused.
~20 months
~32 months
field correction
Capacity built for 30%/60% paths is useful. Capacity built for 40% path is also useful (for field correction). There is no scenario where building response capacity now is wasted.
Clark stares into the black hole and says he’s persuaded. The franchise has been about reading that statement seriously. The reading: he should be. The implication: so should we.
Modern Generative AI with ChatGPT and OpenAI Models: Leverage the capabilities of OpenAI's LLM for productivity and innovation with GPT3 and GPT4
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Implications of the Bivalent AI Forecast for Industry and Policy
This forecast challenges the common assumption that slower progress means merely delayed AI timelines. Instead, Clark’s analysis suggests that hitting a ceiling in capability growth could reveal fundamental limits in current AI paradigms, prompting a potential paradigm shift. Such a realization would fundamentally alter research priorities, investment strategies, and regulatory approaches, as stakeholders recognize the need for new technological foundations.
Understanding whether AI development is on a trajectory toward rapid automation or facing fundamental barriers is crucial for policymakers, industry leaders, and researchers planning for the future. The 40% probability underscores that a significant structural change in AI could be imminent, with profound implications for societal impact and technological innovation.
The Scaling Era: An Oral History of AI, 2019–2025
As an affiliate, we earn on qualifying purchases.
As an affiliate, we earn on qualifying purchases.
Background on Clark’s Probabilistic Forecasts and Paradigm Shifts
Jack Clark’s essay builds on his prior analyses of AI progress, emphasizing that traditional forecasts often overlook the possibility of paradigm shifts. His recent framing introduces a bivalent outlook: either AI will rapidly reach automation milestones, or progress will stall, revealing limitations in current approaches. Clark’s analysis is informed by recent corporate targets, technological bottlenecks, and the history of AI development, which has often seen periods of rapid progress followed by plateaus.
This perspective aligns with broader debates about the sustainability of current AI paradigms, which rely heavily on compute, data, and architectural improvements. Clark’s framing underscores that the future may depend on whether these paradigms are fundamentally sound or inherently limited, a question that remains open as of May 2026.
“The 40% probability that we hit a fundamental ceiling means we may need entirely new approaches to AI development, not just incremental improvements.”
— Jack Clark
Uncertainties Surrounding the Paradigm Shift and Timeline
It remains unclear whether the 40% probability will materialize as a fundamental limitation or if technological breakthroughs will overcome current bottlenecks. The precise nature of these potential limitations—whether in compute, data, or architecture—is still under investigation, and the timeline for any paradigm shift is uncertain. Additionally, corporate targets and technological progress are subject to change, adding further unpredictability.
Next Steps for Researchers and Policymakers in Light of the Forecast
Stakeholders should prepare for both scenarios: continued rapid progress or a paradigm shift requiring new foundational research. Monitoring corporate commitments, technological breakthroughs, and paradigm debates will be critical in the coming months. Further analysis of emerging bottlenecks and paradigm indicators is expected to clarify which trajectory is unfolding, shaping future AI policy and investment.
Key Questions
What does Clark’s bivalent forecast mean for AI timelines?
It suggests there is a 60% chance AI automation will occur by 2028, but also a 40% chance that progress will stall due to fundamental limitations, potentially delaying or altering the development path.
Why is the 40% probability significant?
This figure indicates that a major paradigm shift could be imminent, meaning current approaches might be inherently limited and require new foundational inventions, not just incremental improvements.
How should policymakers respond to this forecast?
Policymakers should consider both scenarios—accelerated automation and fundamental limits—and prepare flexible strategies that can adapt to either outcome, including supporting foundational research and regulatory frameworks.
What is the basis for Clark’s confidence in these probabilities?
Clark’s assessment is based on recent corporate targets, technological bottlenecks, and historical patterns of AI development, combined with his analysis of current paradigm limitations.
When will the actual trajectory become clearer?
Monitoring ongoing corporate milestones, technological breakthroughs, and paradigm debates over the next 12-24 months will be essential to determine which scenario is unfolding.
Source: ThorstenMeyerAI.com
