AGI Timeline

Nobody knows when artificial general intelligence will arrive. At Layer3Labs, we build AI systems inside small and mid-sized businesses, so we read these forecasts for what they change in a 12 to 36 month plan.

Published estimates from major artificial intelligence lab leaders cluster between 2026 and 2030, while a survey of 2,778 researchers places a 50 percent chance of machines outperforming humans at every task at 2047. Because target dates vary by decades, business leaders cannot base capital allocation or operational strategy on a single predicted year. Plan around model capabilities your team can test on current workflows, so a slipped forecast date does not break the budget.

How Far Away Is AGI? The Forecasts in One Table

Published AGI forecasts run from one to two years out to 2047. Lab leaders' targets cluster between 2026 and 2030, while the largest researcher survey put even odds of machines beating humans at every task at 2047.

Each row is what that person or group said, on the date shown.

What to Plan For Either Way

Plan on the AI you can test in the next 12 to 36 months, and treat any AGI date as unknown.

In the implementations we run for clients, the plans that hold up are built on what a model does on your own tasks this quarter.

To identify which internal workflows offer the highest return on investment, teams can run an AI workflow audit. For structured deployment steps, see our AI implementation timeline and our AI strategy framework. Leaders evaluating workforce impact can also review what jobs will AI replace.

• List the repetitive tasks in each team before buying anything
• Pilot AI on one workflow before rolling it out to a department
• Keep a person approving payments, contracts and client messages
• Re-test vendor claims every quarter on your own files
• Budget for model prices and features to change each quarter

Why the Forecasts Disagree

Forecasts diverge because each source defines the milestone differently. Demis Hassabis defines artificial general intelligence as a system possessing the full cognitive range of the human brain.

Dario Amodei talks about "powerful AI", which he described as "a country of geniuses in a datacenter" in his October 2024 essay Machines of Loving Grace. He repeated it in his January 2026 essay The Adolescence of Technology (Axios coverage). OpenAI's goals are functional. In October 2025 Sam Altman set an automated AI research intern by September 2026 and a true automated AI researcher by March 2028, adding, "We may totally fail at this goal."

The researcher survey asked about every task, a stricter bar. In the survey, published in the Journal of Artificial Intelligence Research, the 2047 date required unaided machines to beat humans at every task. Full automation of all occupations came out at 2116. The dates also move. Hassabis said five to ten years in January 2026 and, as reported, two to three years in October 2026. The AI Futures Project moved Daniel Kokotajlo's Automated Coder median from late 2029 to mid-2028. To explore core definitions, read what is AGI and review AGI vs ASI.

AGI vs Superintelligence Timelines

Forecasters usually place superintelligence after AGI, and none of the superintelligence statements below names a calendar year. In May 2023, OpenAI leadership wrote in a post on the governance of superintelligence that it was conceivable AI systems would exceed expert skill in most domains within ten years. The post gave no year for AGI.

In the October 2025 announcement that set OpenAI's March 2028 automated researcher goal, chief scientist Jakub Pachocki said deep learning systems could be less than a decade away from superintelligence.

Corporate forecasts should not be confused with federal naming conventions. Executive Order 14434, issued on September 29, 2026, renames artificial intelligence to Super Intelligence or SI across federal agency documents, but it relies on existing statutory AI definitions and makes no timeline or capability claims. To analyze these distinctions, review our guides on what is superintelligence, the Super Intelligence Executive Order explained, and AI 2027 explained.

What Researchers Expect Before 2030

The largest researcher survey rated several concrete, business-relevant milestones as at least even odds by 2028, long before its date for machines beating humans at every task. Grace et al. surveyed 2,778 researchers who published at top AI venues in October 2023, for "Thousands of AI Authors on the Future of AI". Respondents placed at least a 50 percent chance on multiple technical achievements becoming feasible by 2028, including building a payment processing site and fine-tuning a large language model (LLM).

That survey's 2047 date for machines beating humans at every task is 13 years earlier than the 2022 survey's. Lab leaders' timelines moved forward as well. Demis Hassabis told NBC News in 2025 that matching humans at any task was five to 10 years away, before putting AGI at two to three years in an October 2026 WSJ Magazine profile, as reported.

Whichever AGI date proves right, a 12 to 36 month plan can rest on the 2028 milestones above. Test those near-term capabilities on your own work by scheduling an AI workflow audit.

Signals Worth Watching

Dated milestones are better planning signals than predictions. Corporate milestones offer near-term verification points, such as whether OpenAI reports meeting the automated AI research intern goal it set for September 2026.

The AI Futures Project publishes timelines updates. In August 2026 it said that technological progress was tracking between 70 percent and 90 percent as fast as its original AI 2027 project scenarios. Future academic surveys will also capture whether working researchers pull their estimates earlier again, as the 2023 survey did against its 2022 predecessor.

Operators should prioritize models they can test on their own workflow files. When labs announce architectural advances or expanded reasoning capabilities, test those models against your own document extraction, customer routing, or code refactoring tasks to verify whether practical efficiency improves.

• OpenAI research milestones: monitor progress against the stated goals of an automated research intern and a full automated researcher
• AI Futures Project updates: track revisions to its Automated Coder median
• Academic literature: evaluate future multi-author surveys assessing whole-occupation automation thresholds
• Hands-on operational benchmarks: test newly released commercial models directly on internal company data