We are on the verge of something extraordinary: humanity’s final invention.
That invention is Artificial General Intelligence, and this AGI timeline is my year-by-year map of how we get there. I first published it in 2024. I’m updating it now because something unusual has happened since: enough of the timeline has played out that I can grade my own predictions.
Some aged well. One arrived faster than I expected, and it’s already showing up in layoff reports.
AGI promises to reshape our world in ways we’re just beginning to grasp. We’re headed toward a jobless society with global dematerialization, and when automation replaces human labor, capital itself will have to change.
Here’s where we actually stand, with receipts.
What Is AGI, Humanity’s Last Invention?
AGI is the point where machines achieve cognitive abilities on par with humans across the full range of tasks, a convergence of automation, AI, and robotics operating at warp speed.
It’s called humanity’s last invention because an intelligence that matches ours can take over the inventing. Some experts believe it will quickly exceed human performance and lead to artificial superintelligence (ASI), an intelligence far beyond us.
If it lives up to the hype, AGI could accelerate scientific discovery in medicine, energy, and longevity in ways that change the human trajectory itself. Imagine AGI developing life-extending therapies or pioneering new energy sources.
I’ve covered AGI extensively on my YouTube channel: how it evolves from narrow AI, the job losses that come with full automation, and how to prepare for the age of robotics. So how close are we, really?
The AGI Timeline, Year by Year
Researchers estimate wildly different timelines, from 2 years to 20. Here’s my breakdown, updated with what has actually happened through mid-2026.
2024: Laying the Groundwork (What Happened)
2024 delivered the reasoning breakthrough. OpenAI’s o1 showed that models could think step by step before answering, and every major lab followed with reasoning models of their own.
The humanoids showed up too. Tesla’s Optimus went from barely walking in 2022 to folding laundry, Agility’s Digit worked Amazon warehouse pilots, and Figure partnered with OpenAI to give a robot conversational ability.
I said 2024 was groundwork, and it was. The groundwork just got poured faster than almost anyone outside the labs expected.
2025: Benchmarks and the AI Operator Economy (What Happened)
I predicted 2025 would bring better benchmarks, enterprise adoption, and a surge in hiring for AI optimizers, drivers, and operators.
The operator prediction hit dead center. AI-skilled roles became some of the fastest-growing job postings in the market, and I built a whole company on training exactly these people.
GPT-5 shipped. Hallucinations improved rather than disappeared, which kept some regulated industries cautious, but enterprise adoption surged anyway. OpenAI reported majority adoption across the Fortune 500 by late 2025.
What I underestimated: the layoffs started here, not in 2028. More on that in a moment.
2026: Integration Everywhere (Happening Now)
I predicted deep industry integration, new frontier models like Claude 5, smart home AI, and AGI entering daily conversation. Scorecard so far: Claude 5 is here, agents run real workflows inside businesses, and 1X began piloting its Neo humanoid in actual homes.
The robots crossed from demo to deployment this year. Figure passed 1,000 units of its Figure 03 robot produced at its BotQ factory in July, after an 11-month pilot on BMW’s Spartanburg line. Boston Dynamics’ electric Atlas is working at Hyundai plants. Tesla has converted Fremont’s Model S/X line into its first dedicated Optimus factory, says Gen 3 production is anticipated later this year, and Musk has targeted consumer sales for the end of 2027.
And AGI is absolutely part of the daily conversation now. It dominated Davos in January. It’s in earnings calls, Senate hearings, and my comment section.
The prediction I’d revise: the job displacement I penciled in for 2028 showed up 2 years early. Challenger, Gray & Christmas recorded AI as the single most-cited reason for U.S. layoffs for 4 straight months in 2026, peaking at 38,579 AI-attributed cuts in May alone. Block cut 40% of its workforce, with the CEO stating plainly that smaller teams using AI tools can do more.
2027: The Turning Point (Prediction)
2027 remains the earliest year I believe we get AGI, and the labs keep pointing at the same window. Dario Amodei told the World Economic Forum in January that AGI-level systems will likely arrive within a few years, possibly sooner than expected, driven by AI systems accelerating their own development.
When will we reach AGI? Here’s what the predictions have been saying throughout the years:
DeepMind cofounder Shane Legg puts a 50% chance on minimal AGI by 2028.
Here’s where it gets interesting, and this part of my original post has only gotten stranger: OpenAI increasingly talks about superintelligence as if AGI is a settled matter. Sam Altman has even called AGI “not a super useful term” while the company builds toward what comes after it.
