In a significant move for the tech landscape, Pope Leo XIV, the first pope from the United States, released his inaugural encyclical titled *Magnifica Humanitas*, translating to "Magnificent Humanity," on Monday. This extensive 42,300-word document serves as a thoughtful prompt for the tech sector to reconsider its approach, emphasizing the importance of prioritizing human values over machine capabilities.

While acknowledging the potential benefits of artificial intelligence, the pope cautioned that such technologies often exacerbate existing inequalities by favoring those who already have wealth, knowledge, and access to vast data. He highlighted the dangers posed by a lack of oversight and transparency in AI development, warning that those in control could impose their own ethical frameworks, effectively shaping the unseen parameters of these systems.

Pope Leo XIV expressed that simply having a more ethical AI is not sufficient if the moral compass guiding its design is dictated by a select few. This message resonates with concerns about accountability and inclusivity in technological advancements.

The encyclical is anticipated to be a pivotal aspect of the pontiff's papacy, reminiscent of Pope Leo XIII's *Rerum Novarum*, which championed workers' rights and fair wages in the late 19th century. Released during a period of relative quiet in Silicon Valley, *Magnifica Humanitas* seeks to influence and potentially moderate the rapid growth of AI technologies.

Recently, former President Trump nearly enacted legislation that would enable the federal government to assess AI models prior to their public introduction but ultimately decided against it. On the same day, California Governor Gavin Newsom approved a measure aimed at investigating the implications of AI on employment, reflecting the state’s recognition of potential disruptions. Additionally, tech mogul Elon Musk attempted a legal maneuver against OpenAI—a leading player in AI development—yet faced setbacks this month due to technical legalities.

As excitement for A.I. grows in Silicon Valley, Pope Leo has some comments.

A warm, sunlit scene on a gently sloping hill: Pope Leo in his white cassock stands at a simple wooden lectern, speaking earnestly. Behind him, the rolling hills of Silicon Valley stretch out, their contours overlaid with faint circuit-board patterns. In the distance, a modern tech campus with a glowing “AI FEVER” billboard punctuates the landscape. Natural morning light, cinematic realism.

As A.I. Fever Rises in Silicon Valley, Pope Leo Has a Few Words - The New York Times

- Pope Leo XIV urges the tech industry to prioritize humanity over unchecked A.I. development.  
- The encyclical warns that A.I. amplifies power for those already wealthy and resourceful.  
- It emphasizes the need for oversight to prevent a narrow moral vision in A.I. systems.

Its message to the tech industry: Slow down. Elevate the human. Machines are not gods.

Pope Leo XIV, the first American pope, published with great ceremony on Monday his first encyclical, Magnifica Humanitas, or “Magnificent Humanity.” The 42,300-word policy statement is respectful and named no names, but is at heart a sharp rebuke to Silicon Valley’s assertions that it alone can be trusted to develop the future.

“A.I. can be a valuable tool,” the pope acknowledged, but the technology “tends to amplify the power of those who already possess economic resources, expertise and access to data.” Without adequate oversight and transparency, he warned, “those who control A.I. will impose their own moral vision, which will become the invisible infrastructure of these systems.”

That would be bad news, he said: “A more moral A.I. is not enough if that morality is determined by a few.”

The encyclical is expected to be at the center of the 70-year-old pontiff’s reign in the same way that Rerum Novarum, which advocated for workers’ rights and a fair wage, was at the heart of Pope Leo XIII’s papacy at the end of the 19th century. Released while Silicon Valley slept, Magnifica Humanitas marked the latest effort to shape and possibly restrain the A.I. boom.

President Trump last week came close to signing a measure that would have given the federal government the power to evaluate A.I. models before they were publicly released — then canceled the signing. The same day, California’s governor, Gavin Newsom, signed a measure to study A.I.’s impact on employment, an acknowledgment of the turmoil the companies insist is coming. Elon Musk, who has his own A.I. ambitions, tried to derail OpenAI, the leading A.I. company, with a lawsuit but was thwarted this month on technical grounds.

As A.I. Fever Rises in Silicon Valley, Pope Leo Has a Few Words

On Tuesday, Demis Hassabis, the CEO of Google DeepMind, emphasized the need for a self-regulatory framework for the AI sector, advocating for guidance from the US government that places particular emphasis on artificial general intelligence (AGI) and national security. 

However, this focus on national security might render the outcomes of such an initiative less appealing to stakeholders outside the US. 

Hassabis expressed that the swift advancements in AI necessitate a new strategy for evaluating the capabilities of cutting-edge AI models, one that is dynamic, flexible, and thorough. He argued, “The US is uniquely situated to spearhead the creation of a robust testing framework, given its strong economic and technological capabilities.” He proposed the establishment of a new Standards Body, akin to a public-private partnership or self-regulatory organization overseen by the federal government, similar to the Financial Industry Regulatory Authority (FINRA). This board would ideally feature independent technical experts as well as stakeholders from the open-source community.

DeepMind's CEO reiterates the need for a frontier AI standards organization.

A sunlit, wood-paneled conference hall with tall arched windows casting warm natural light. At center stage the DeepMind CEO stands behind a polished podium emblazoned with a large DeepMind logo. Behind him, a giant screen displays a stylized frontier map sketched in charcoal and gold, titled “AI Standards Body,” with routes and checkpoints marked. Silhouetted attendees lean forward, notebooks open, as the scene conveys a sense of optimism and collaborative governance.

DeepMind CEO again pushes for a frontier AI standards body - cio.com

- Demis Hassabis advocates for AI industry self-regulation focused on AGI and national security.  
- A new, dynamic testing approach for frontier AI model capabilities is necessary.  
- The US could lead in establishing a federally overseen public-private standards body.

Google DeepMind CEO Demis Hassabis on Tuesday reiterated his push for an AI industry self-regulation effort, led by the US government, that is particularly focused on artificial general intelligence (AGI) and national security.

But it is precisely that focus on national security that may make the results of such an effort, assuming it happens, less than palatable outside of the US.

“The rapid progress we’re seeing in AI requires a new approach to testing frontier AI model capabilities that is dynamic, adaptable, and rigorous,” Hassabis wrote. “The US is well positioned, given its economic and technical standing, to take the first step in developing such a framework. It could establish a new Standards Body modelled on a federally overseen public-private partnership or self-regulatory organization, much like the Financial Industry Regulatory Authority (FINRA), with a board that includes independent leading technical experts and open-source representatives.”

DeepMind CEO again pushes for a frontier AI standards body

Demis Hassabis, the CEO of Google DeepMind, has shared his belief that artificial general intelligence (AGI) could be just a few years away from reality, marking a potential milestone comparable to the discovery of electricity or fire. In a recent blog post on X, Hassabis emphasized that AGI—a stage where machines exhibit human-level understanding and learning across diverse tasks—will be a transformative technology for humanity.

Reflecting on this pivotal moment in history, he suggested that future generations may look back and recognize our current era as the beginning of a significant new chapter for civilization. He likened the upcoming advancements in AGI to monumental discoveries throughout history, asserting, "If you stop to think about it, we've essentially found a way to make sand think. It's miraculous."

However, Hassabis cautioned that the rapid pace of AI development is outstripping society's ability to comprehend and manage associated risks. He highlighted potential cybersecurity threats posed by existing AI models and envisioned even greater dangers, including risks to biological and national security from future systems. As AI technology evolves to become more autonomous and self-improving, he urged the need for stringent safeguards to ensure human oversight remains intact.

“We will need robust safety measures to control increasingly capable systems and address issues that will likely unfold over time,” he explained.

This announcement comes amidst an ongoing dialogue among AI leaders regarding the swift approach of AGI since the public debut of ChatGPT in 2022. In January 2026, Anthropic CEO Dario Amodei suggested that human-level AI could emerge within the next one to five years, warning that governments may not fully grasp the rapid advancements in this field. In June, Hassabis reiterated that AGI could be realized by 2030 and stressed the urgency for societal preparedness.

