Unpacking AI investment giant Sequoia Capital’s declaration: ‘AGI is here, now’

Sequoia Capital, arguably the most influential technology investment firm in the world, recently declared that artificial general intelligence had been achieved. 

There were no bells or whistles accompanying this announcement. This wasn’t the arrival of the singularity that Kurzweil and others have predicted; it was the supposed revelation that AGI is already here. 

Per a Sequoia blog post:

“While the definition is elusive, the reality is not. AGI is here, now. Coding agents are the first example. There are more on the way. Long-horizon agents are functionally AGI, and 2026 will be their year.”

It’s hard to overstate the impact this statement could have on the technology industry. According to data from Tracxn, Sequoia Capital has invested in 1638 companies as of January, 2026, with 53 new investments in the past year alone. 

It’s invested in big tech firms including Apple, Google, and Meta. Beyond that, it’s also heavily invested in the generative AI stack from foundational development, with large investments in OpenAI, xAI, and SSI, to the application and integration layers with investments including Hugging Face, LangChain, Pinecone, and Sierra. 

Overall, data indicates the firm has supported the launch of 130 unicorns (startups reaching $1 billion valuations), 119 IPOs, and 412 acquisitions. 

This positions Sequoia as one of the tech world’s most important gatekeepers. And, as such, the declarations made by its analysts and leaders carry legitimate financial weight throughout the tech and investor worlds.

A rose by any other name?

Sequoia Capital’s Pat Grady and Sonya Huang, the authors of the aforementioned blog post declaring AGI has arrived, are careful to point out that they’re defining AGI as investors. 

“We do not have the moral authority to propose a technical definition of AGI,” they write, adding that “we are investors. We study markets, founders, and the collision thereof: businesses. Given that, ours is a functional definition, not a technical definition.”

They sum up by saying: “AGI is the ability to figure things out. That’s it.”

Sequoia’s definition isn’t too far off from the one that we tend to cite here at AGI Ethics News, IBM’s:

“Artificial general intelligence (AGI) is a hypothetical stage in the development of machine learning (ML) in which an artificial intelligence (AI) system can match or exceed the cognitive abilities of human beings across any task.”

But the key part of IBM’s definition is the phrase “across any task.” Those three words almost always spell the difference between hype and reality.

Sequoia bases their argument on the notion that AI models can currently produce code at a human level (many developers disagree: see here, here, here, here, here, here, here, and here). 

Grady and Huang claim that the progress made in developing large language models that can code as well or better than a human developer indicates that AGI is already here. From this point, they say, “soon you’ll be able to hire an agent.”

They point out that you can “hire” GPT-5.2 or Claude or Grok or Gemini now and state that the following agents are “on the way”:

• Medicine: OpenEvidence’s Deep Consult functions as a specialist
• Law: Harvey’s agents function as an Associate
• Cybersecurity: XBOW functions as a pen-tester (ed: a “pen-tester” is industry jargon for “penetration tester,” a program or agent designed to simulate attacks on networks)
• DevOps: Traversal’s agents function as a Site Reliability Engineer (SRE)
• GTM: Day AI functions as a BDR, SE, and Rev Ops leader (ed: GTM refers to go-to-market intelligence, BDR, SE, and Rev Ops represent “Business Development Representatives,” “Sales Enablement,” and “Revenue Operations,” respectively)
• Recruiting: Juicebox functions as a recruiter
• Math: Harmonic’s Aristotle functions as a mathematician 
• Semiconductor Design: Ricursive’s agents function as chip designers 
• AI Researcher: GPT-5.2 and Claude function as AI researchers

Instead of breaking down each of the above bullets and presenting case-by-case counterarguments from scientists and technology experts, I’ll let Grady and Huang’s own words summarize the problem:

Per the Sequoia blog post: 

“To be clear: agents still fail. They hallucinate, lose context, and sometimes charge confidently down exactly the wrong path. But the trajectory is unmistakable, and the failures are increasingly fixable.”

It’s unclear what Sequoia means by “increasingly fixable,” but this reasoning underpins the growing movement in the AI development community to declare AGI has already arrived. 

Those who subscribe to this idea believe that the hardware necessary to produce AGI has already been engineered and that human-level artificial intelligence is now purely a tuning and alignment issue. 

Many of the researchers pushing this narrative, however, are employed by companies whose bottom line depends on LLMs being defined as AGI. 

Anthropic co-founder Daniela Amodei, for example, recently told Forbes “AGI is such a funny term because many years ago it was kind of a useful concept to say: when will artificial intelligence be as capable as a human, and what’s interesting is, by some definitions of that, we’ve already surpassed that.”

Related: AGI and Moral Intelligence

What does it mean to be as capable as a human? Per Sequoia and Amodei, the capabilities of current AI tools surpass that threshold. AGI is, by their definition, already here.

But, if we refer back to IBM’s definition, we’re left wondering whether an AI system can truly be considered an AGI by simply being better than humans at some or most tasks. 

Ethical concerns related to this article:

• Who should set the standards for the validation and acceptability of AGI-driven breakthroughs, especially in domains where rapid acceleration is expected (e.g. biology, economics)?
• What is the ethical duty of governments and corporations to citizens whose cognitive labor is rendered obsolete by AGI?
• How can we ensure that AGI development is aligned with the broader interests and values of global society, rather than those of a narrow group?

Tristan Greene

A veteran AI journalist, Tristan was the creator and managing editor of The Next Web’s “Neural” imprint. He is one of the first and most prolific reporters to cover artificial intelligence as a full-time beat.