Will any company announce that it has achieved Artificial General Intelligence (AGI) before Oct 1, 2026?
When I first saw the market for whether any company will announce Artificial General Intelligence (AGI) before October 1, 2026, I genuinely paused. An 11% chance? Really? My immediate thought was, "Are people actually serious about this?"
Here's the deal on Kalshi: the market asks, "Will any company announce that it has achieved Artificial General Intelligence (AGI) before Oct 1, 2026?" Right now, the 'YES' side is trading at 11%, meaning bettors believe there's an 11% probability of this happening. On the flip side, the 'NO' side is at a dominant 88%. This isn't some niche market; we've seen a solid 7,250 contracts traded, showing real interest and some serious money on the line. The open interest, at 3,639 contracts, tells me people aren't just making quick trades; they're holding positions, ready to see this through until the market closes on October 1, 2026.
So, what does that 11% really mean? It means if you bet $100 on 'YES' and a company does announce AGI by the deadline, you'd net a cool $809 profit. Pretty enticing, right? But the 88% 'NO' price is saying, in no uncertain terms, that the overwhelming majority of participants think this is a long shot. And frankly, I tend to agree with the majority here, though that 11% still niggles at me. It's not zero, and that's what's fascinating.
My read on this is pretty clear: I just don't see it happening within the next two years. I know we've seen absolutely mind-boggling advancements with large language models and multimodal AI lately. Every few months, it feels like a new model shatters some benchmark we thought was years away. But AGI? That's a different beast entirely. We're talking about a system that can understand, learn, and apply intelligence across a wide range of tasks at a human or superhuman level. That includes common sense reasoning, abstract thought, creativity, and the ability to adapt to novel situations without specific retraining. That's a tall order.
Think about it. Even the most ambitious figures in AI, like OpenAI CEO Sam Altman, have often spoken about AGI being more in the realm of decades than mere years. I remember him explicitly stating in a 2023 interview that while progress is rapid, achieving true AGI is an undertaking that will likely span many more years, citing the immense scientific and engineering challenges. We're not just iterating on existing models; we're talking about fundamental breakthroughs in architecture, learning paradigms, and even our understanding of intelligence itself. Current frontier models, while astonishing, still exhibit glaring weaknesses in robust reasoning, long-term memory, and hallucination control. We're great at predicting the next word; we're not yet great at predicting the next scientific discovery or writing a novel with genuine human insight.
Then there's the sheer, ungodly cost of compute. Training the largest current models already demands hundreds of millions of dollars in specialized hardware and electricity. For a system truly deserving of the AGI label, we're likely talking about orders of magnitude more, pushing into the billions, potentially trillions, for development and deployment. The infrastructure simply isn't there yet, and building it out takes time, even with all the money pouring into the sector. So, for a company to not only achieve AGI but also announce it and have it be recognized as such within this tight timeframe feels incredibly ambitious, bordering on unrealistic.
So, who's betting 'YES' at 11%? I figure it's a mix. Maybe it's the true believers, the optimists who think an 'alpha' moment is around the corner. Perhaps it's people who believe a company might declare they've achieved AGI, even if the broader scientific community disagrees on the definition. That's the tricky part, isn't it? The definition of AGI itself is somewhat fluid. One company's 'AGI' might be another's 'highly advanced narrow AI.' If a major player like Google DeepMind or OpenAI makes a bold claim, even if it's contested, it could trigger this market. That's a risk. I imagine some of those 'YES' bettors are banking on a definitional dispute, or a PR-driven announcement designed to boost stock prices or attract talent.
But for me, putting my money where my mouth is, I'm siding with the 88%. The technical hurdles are just too high, and the timeframe is too short. While I'm as excited as anyone about AI's progress, I think we're a good deal further out from general intelligence than October 2026. Iām going to be watching the 'NO' side to continue its dominance, and probably adding a few contracts to it myself. It feels like a relatively safe bet, and sometimes, the best trades are the ones where the crowd has it mostly right.