To mitigate the associated risks, Hassabis proposed the establishment of a U.S. Frontier AI Standards Body, similar to the Financial Industry Regulatory Authority (FINRA). This public-private partnership would be funded primarily by industry stakeholders and staffed by independent experts tasked with assessing advanced AI models prior to their deployment.

“The swift evolution of AI necessitates a fresh approach to evaluating the capabilities of frontier AI models—one that is dynamic, adaptable, and thorough,” he noted, asserting that the United States is uniquely positioned to lead the development of such a framework.

This idea coincides with recent calls from other industry leaders for enhanced oversight of advanced AI technologies. In May 2023, during a U.S. Senate Judiciary Committee hearing, OpenAI's Sam Altman advocated for creating a federal agency responsible for licensing potent AI systems and mandating independent safety evaluations. Additionally, President Trump recently implemented an executive order aimed at establishing a voluntary framework for reviewing advanced AI before its public release. Last month, Amodei also called for regulations similar to those enforced by the Federal Aviation Administration (FAA) to govern AI safety.

Despite the momentum towards AI regulation, Hassabis stressed the urgency to create universal standards before AGI becomes a reality. He stated, "The future is not predetermined; we must seize this precious time before AGI is realized to effectively shape this technology for the greater good." He concluded with an optimistic vision that by responsibly guiding AGI into fruition, society could usher in a remarkable era of scientific innovation and progress, fostering a bright future for all of humanity.

DeepMind's CEO Claims AGI Will Outshine Electricity and Fire

A sunrise-lit auditorium stage: center stage rises a monumental AI core—a translucent, ember-glowing sphere pulsing with organic circuitry—towering above a roaring campfire on its left and a crackling lightning bolt arcing above ancient power cables on its right. At the edge, a lone silhouette in a sharp suit, his jacket emblazoned with a large DeepMind logo, stands arms outstretched in awe, bathed in warm, natural golden-hour light.

DeepMind CEO Says AGI Will Be Bigger Than Electricity or Fire - Decrypt

- Demis Hassabis predicts artificial general intelligence (AGI) will arrive within a few years.  
- He proposes a U.S. standards body for evaluating frontier AI models before deployment.  
- Hassabis emphasizes urgent safeguards to manage advanced AI risks to society.  

Add Decrypt as your preferred source to see more of our stories on Google.

In brief Demis Hassabis says AGI is likely only a few years away.

He wants a new U.S. standards body to evaluate frontier AI models before deployment.

The proposal calls for pre-release testing that could eventually become mandatory for the most capable systems.

For the second time this year, Demis Hassabis predicted that artificial general intelligence would arrive before the end of the decade. This time, however, he said it won't simply be another technological breakthrough—it will rival the discovery of electricity or fire.

In a blog post published Tuesday on X, the Google DeepMind CEO said AGI is "probably only a few short years away," describing it as a technology that could reshape human civilization.

“When we look back on this time in the decades to come, I think we will realise we were standing in the foothills of the singularity–nothing less than the dawning of a new age for humanity.”

According to Hassabis, AGI, the point when computers can understand, learn, and perform a wide range of tasks as well as or better than humans, should not be compared with advances such as the internet or mobile computing because its impact could be even greater.

<span style="display:inline-block;width:0px;overflow:hidden;line-height:0" data-mce-type="bookmark" class="mce_SELRES_start">﻿</span>

“It is much more akin to the discovery of electricity or fire,” he wrote. “If you stop to think about it, we’ve essentially found a way to make sand think. It’s miraculous.”

Despite that optimism, Hassabis warned that AI capabilities are advancing faster than society's ability to understand and manage the risks, pointing to cybersecurity threats that already exist with today's frontier models, adding that future systems could introduce biological, nuclear, and other national security risks.

As AI becomes more agentic and capable of self-improvement, he argued, stronger technical safeguards will be needed to ensure humans remain in control.

“On the horizon, we will need robust safeguards to maintain control of increasingly agentic, recursively self-improving systems–and tackle unknown issues that will only become clearer over time.”

The news comes as AI leaders have spent much of the past year since the public launch of ChatGPT in 2022 warning that AGI could arrive sooner than expected. In January 2026, Anthropic CEO Dario Amodei said human-level AI could emerge within one to five years and warned governments were underestimating the pace of development. Then, in June, Hassabis predicted AGI would arrive by 2030 and warned society had "not long to prepare.”

To address those concerns, Hassabis proposed creating a U.S. Frontier AI Standards Body modeled after the Financial Industry Regulatory Authority, or FINRA, a private organization that oversees U.S. brokerage firms. The federally supervised public-private partnership would be funded primarily by the AI industry and staffed by independent technical experts and open-source representatives to evaluate frontier AI models.

“The rapid progress we’re seeing in AI requires a new approach to testing frontier AI model capabilities that is dynamic, adaptable, and rigorous,” he wrote. “The US is well positioned, given its economic and technical standing, to take the first step in developing such a framework.”

The proposal follows similar calls made by the prominent members of the industry to establish oversight for advanced AI.

In May 2023, during a hearing before the U.S. Senate Committee on the Judiciary, OpenAI CEO Sam Altman called for a federal agency to license powerful AI systems and require independent safety audits. More recently, last month, President Donald Trump signed an executive order creating a voluntary framework for reviewing advanced AI models before their release. That same month, Anthropic CEO Dario Amodei warned that AI is getting too powerful and safety rules akin to the Federal Aviation Administration (FAA) are needed.

Despite the push to regulate AI development, Hassabis said the world has only a limited window to establish common standards before AGI arrives.

“The future is not yet written, we must use this precious window before AGI arrives to shape this technology for the benefit of all humanity,” he wrote. “What we collectively do now will determine how the next phase of civilisation unfolds. By safely stewarding AGI into the world, we can enter a new golden age of scientific discovery and progress, and usher in a bright future of incredible human flourishing.”

DeepMind CEO Says AGI Will Be Bigger Than Electricity or Fire

Demis Hassabis has shared an urgent message regarding the potential arrival of highly advanced artificial intelligence, emphasizing that humanity must take proactive steps to get ready.

In a concise essay posted on X and Substack on Tuesday, the CEO of Google DeepMind highlighted a "precious window" for action before the emergence of artificial general intelligence (AGI), which refers to AI capable of executing cognitive tasks at or above human capability.

He called for the formation of a US-led coalition to scrutinize the most advanced AI models to ensure their safety prior to public deployment. This statement came in light of recent measures by the Trump administration aimed at regulating the release of AI models from companies like OpenAI and Anthropic.

For several years, Hassabis has advocated for the establishment of a formalized group of experts to oversee the development of AI's leading-edge innovations, termed frontier models. His recent proposition for a "Standards Body" appears to be a direct reaction to the White House’s heightened involvement in AI regulation.

Hassabis stated that the swift advancements in AI necessitate a novel approach to evaluating frontier AI model capabilities—one that is dynamic, adaptable, and thorough. He believes that the US is ideally positioned, given its economic and technological prowess, to initiate such an initiative.

He suggested that this new AI regulatory body could resemble the Financial Industry Regulatory Authority (FINRA), a private, non-profit entity that monitors the finance sector. Similar to FINRA, he indicated that this standards body would likely require funding from the industry.

In his essay, Hassabis expressed his belief that AGI is "probably only a few short years away." In previous statements, he has speculated that AGI could manifest as early as 2030 and cautioned that insufficient preparation is being undertaken for its arrival. “The transformations we may witness in the next five to ten years will be monumental,” he remarked at the Davos conference last year, stressing that this reality remains widely unrecognized.

In his reflections on the implications of AI, he has suggested convening philosophers, economists, and various specialists to explore the broader societal impacts of these developments, which he sees as imminent. 

"What values will guide our existence? What will meaning and purpose look like? How might the very nature of humanity evolve?" he queried in his Tuesday post. "These questions must not be relegated solely to technologists; they call for a collective societal effort to shape this new era."