I’ve seen this question in forums and on my own channel constantly: have these labs already uncovered AGI internally? Are they holding it until society is ready, because dropping it cold could cause total market collapse and the rebirth of society itself? Or is “superintelligence” just their word for the thing the rest of us call AGI?
Only time will tell. But watch what the labs build, not just what they announce.
2028: Governments Forced to Figure It Out (Prediction)
2028 is the year governments run out of room to postpone the AI question, and in the US it’s an election year, which guarantees the issue lands on every debate stage.
The early numbers make the stakes concrete. A National Bureau of Economic Research survey of 750 CFOs projects roughly 502,000 U.S. roles lost to AI in 2026 alone, about half of them white-collar. That’s 9x the prior year, and we’re still in the warm-up phase of this AGI timeline.
We’re going to need a mass societal transition from a traditional labor economy to something new, and the planning has to start before the displacement peaks, not after. My hope is that world governments come together with technology leaders, founders, authors, and innovators to design it.
2029: The Quantum Leap (Prediction)
If 2027-2028 plays out, 2029 is when humanity grapples with AGI’s full impact at global scale.
Expect gigantic data centers, serious quantum computing and fusion milestones, cybernetics, mass robot integration, and early longevity escape velocity conversations moving from fringe to mainstream.
This could also be the year Rosie the Robot finally rolls into ordinary homes. I joked about that in 2024, and then 1X started home pilots in 2026, so the joke is aging into a forecast. Unitree already sells a humanoid for $16,000, which is used-car money.
2030 and Beyond: The Age of Superintelligence (Prediction)
By 2030, many predict a fundamentally different world: robotics everywhere, including our homes, and possibly the emergence of ASI.
The forecasting community has compressed hard toward this decade. Metaculus forecasters now put a 25% probability on AGI by 2029 and 50% by 2033, down from a median of 50 years away as recently as 2020.
Demis Hassabis says AGI’s impact will be 10x the Industrial Revolution, compressed into a decade instead of a century.
ASI, by definition, surpasses our comprehension. It may end up defining its own intelligence to us, making it something closer to alien intelligence than a smarter chatbot.
FAQ: The AGI Timeline
When Will AGI Arrive?
The honest answer is a range, not a date. Lab leaders cluster around 2027-2028: Amodei says within a few years, Shane Legg gives 50% odds by 2028, and Altman has said AGI will probably arrive during the current U.S. presidential term. Independent forecasters run later, with Metaculus at 50% by 2033 and academic surveys stretching to 2040. My own call stays where it’s been since 2024: 2027 is the earliest realistic year.
What’s the Difference Between AGI and ASI?
AGI matches human cognitive ability across essentially all tasks, meaning it can do what a capable human can do, including learning new domains. ASI exceeds human ability by a wide and growing margin, potentially improving itself faster than we can follow. Most experts who expect AGI this decade expect ASI to follow within a few years of it, because an AGI that can do AI research compresses its own development cycle.
Has AGI Already Been Achieved in Secret?
There’s no evidence any lab has working AGI behind closed doors, and the compute and staffing footprint of frontier training runs makes a full secret hard to keep. What is true: labs see capabilities months before the public does, and OpenAI’s shift to superintelligence language suggests they consider the AGI threshold close enough to look past. I treat the secret-AGI theory as unproven, but the capability gap between labs and the public as very real.
Will AGI Take My Job?
Some version of your tasks, almost certainly, and the data says it’s already started: AI was the top-cited reason for U.S. layoffs for 4 consecutive months in 2026. The people thriving are the ones who moved early into operating the tools rather than competing with them. Upskilling into an AI operator role is the single most concrete defensive move available right now.
The Time to Prep for AGI Is Now
I predict a messy rollout. In some instances, humans might not even know what to do when AGI arrives, and the transition will be hardest for the people who ignored it longest.
So my call to action to legislators, innovators, and founders stands: let’s build the foundation of a new society where the gains reach all of us, no matter our age, race, belief system, political party, or access to resources.
And my call to you is simpler. Get ahead of this while the window is open. AI Labs trains you to become the AI operator your industry will be hiring for, and our consulting team builds these systems inside businesses doing $250K+ a year.
It’s my belief that if just 1% more of the human race were aware of what’s coming, we’d see a more optimistic adaptation everywhere. Be part of the 1%.