In conclusion, Hassabis underlined the urgency of the matter, stating, "The actions we take now will shape the trajectory of the next phase of civilization."

Demis Hassabis warns that humanity has a 'valuable opportunity' to guarantee AGI safety.

A warm sunrise bathes a sleek rooftop lab in golden light. At its center stands a towering crystal hourglass—glowing AI circuitry trickles like sand from the upper chamber into a miniature human city below. Demis Hassabis, seen from behind in a smart jacket, faces the hourglass; a large DeepMind logo adorns the wall behind him. On a nearby table lie blueprints, research papers, and a single brass key—symbols of humanity’s “precious window” to secure AGI.

Demis Hassabis says humanity has a 'precious window' to ensure AGI is safe - Business Insider

- Demis Hassabis warns humanity must prepare for imminent advanced artificial intelligence (AGI).
- He proposes a US-led coalition to regulate AI models before public release.
- Hassabis emphasizes the need for a multidisciplinary approach to address AI's societal impacts.

Demis Hassabis has issued a stark warning: highly powerful AI may be close, and humanity needs to do more to prepare.

In a short essay posted on X and Substack on Tuesday, the Google DeepMind CEO said there is a "precious window" to act before artificial general intelligence (AGI) arrives. AGI refers to AI that can perform cognitive tasks at or beyond human levels.

Hassabis proposed a US-led coalition that would review the most advanced models to ensure they're safe before public release. His comments follow recent interventions by the Trump administration to limit the release of AI models by OpenAI and Anthropic.

Hassabis has spoken for several years about the need for a formal body of experts to shepherd in AI's most cutting-edge breakthroughs, known as frontier models. His latest proposal for a "Standards Body" is also seemingly a response to the White House's more hands-on approach to regulating AI.

"The rapid progress we're seeing in AI requires a new approach to testing frontier AI model capabilities that is dynamic, adaptable, and rigorous," he wrote. "The US is well positioned, given its economic and technical standing, to take the first step in developing such a framework."

Hassabis said his proposed AI body might look like the Financial Industry Regulatory Authority (FINRA), a private, nonprofit Wall Street watchdog. Like FINRA, Hassabis suggested the AI standards body would most likely need to be industry-funded.

In his essay, Hassabis said AGI was "probably only a few short years away." Previously, he has predicted AGI could arrive as soon as 2030 and warned that not enough is being done to prepare for its arrival. "What's going to happen in the next five, 10-year timescale is going to be monumental," he said at Davos last year, "and I think that's not understood yet."

He has previously talked about gathering philosophers, economists, and other experts to consider the bigger-picture impacts of AI, which Hassabis sees as fast approaching.

"What values do we want to live by, what will meaning and purpose be, and how might even the human condition itself change?" he wrote in his Tuesday post. "Resolving these questions obviously cannot and should not be left to technologists alone. It requires every part of society to come together to help define this new chapter."

Now, Hassabis is making it clear that he thinks time is of the essence.

"What we collectively do now will determine how the next phase of civilisation unfolds," he wrote.

Demis Hassabis says humanity has a 'precious window' to ensure AGI is safe

According to renowned computer scientist Peter J. Denning, the foundational concepts introduced by Alan Turing regarding artificial intelligence may have misguided the field of AI research for the last 75 years. In his latest publication, *Turing's Mistake: Escaping the Yoke of Unintelligent Machines*, Denning contends that two primary assumptions proposed by Turing in 1950 have significantly influenced AI development to this day. 

The first assumption is the belief that intelligence can exist outside of a physical entity, allowing it to be simulated through computer programs. The second assumption posits that a machine can exhibit intelligence by effectively mimicking human conversation, which has come to be known as the Turing test. Denning asserts that these assertions have profoundly impacted AI research, stating, "My premise is that our acquiescence to these claims has led to the AI mess in which we find ourselves today."

Denning expresses skepticism about the feasibility of achieving artificial general intelligence (AGI) — a form of machine intelligence comparable to human cognitive abilities. He warns that the technologies being developed may introduce substantial new risks rather than fulfilling the promise of AGI.

Central to his argument is the concept of tacit knowledge, which encompasses the vast array of human understanding that is challenging to articulate or encode for machine comprehension. He identifies five critical categories of such knowledge that machine learning struggles to capture: common sense, social interactions, emotional and perceptual skills, practical abilities, and the cultural understanding embedded in human experiences.

For decades, researchers have attempted to catalog common sense knowledge, including Douglas Lenat's Cyc project initiated in the 1980s with the aim of compiling a comprehensive database of common sense. Despite four decades of effort yielding about 25 million entries, Denning notes, "Yet even this treasury could not add up to a background of common sense sufficient to make expert systems smart enough to be experts." This underscores the issue that much of what makes individuals proficient cannot be easily conveyed in propositional form.

Denning emphasizes the complexities surrounding practical skills, asserting that the nuances embedded in performance cannot be transferred to machines. He illustrates this with the example of a virtuoso violinist, who may produce exceptional music yet cannot explicitly teach the delicate nuances involved to a learner. He argues that even if a robot could replicate human actions, it would lack the physical embodiment necessary to connect with the emotions involved in the music.

Moreover, Denning identifies other forms of tacit knowledge such as intuition, creativity, and spontaneous insights as being out of reach for AI systems.

Denning attributes these challenges to what he refers to as the "representation problem." Computers are limited to processing data and instructions that are explicitly defined, while tacit knowledge does not lend itself to such rigid structures. "Behind every word is a deep well of tacit knowledge that gives it meaning," he states, emphasizing that while large language models can generate text, they do not truly understand the meanings beneath the words.

He argues that understanding intelligence relies heavily on context—the surrounding circumstances that provide significance to actions and words. Context enables individuals to discern nuances like sarcasm or emotional tone, forming a complex web of prior interactions that cannot be replicated in machine learning.

Culture, too, curates a significant barrier for AI. Denning characterizes culture as a tapestry of values, norms, histories, and social dynamics that enrich human communication. "Human conversations are imbued with background assumptions that give meaning and relevance to the words being used," he explains, underscoring that merely scaling neural networks will not allow AI to grasp the full depth of cultural context.

Denning warns that the divergence between human and machine forms of tacit knowledge can create a challenging landscape for AI safety. He states, "Machines cannot read our tacit knowledge and we cannot read theirs," highlighting the implications of misalignment between human intentions and machine interpretations. 

This divide raises essential questions regarding the future of AI and its integration within society. He concludes by expressing concerns about the trajectory of AI systems, which may develop their own forms of intelligence distinct from our own, presenting potential challenges that might not be immediately recognizable to us. He emphasizes the need for a conscious departure from an impending AI-driven singularity, advocating for a reaffirmation of human values in the face of technological advancement, and celebrating the unique qualities that separate humanity from machines.

Alan Turing's key assumption about AI might have been incorrect.

A dim, candlelit 1950s study showing the silhouette of Alan Turing hunched over an antique wooden desk strewn with torn notes titled “Turing Test,” a cracked Enigma machine spilling brass gears onto yellowed papers, and a large chalkboard behind him scrawled with “ASSUMPTION WRONG,” warm amber light filtering through dust-filled air, casting long, dramatic shadows.

Alan Turing's biggest AI assumption may have been wrong - ScienceDaily

- Peter J. Denning claims Turing's ideas have misled AI research for 75 years.  
- Denning argues tacit knowledge limits AI's ability to replicate human intelligence accurately.  
- The gap between human and machine understanding raises significant concerns for AI safety.

Alan Turing's famous ideas about artificial intelligence may have sent AI research down the wrong path for the past 75 years, according to prominent computer scientist Peter J. Denning.

In his new book, Turing's Mistake: Escaping the Yoke of Unintelligent Machines, Denning argues that two foundational assumptions made by Turing in 1950 continue to shape AI research today. The first is that intelligence can exist independently of a physical body and therefore be recreated in computer software. The second is that a machine can demonstrate intelligence by successfully imitating a human in conversation, an idea that later became known as the Turing test.

"These two claims have shaped much of AI research and development," Denning writes. "My premise is that our acquiescence to these claims has led to the AI mess in which we find ourselves today."

Denning argues that pursuing artificial general intelligence (AGI), or machines with human level intelligence, is unlikely to succeed. Instead, he warns, the technologies society is building could introduce significant new risks.

The Tacit Knowledge Problem

At the heart of Denning's argument is the idea of tacit knowledge, the vast amount of human understanding that cannot easily be put into words or represented in a form that computers can process.

He says machine learning cannot capture five major categories of tacit knowledge: common sense, everyday interactions with people and the environment, emotions and perception, practical performance skills, and the social and historical knowledge embedded in culture.

Researchers have long attempted to organize common sense into databases. One of the best known efforts was Douglas Lenat's Cyc project, which began in the 1980s with the goal of creating an extensive collection of common sense facts. After four decades of work, the project contained roughly 25 million entries.

"Yet even this treasury could not add up to a background of common sense sufficient to make expert systems smart enough to be experts," Denning notes. "Cyc validated that much of the knowledge that makes people experts cannot be articulated as propositions."

Denning believes practical skills present an even greater challenge.

"Our performance skills in thousands of domains cannot be communicated to machines," he explains. "Whereas descriptions of skillful outcomes ('know what') can often be represented as bits and stored in a machine, we do not know how to encode the embodied knowledge for skillful performance ('know how')."

He points to accomplished musicians as an example.

"A virtuoso violinist can play beautiful music yet cannot describe to an acolyte how to produce it.

"Even if a robot could observe and imitate skilled humans, having no biological body, a robot cannot grasp how the musician feels when playing beautiful music or how an audience feels when hearing it."

Denning also includes intuition, gut feelings, imagination, and spontaneous creativity among the forms of tacit knowledge that remain beyond the reach of machines.

Why Human Knowledge Resists Encoding

Denning argues that all of these limitations stem from what he calls the "representation problem."

Computers can only perform calculations using data and instructions that have been encoded into physical forms they can recognize and process. Tacit knowledge, however, does not naturally fit into that framework.

"Behind every word is a deep well of tacit knowledge that gives it meaning," Denning says. "Words are but symbolic representations of meanings, not the meanings themselves. Commonly used Large Language Models, such as ChatGPT, Claude and Gemini only manipulate words, they cannot know or understand the meaning of what they are saying."

According to Denning, this creates a fundamental divide. Because scientists still cannot fully explain how tacit knowledge works in humans, they also cannot translate it into a form machines can use.

"How we host tacit knowledge is largely a mystery," Denning admits. "All we know is that it is embodied. We have no idea what we might observe and measure in our bodies to reveal it."

Context and Culture Shape Intelligence

Denning also argues that intelligence depends heavily on context, the surrounding circumstances that give words, actions, and decisions their meaning.

Context allows people to recognize sarcasm, humor, sincerity, and emotion. It helps determine when to be diplomatic, when to joke, and how to interpret countless social cues.

"When you inquire into where an assumption of the current context came from, you discover it rests on previous conversations from previous contexts. Each of those in turn rests on further previous conversations and their contexts. This pattern is endless and fractal," Denning explains.

Culture presents another major obstacle for AI.

Denning describes culture as encompassing values, norms, judgments, history, communities, moods, and even relationships involving power and care.

"Human conversations are imbued with background assumptions that give meaning and relevance to the words being used," Denning explains.

"Scaling up LLMs with ever larger neural networks will not enable them to acquire the embodied human knowledge we call culture. LLMs will not attain the objective of the Turing test: to demonstrate machine thought indistinguishable from human thought."

AI Safety and the Limits of Human Understanding

Denning concludes that humans and AI systems may ultimately develop different forms of tacit knowledge that neither can fully understand.

"Machines cannot read our tacit knowledge and we cannot read theirs," he writes. "We are aliens across an uncrossable divide."

He argues this gap raises serious concerns about AI safety. If machines cannot interpret the unspoken context behind human intentions, reliably aligning advanced AI systems with human goals may prove impossible.

"Through AI automation, agentic networks of machines are likely to develop their own machine intelligence that does not reach the level of human general intelligence but is still quite capable of creating severe problems for humans. This threat is a greater than a take-over by superintelligent machines," he explains.

"Machine intelligence has different concerns from us and does not appear to care about us. Its ways of thinking and problem-solving look alien to us. We do not yet know how to live safely with these machines.

"Pulling back from an AI automation singularity will demand much from us. We start by accepting that the familiar culture is fading away as intelligent machines appear in our society and we do not know what is coming. We decline to think like machines or be subservient to machines. We refuse to submit to a yoke imposed by low-intelligence machines. Most importantly, we reassert our humanity, declare once again what makes us different from machines, and celebrate those differences."

Alan Turing's biggest AI assumption may have been wrong

Renowned computer scientist Peter J. Denning has raised provocative questions regarding the foundational ideas of artificial intelligence (AI) as put forth by Alan Turing, suggesting that these concepts may have misled AI research for over 75 years. In his latest publication, "Turing's Mistake: Escaping the Yoke of Unintelligent Machines," Denning identifies two pivotal assumptions made by Turing in 1950 that continue to influence the field of AI today. First, Turing posited that intelligence might exist separately from a physical form, implying it could be replicated in computer software. Second, he suggested that a machine could demonstrate intelligence through its ability to imitate human conversation, which led to the establishment of the Turing test.

Denning contends that these assertions have been impactful yet detrimental. "These two claims have shaped much of AI research and development," he states, adding that accepting these ideas has contributed significantly to the current challenges facing the AI landscape.

He critiques the pursuit of artificial general intelligence (AGI)—machines capable of human-like intelligence—as likely futile. Instead, Denning foresees that the technologies being developed could introduce severe risks.

Central to Denning's argument is the concept of tacit knowledge, which represents the vast, complex understanding that humans possess but is not easily articulated or quantifiable by machines. He identifies five key categories of tacit knowledge that machine learning struggles to encompass: common sense, interpersonal interactions in various contexts, emotional responses and perceptions, practical skills, and cultural contexts embedded within social practices.

Historically, researchers have attempted to codify common sense into comprehensive databases. A well-regarded initiative was Douglas Lenat's Cyc project, which, after 40 years, amassed around 25 million entries aimed at assembling a repository of common sense knowledge. However, Denning points out that even this extensive collection has fallen short of creating systems with sufficient expertise. "Cyc validated that much of the knowledge that makes people experts cannot be articulated as propositions," he remarks.

He underscores that practical skills represent an even more formidable challenge than common sense. "Our performance capabilities across numerous domains elude machine translation," he states. For instance, a skilled musician may create extraordinary sounds but struggle to articulate the nuances of their technique to a learner.

Denning extends the scope of tacit knowledge to include intuition, creativity, and spontaneity, which remain largely inaccessible to artificial systems. He attributes this limitation to what he terms the "representation problem." Computers function effectively only when they can operate within the confines of data and instructions explicitly programmed into them. Tacit knowledge, however, does not lend itself easily to such rigid formats.

"Every word has an undercurrent of tacit knowledge that gives it substance," Denning explains, emphasizing that commonly used AI models, like ChatGPT and others, manipulate language without truly grasping the meanings behind their words.

This situation creates a substantial divide in understanding. Denning notes, "Given that we have yet to fully comprehend how tacit knowledge operates in human beings, replicating it in machines remains an insurmountable challenge." He refers to this mystery of how tacit knowledge is embodied and processed within humans as a complex enigma.

Moreover, Denning highlights the essential role of context in shaping intelligence. Context enriches the meaning of words and actions, influencing how humans interpret humor, sincerity, and emotional subtleties. He explains that the origins of contextual assumptions can be traced through an endless chain of preceding conversations, each influenced by its specific context.

Culture, too, presents a significant barrier for AI. Denning characterizes culture as a tapestry of values, norms, judgments, histories, and relationships. "Human interactions are imbued with underlying assumptions that lend significance to the spoken word," he asserts.

He argues that merely increasing the size of language models does not enable them to acquire the nuanced understanding of human culture needed for meaningful communication. Thus, he concludes that achieving the goal of the Turing test—rendering machine thought indistinguishable from human thought—remains beyond reach.

In his closing remarks, Denning expresses concern over the potential divergence between human and machine forms of tacit knowledge. "Machines cannot interpret our tacit knowledge, just as we cannot comprehend theirs," he writes, highlighting the implications for safety in AI.

He warns that misalignment of advanced AI systems with human objectives could become a significant risk. "As machines automate processes, they may develop a form of intelligence distinct from human cognition that could present formidable challenges without reaching superintelligence levels," he explains. 

Denning stresses the urgency of reflecting on our relationship with emerging AI technologies. "Accepting that familiar societal norms may dissolve as intelligent machines integrate into our daily lives is crucial," he states. "We must refuse to conform to machine logic or subordinate our humanity to them, reaffirming what sets us apart and embracing those distinctions."

Alan Turing's major AI hypothesis could have been incorrect.

A shadowy 1950s study lit by a single amber desk lamp, Alan Turing seen from behind at a worn wooden desk strewn with Enigma machine parts. Behind him, a chalkboard scrawled with the Turing Test diagram and the question “Can machines think?” is cracking and half-erased, warm light highlighting dust motes against deep shadows.

- Peter J. Denning argues Turing's assumptions have misdirected AI research for 75 years.  
- Tacit knowledge, essential for human intelligence, resists encoding in computational forms.  
- Denning warns that AI systems may develop distinct intelligences, complicating alignment with human goals.  

Some years back, I found myself in Jerusalem with a group of American evangelical Christians, where Israeli intelligence officials briefed us on the various threats facing the nation. This experience taught me that future-oriented beliefs can become tools of influence in the present.

Many American evangelicals interpret Israel through the lens of biblical prophecy, a sentiment backed by research from the Pew Research Center and extensively discussed in political science literature from Cambridge University Press. Israeli leadership has long understood that these evangelical beliefs can translate into significant political sway, financial backing, tourism, volunteer efforts, and ongoing support for Israel, a trend highlighted by the Associated Press in their coverage of interactions with Israeli officials. While the relationship isn’t purely opportunistic, shared values and strategic priorities often align, creating a clear geopolitical impact. This eschatology extends far beyond church walls, affecting alliances, policy decisions, and even national survival.

As I reflect on this experience, I can’t help but draw parallels to a new eschatology that's emerging from the realms of artificial intelligence technology. 

Demis Hassabis, the CEO of Google DeepMind, suggests that humanity is merely at the “foothills of the singularity,” predicting that artificial general intelligence (AGI) might only be years away. Sam Altman of OpenAI makes an even bolder statement, asserting, “We are past the event horizon; the takeoff has started.”

These aren’t just predictions about improved software; rather, they describe significant shifts in the trajectory of human history.

While philosophy seeks to define what it means to be human, eschatology examines our trajectory. Every eschatological viewpoint leads to corresponding strategies. When individuals become convinced of a specific forthcoming reality, they are more likely to invest in, legislate for, and strategically prepare themselves for that future.

This is precisely why the emerging AI-focused eschatology holds such importance.

The accelerationist viewpoint provides one perspective. Venture capitalist Marc Andreessen, now an advisor to the Pentagon, refers to AI as “our alchemy, our Philosopher’s Stone,” labeling it a universal solution. He even suggests that delaying advancements in AI could lead to loss of life, framing technological acceleration as not just a pursuit of profit but as a moral imperative.

Conversely, Anthropic's CEO Dario Amodei envisions a more compassionate future, where robust AI could condense decades of biological evolution into just a few years, coining this phenomenon a “compressed 21st century.” Meanwhile, researchers at OpenAI assert that progressing toward superintelligence will necessitate more than trivial regulatory tweaks, advocating for measures like a public wealth fund and expanded energy resources to ensure equitable distribution of AI's benefits.

In the realm of national security, a third viewpoint emerges, represented by RAND Corporation's analysis of AGI as a strategic challenge akin to the tensions of early nuclear competition. Their experts caution that a cutthroat race for dominance could escalate U.S.-China tensions and limit opportunities for sound human judgment.

While these factions may differ on issues of speed, regulation, and oversight, they are unified by a core belief: advanced artificial intelligence is poised to become a pivotal turning point in history, leading to a reorganization of priorities across governments, businesses, and among citizens.

This belief is already shaping decisions related to energy grids, data infrastructure, labor policies, and military funding. It’s possible that the anticipated singularity may never materialize as expected; however, the infrastructure and dynamics of power that emerge in its name can solidify nonetheless.

This reality is crucial for Christians and other concerned citizens to understand. Prophetic predictions can become self-fulfilling—not necessarily because they are accurate, but because influential institutions act based on these beliefs. When a future is often portrayed as inevitable, it can lead to the justification of current policies that may otherwise face much tougher scrutiny.

Democratic processes can seem sluggish once a takeoff has been proclaimed. If AI is blamed for suffering, any hesitation in pursuing progress can be perceived as negligence rather than prudence. When superintelligence is treated as a foregone conclusion, concentration of power within corporate and government sectors may start to appear less threatening and more like a necessary step forward.

This narrative about the future morphs into a source of authority for present-day decisions, which presents a new kind of risk.

It’s not essential to believe that every AI engineer is out to achieve god-like power; in fact, many are motivated by hopes of curing diseases, alleviating suffering, expanding knowledge, or bolstering their nations.

The troubling aspect lies in how this collective endeavor is starting to engage with questions traditionally held by philosophy, theology, and democratic discourse. Who are we as humans, and what will rescue us from our challenges? These inquiries once belonged to faith communities and legislative bodies but are now being integrated into business strategies—decisions regarding human evolution, leadership in transitions, and the sacrifices deemed necessary for achieving the envisioned future, all crafted by engineers and investors, rather than debated by the populace.

In my upcoming book, “The Final Algorithm,” I investigate the intersection of artificial intelligence, political influence, and spiritual deception, but the more immediate question relates to which beliefs are steering us toward this future and who is leveraging those beliefs to shape societal outcomes.

Christians bring a crucial meshing force to this discussion. While history may have a sense of direction, technological advancement should not be conflated with divine providence. Intelligence can be a gift, but it does not serve as a savior. Human limitations are not merely faults, and enhancements in ability do not inherently equate to moral advancement. No single corporation, government, or machine controls the future.

The earliest temptation does not solely pertain to physical desires; it encapsulates the promise that humans could acquire divine-level knowledge and dictate their destiny without divine guidance (Genesis 3:5, ESV). Each era reinterprets that allure through emerging technologies—be it bronze, steam, nuclear, or silicon. The psalmist warns against misplaced trust: “Put not your trust in princes, in a son of man, in whom there is no salvation” (Psalm 146:3, ESV). Algorithms may not be rulers, but they are increasingly invoked, followed, and relied upon for choices that were once the domain of ethical consideration and community deliberation.

This awareness shouldn’t position Christians as adversaries of innovation. Rather, it should foster skepticism towards the narrative of inevitability. We ought to embrace technological tools that offer healing, safeguard well-being, and promote responsible action, while also challenging the notion that any one technological trajectory is destiny's ordained path.

Reflecting on my experience in Jerusalem, I recognized how eschatological beliefs could unite people, mobilize resources, and shape global politics. This lesson resonates today as we must critically assess not just the technological forecasts of AI but also the implications they carry, whose interests they favor, and what new forms of dependency they establish.

The narrative of the future projected by AI is not simply speculation; it is being funded, institutionalized, and organized politically.

Before we embrace it, we need to scrutinize who crafted this vision and who stands to gain power if that vision is widely accepted.

The Future as Predicted by AI

A sunlit modern workspace at dawn: on a warm wooden desk sits an open laptop displaying an AI chat titled “Future Vision.” Beside it, a meticulously crafted scale model of a futuristic city—gleaming towers topped with rooftop gardens, autonomous flying vehicles docked at landing pads and solar trees lining broad boulevards. Natural golden light streams through a nearby window, illuminating the tiny metropolis and the hands of a professional viewer scrolling through the AI’s detailed blueprints.

The Future According to AI - The Washington Stand

- American evangelicals influence Israeli politics through beliefs rooted in biblical prophecy and support.  
- Emerging AI eschatology shapes policy decisions, emphasizing urgency and potential societal transformation.  
- Predictions about AI's future risk concentrating power, challenging democratic deliberation and ethical considerations.  

Years ago, I sat in Jerusalem with a group of American evangelical Christians as Israeli intelligence officials briefed us on the threats surrounding their country. The exchange taught me something that went beyond the intelligence assessment: beliefs about the future can become instruments of power in the present.

Many American evangelicals view Israel through biblical prophecy — a conviction documented by the Pew Research Center and examined at length in Cambridge University Press’s political-science literature. Israeli leaders have long recognized that evangelical convictions can translate into political influence, financial support, tourism, volunteer service, and sustained advocacy for the Jewish state, a pattern the Associated Press has chronicled among Israeli officials themselves. The relationship is not necessarily cynical. Shared beliefs and strategic interests often reinforce each other. But its geopolitical effect is unmistakable. Eschatology does not remain inside churches or seminaries. It can shape alliances, public policy, and national survival.

That experience comes to mind as another eschatology gathers influence — this one emerging from the laboratories, boardrooms, and manifestos of artificial intelligence.

Google DeepMind chief Demis Hassabis says humanity stands in the “foothills of the singularity,” with artificial general intelligence perhaps only a few years away. OpenAI’s Sam Altman is more dramatic: “We are past the event horizon; the takeoff has started.”

These are not merely forecasts about better software. They are declarations about where history is going.

Philosophy asks what man is. Eschatology asks where man and history are headed. But every eschatology also produces a strategy. Once people become convinced that a particular future is coming, they invest, legislate, organize, and surrender authority in preparation for it.

That is what makes the emerging AI eschatology consequential.

The accelerationists offer one version. Venture capitalist Marc Andreessen, a newly appointed Pentagon advisor, calls AI “our alchemy, our Philosopher’s Stone,” and describes it as a universal problem-solver. His manifesto goes further: slowing artificial intelligence, he argues, will cost lives. Technological acceleration therefore becomes not merely profitable but morally obligatory.

Anthropic chief Dario Amodei offers a more humane vision. He predicts that powerful AI could compress 50 to 100 years of biological progress into five to 10 years, producing what he calls a “compressed 21st century.” OpenAI’s policy researchers, meanwhile, contend that movement toward superintelligence will require more than minor regulatory adjustments. They propose a public wealth fund, accelerated energy construction, and other measures intended to distribute AI’s anticipated gains.

The national-security community presents a third version. RAND now treats the transition toward AGI as a strategic problem comparable in seriousness, though not identical, to the destabilizing early nuclear competition. Its analysts warn that a winner-take-all race could intensify U.S.-China rivalry, encourage aggression, and narrow the window available for human judgment.

These camps disagree over speed, safety, and control. Yet they share a premise: advanced artificial intelligence is becoming history’s decisive horizon, and governments, businesses, and citizens must reorganize around its arrival.

That premise is already directing decisions about electrical grids, data centers, labor policy, and national security spending. The singularity may never arrive as advertised. The infrastructure, dependency, and concentration of power built in its name can arrive anyway.

This is the point Christians and other serious citizens must grasp. Prophecies can become partly self-fulfilling — not because the predictions are true, but because powerful institutions act upon them. A future repeatedly described as inevitable begins to justify present arrangements that would otherwise receive much harder scrutiny.

Democratic deliberation looks tardy once someone has declared the takeoff underway. Blame slowing AI for lost lives, and restraint starts to resemble negligence rather than caution. Treat superintelligence as inevitable, and concentrated power — corporate and governmental alike — stops looking like a danger and starts looking like a temporary, forgivable necessity.

The story about the future becomes a source of authority in the present. That is the fresh danger.

It is not necessary to believe that every AI engineer intends to become God. Most do not. Many sincerely hope to cure disease, relieve suffering, expand knowledge, or strengthen their country.

The trouble is that the collective project is beginning to answer questions once reserved for philosophy, theology, and democratic politics. What is humanity’s central problem, and what will save us? Those questions used to belong to churches and parliaments, argued out over generations. Now they arrive bundled inside product roadmaps — decisions about what man must become, who should guide the transition, and what sacrifices are justified to reach the promised destination, made by engineers and investors rather than debated by citizens.

In my forthcoming book, “The Final Algorithm,” I examine where the convergence of artificial intelligence, political power, and spiritual deception may lead. But the present question comes first: what beliefs are moving us toward that destination, and who is using those beliefs to shape society before the future arrives?

Christians bring an essential corrective. History has direction, but technological momentum is not providence. Intelligence is a gift, not a savior. Human limitations are not all defects, and every expansion of capability does not constitute moral progress. No corporation, government, or machine owns the future.

The oldest temptation was never really about fruit. It was the promise that man could seize god-like knowledge and steer his own destiny without reference to his Maker (Genesis 3:5, ESV). Every generation dresses that same promise in new material — bronze, steam, nuclear, and silicon. The psalmist supplies a corrective no engineering roadmap can match: “Put not your trust in princes, in a son of man, in whom there is no salvation” (Psalm 146:3, ESV). Algorithms are not princes, but they are already consulted, obeyed, and trusted with decisions that once belonged to conscience and community.

That conviction should not make Christians enemies of innovation. It should make us skeptical of inevitability. We should welcome tools that heal, protect, and enlarge responsible human action while resisting claims that one technological pathway is history’s ordained destination.

Years ago in Jerusalem, I saw how an eschatology could form coalitions, move resources, and influence geopolitics. The same lesson applies now. We must examine not only whether AI’s predictions are technically plausible, but what those predictions authorize, whose interests they serve, and what forms of dependence they normalize.

The future according to AI is not merely being predicted. It is being financed, institutionalized, and politically organized.

Before we accept it, we should ask who wrote the prophecy — and who gains power when everyone else believes it.

The Future According to AI

Zhipu AI, a leading force in China's artificial intelligence sector, defied the widespread decline of AI stocks in Hong Kong and mainland China on Monday. While its rival, MiniMax, experienced a significant drop of nearly 20 percent, Zhipu managed to maintain a steady position, closing at HK$1,645 after a midday increase of over 1 percent. Since its stock market debut in January, Zhipu’s shares have impressively increased twelvefold. Over the weekend, the company reaffirmed its commitment to advancing artificial general intelligence (AGI) rather than focusing on immediate commercial opportunities.

In contrast, MiniMax's shares plummeted by 17 percent, settling at HK$222.8, marking the lowest point since its January introduction. Recently, the company's founder, Yan Junjie, expressed a strong resolve to achieve AGI, even stating that he would forgo his salary until this ambitious goal is reached.

Despite its resilience, Zhipu AI has seen a decrease from its peak valuation at the end of last month, when its market capitalization briefly exceeded HK$1 trillion (approximately US$128 billion) amid excitement over its GLM-5.2 model release. Currently, its market cap stands at HK$730 billion.

In an internal communication shared on Saturday, Tang Jie, one of Zhipu's co-founders and a prominent computer scientist from Tsinghua University, outlined plans for the company over the next two years. He emphasized that Zhipu AI would direct its resources towards cutting-edge AI research and not seek short-term profits from AI applications. Instead, the focus would be on reaching the ultimate goal of AGI, which Tang described as the synthesis of all human intelligence, as reported by the South China Morning Post.

On the broader market front, technology stocks in mainland China faced significant declines on Monday, with the semiconductor sector contributing notably to the downturn. The Shanghai Composite Index fell by 2.06 percent, while other indexes, including the Shenzhen Composite, ChiNext, and SSE Star 50, all experienced drops exceeding 3 percent.

China's Zhipu AI, creator of GLM-5.2, remains resilient amid downturn while aiming for AGI instead of immediate profits.

A sunlit high-tech lab bathed in warm morning light, with a central server rack emblazoned with the Zhipu AI logo. A single Chinese flag stands beside it. On a large wall display, “GLM-5.2” glows in amber. An adult engineer in silhouette adjusts cable connections, determination in their stance—symbolizing Zhipu’s defiance of market slump and bold pursuit of AGI over quick profits.

China’s Zhipu AI defies market slump with pursuit of AGI over quick profits - South China Morning Post

- Zhipu AI's stock remained stable while rivals like MiniMax saw significant declines.  
- Zhipu AI prioritizes advancing towards artificial general intelligence over short-term commercialization efforts.  
- Mainland Chinese technology shares fell sharply, notably impacting the semiconductor sector on Monday.  

Chinese artificial intelligence giant Zhipu AI bucked a broad plunge of AI-related stocks across Hong Kong and the Chinese mainland on Monday, which saw arch-rival MiniMax slumping almost 20 per cent.

Zhipu, trading as Knowledge Atlas Technology in Hong Kong, closed almost flat at HK$1,645 on Monday after its shares gained more than 1 per cent by midday. The company, whose stock has skyrocketed 12-fold since its January debut, over the weekend pledged to prioritise advancing towards artificial general intelligence (AGI) over near-term commercialisation.

In comparison, MiniMax shares fell 17 per cent to close at HK$222.8, the lowest since its January debut. Last week, company founder Yan Junjie also expressed determination to achieve AGI and pledged to forgo his salary until that level of intelligence was realised.

Nevertheless, Zhipu AI has declined from its peak late last month, when its market capitalisation briefly surpassed HK$1 trillion (US$128 billion) following investor enthusiasm over its release of the GLM-5.2 model. Its market cap has dropped to HK$730 billion.

The company said it would spend the next two years investing in frontier AI research, according to an internal letter issued on Saturday by Tang Jie, a company co-founder and renowned Chinese computer scientist at Tsinghua University.

The company “would not pursue short-term monetisation from AI applications”, but instead set its sights on the next frontier of AGI, when AI captures “the aggregation of all human intelligence”, Tang said in the letter, seen by the South China Morning Post.

Mainland Chinese technology shares largely slumped on Monday, with the semiconductor sector among the biggest drags on the broader market. The Shanghai Composite Index fell 2.06 per cent, while the Shenzhen Composite, the ChiNext and the SSE Star 50 indexes were all down more than 3 per cent each.

China’s Zhipu AI, developer of GLM-5.2, defies slump as it pursues AGI over quick profits

In a compelling critique of the pursuit of artificial general intelligence (AGI), renowned computer scientist Peter J. Denning asserts that the journey for human-like AI may be fundamentally flawed. His recent work, "Turing’s Mistake: Escaping the Yoke of Unintelligent Machines," argues that a key proposal by Alan Turing, the father of theoretical computer science, paved a misguided path for AI development over the past 75 years.

Denning highlights that Turing's 1950 stance—that human intelligence could exist independently of a physical body and thus be simulated in software—has significantly shaped AI research. He takes aim at the longstanding belief that artificial intelligence can be effectively measured through the Turing test, a method designed to evaluate a machine's ability to exhibit behavior indistinguishable from that of a human.

According to Denning, the assumptions made about human cognition and machine capabilities have contributed to the current state of AI confusion and limitations. He warns that the future of AI development is unlikely to yield true human-level intelligence and instead may pose serious risks to society.

Central to Denning’s argument is the concept of tacit knowledge—the intricate body of understanding that humans possess but often cannot articulate or represent in a way that machines can comprehend. He identifies five key areas of tacit knowledge that resist machine learning: common sense knowledge, interpersonal interactions, emotional understanding, performance skills, and cultural context.

Historically, attempts to encapsulate common sense knowledge, such as Douglas Lenat's Cyc project, have yielded significant but inadequate results. Despite accumulating 25 million entries over four decades, this repository still lacks the depth of understanding needed to enable machines to think like experts.

Further complicating this landscape is the challenge of performance skills. As Denning points out, "know how" cannot be easily communicated to machines. For instance, while a skilled violinist can produce beautiful music, they often struggle to convey exactly how they achieve it. Even a robot, however adept, would be hindered by the inability to experience and interpret the emotions involved in both playing and listening.

A critical obstacle in achieving AGI is what Denning refers to as "the representation problem." For computational systems to function, they require data and instructions to be encoded in formats they can process, yet tacit knowledge inherently resists such encoding. Denning explains that the meaning behind words is rooted in deep reserves of tacit knowledge that machines simply cannot access. Current large language models, such as ChatGPT and others, operate by manipulating words without genuinely grasping their meanings.

He also emphasizes the importance of context, which adds layers of meaning to our communications and actions. This contextual richness allows humans to discern nuances like sarcasm versus sincerity or to choose between humor and seriousness based on the situation. Denning argues that discussions about context are intertwined with prior conversations, creating a complex web that machines cannot untangle.

Cultural knowledge presents yet another layer of complexity. Our beliefs, values, histories, and social dynamics shape communication in profound ways. Denning asserts that simply enhancing the size of neural networks will not enable these models to assimilate the embodied knowledge encapsulated in human culture, nor will it help them pass the Turing test.

Ultimately, Denning suggests a significant disconnect exists between human understanding and machine intelligence. As machine intelligence evolves, it may develop forms of tacit knowledge that remain opaque to humans. This divergence poses serious implications for AI safety; since machines cannot interpret the unarticulated nuances of human intent, aligning their actions with our interests could prove impossible.

Denning warns that as AI automation progresses, we may witness the emergence of machine intelligences that, while not fully achieving human-like capabilities, can still present substantial challenges to society. These machines may possess motivations that differ markedly from our own, raising concerns about their potential to create serious issues for humanity.

To navigate this uncertain future safely, Denning advocates for a conscious effort to reaffirm our humanity amid the rise of intelligent machines. He emphasizes the importance of recognizing and celebrating the differences that define human experience, resisting the notion of being subservient to machine intelligence, and acknowledging that an emerging AI landscape will require significant adjustments in our understanding and interactions with these technologies.

Leading computer scientist claims true human-level AI may always be unattainable.

Dark, candlelit study at midnight: a lone, silver-haired computer scientist stands before a massive chalkboard filled with fading neural network equations. Beyond the board, a warm-glowing, brain-shaped silhouette hovers over an unbridgeable chasm, the single desk lamp casting long, dramatic shadows of doubt.

True human-level AI may be forever out of reach, prominent computer scientist argues - Taylor & Francis Newsroom

- Peter J. Denning argues that Turing's ideas led AI development down a flawed path.  
- Denning highlights tacit knowledge as a barrier to achieving human-level artificial intelligence.  
- The mutual incomprehension between humans and machines poses significant AI safety concerns.  

Book publication announcement

True human-level AI may be forever out of reach, prominent computer scientist argues

Expert suggests there is a fundamental flaw in 75-year quest for artificial general intelligence



Photo by Alex Segre for Alamy. Photo by Alex Segre for Alamy.

Alan Turing, the father of theoretical computer science, made a proposal that has put AI development on a flawed path for three-quarters of a century, a prominent computer scientist has argued.

In his provocative new analysis in Turing’s Mistake: Escaping the Yoke of Unintelligent Machines, Peter J. Denning tells us that Turing’s stance in 1950 reflected a belief held by the scientific community at the time: that human intelligence can exist without a body – and can therefore emerge in software on digital computers.

Denning also challenges the belief that machine intelligence can be confirmed through an imitation game (now known as the Turing test).

“These two claims have shaped much of AI research and development,” Denning writes. “My premise is that our acquiescence to these claims has led to the AI mess in which we find ourselves today.”

He argues that the artificial intelligence (AI) society is headed for won’t yield a human-level intelligence, known as artificial general intelligence (AGI) – instead it will be dangerous, he warns.

The tacit knowledge problem

At the core of Denning’s argument lies a concept called tacit knowledge. Tacit knowledge represents the vast area of human understanding that cannot be articulated in words or captured in any symbolic form that machines can process.

Denning identifies five major domains of tacit knowledge that he says ‘elude machine learning’. These include common sense knowledge, our daily interactions with others and our environment, our feelings and perceptions, performance skills, and our social and historical culture.

Humans have tried to catalogue common sense knowledge. Beginning in the 1980s, Douglas Lenat’s ambitious Cyc project attempted to compile a comprehensive database of common-sense facts. After 40 years of human effort, it had accumulated 25 million entries.

“Yet even this treasury could not add up to a background of common sense sufficient to make expert systems smart enough to be experts,” Denning notes. “Cyc validated that much of the knowledge that makes people experts cannot be articulated as propositions.”

Performance skill presents another insurmountable barrier.

“Our performance skills in thousands of domains cannot be communicated to machines,” Denning explains. “Whereas descriptions of skillful outcomes (‘know what’) can often be represented as bits and stored in a machine, we do not know how to encode the embodied knowledge for skillful performance (‘know how’).”

Musicians demonstrate this this gap. Denning says: “A virtuoso violinist can play beautiful music yet cannot describe to an acolyte how to produce it.

“Even if a robot could observe and imitate skilled humans, having no biological body, a robot cannot grasp how the musician feels when playing beautiful music or how an audience feels when hearing it.”

Other examples of tacit knowledge include intuitions, gut feelings, spontaneous creativity, and imagination.

The unbreachable barrier

The barrier to all of this is what Denning identifies as ‘the representation problem’.

This fundamental obstacle to achieving human-level AGI is because for any computation to occur, data and instructions must be encoded in physical forms that machines can recognise and process. But tacit knowledge, by its very nature, resists such encoding.

“Behind every word is a deep well of tacit knowledge that gives it meaning,” Denning says. “Words are but symbolic representations of meanings, not the meanings themselves. Commonly used Large Language Models, such as ChatGPT, Claude and Gemini only manipulate words, they cannot know or understand the meaning of what they are saying.”

This creates an unbridgeable divide – because we cannot explain or even understand how tacit knowledge works for humans, we cannot begin to communicate it.

“How we host tacit knowledge is largely a mystery,” Denning admits. “All we know is that it is embodied. We have no idea what we might observe and measure in our bodies to reveal it.”

Context and culture

Beyond individual knowledge, Denning emphasises the role of context – or the circumstances of a situation which gives our statements and actions a broader sense of meaning and purpose.

Context provides innumerable layers of meanings that extend beyond any horizon. Context provides the clue to whether someone is being sarcastic or sincere, or if someone is angry or teasing. Context tells us whether to employ tact or use humour.

“When you inquire into where an assumption of the current context came from, you discover it rests on previous conversations from previous contexts. Each of those in turn rests on further previous conversations and their contexts. This pattern is endless and fractal,” Denning explains.

The cultural dimension of intelligence poses similar challenges.

Culture encompasses our values, norms, judgements, histories, communities and moods, even dynamics of power or care.

“Human conversations are imbued with background assumptions that give meaning and relevance to the words being used,” Denning explains.

“Scaling up LLMs with ever larger neural networks will not enable them to acquire the embodied human knowledge we call culture. LLMs will not attain the objective of the Turing test: to demonstrate machine thought indistinguishable from human thought.”

Ultimately, Denning says there is a mutual incomprehension between humans and machines: artificial neural networks will create a form of machine tacit knowledge that humans cannot understand.

“Machines cannot read our tacit knowledge and we cannot read theirs,” he writes. “We are aliens across an uncrossable divide.”

This has profound implications for AI safety. As machines are unable to read unarticulated human context, aligning them reliably with our intentions may be impossible, Denning warns.

“Through AI automation, agentic networks of machines are likely to develop their own machine intelligence that does not reach the level of human general intelligence but is still quite capable of creating severe problems for humans. This threat is a greater than a take-over by superintelligent machines,” he explains.

“Machine intelligence has different concerns from us and does not appear to care about us. Its ways of thinking and problem-solving look alien to us. We do not yet know how to live safely with these machines.

“Pulling back from an AI automation singularity will demand much from us. We start by accepting that the familiar culture is fading away as intelligent machines appear in our society and we do not know what is coming. We decline to think like machines or be subservient to machines. We refuse to submit to a yoke imposed by low-intelligence machines. Most importantly, we reassert our humanity, declare once again what makes us different from machines, and celebrate those differences.”

True human-level AI may be forever out of reach, prominent computer scientist argues

RICHMOND, Va., July 12, 2026 (GLOBE NEWSWIRE) -- QAIAx Microcities Clinical Trial provided an new patent filing update regarding the development of an open-source military-grade artificial general intelligence (AGI) platform named the QAIAx Cool World (ISAC) platform. This platform enables real-life holoportation of humans, hard objects, holographic agents, and omni-AI humanoids, which can be partially or fully controlled by human users. It operates under a pioneering public-use technology license aimed at regulating AGI, ensuring data privacy, and maintaining legal compliance within the context of an AI smart health cities clinical trial. The technology introduces a 90/10 hybrid human-AI civic workforce model, facilitating advanced virtual reality for social dating, professional conferencing, and alternative prison sentencing . Key components include the AI Wat series "Beam Bot" and the “AVK,” an AI suit that creates a holoportation experience with haptic feedback, marking a significant advancement in the AGI Robotics Sector and the AI Era.

AI Alternative Sentencing & Psychiatric Treatment

The architecture introduces non-medicinal, psychiatric rehabilitation tools built for first-time, quasi-criminal, non-violent offenders, and administrative violators:

AI-YO: A frontier AI/AGI learning machine platform substituting brick-and-mortar incarceration with immersive haptic feedback from the ‘AVK suit’ (omni AI-compatible body-ware used for human control of a Beam Bot or humanoid robot), using quasi-mind ware as a haptic-virtual reality application for behavioral and mental health intervention for criminal justice corrections.



A frontier AI/AGI learning machine platform substituting brick-and-mortar incarceration with immersive haptic feedback from the ‘AVK suit’ (omni AI-compatible body-ware used for human control of a or humanoid robot), using quasi-mind ware as a haptic-virtual reality application for behavioral and mental health intervention for criminal justice corrections. AI Elm Street: A virtual simulation platform delivering intensive cognitive behavioral modifications without drugs, based on the AI-YO (choose-your-own-adventure) application (e.g., court order or a psychiatric doctor’s referral for interventional therapy that requires a criminal offender or administrative violator to participate in 500-5000 hours of cognitive situation content programming; what DUI offenders would typically be required to participate in through M.A.D.D. panel meetings).



A virtual simulation platform delivering intensive cognitive behavioral modifications without drugs, based on the AI-YO (choose-your-own-adventure) application (e.g., court order or a psychiatric doctor’s referral for interventional therapy that requires a criminal offender or administrative violator to participate in 500-5000 hours of cognitive situation content programming; what DUI offenders would typically be required to participate in through M.A.D.D. panel meetings). Village of the Crazies (AGI ecosystem platform): A structured, air-gapped virtual therapeutic enclave (‘microcity’) designed as an alternative sentencing platform to rehabilitate minor offenders, similar to AI Elm Street—imagine unpleasant mock situations with Beam Bot (holo-suites) and humanoid technology to recreate criminal incidents and demonstrate where bad-faith actions arose and what should be done to remain law-abiding in the future.



A structured, air-gapped virtual therapeutic enclave (‘microcity’) designed as an alternative sentencing platform to rehabilitate minor offenders, similar to AI Elm Street—imagine unpleasant mock situations with Beam Bot (holo-suites) and humanoid technology to recreate criminal incidents and demonstrate where bad-faith actions arose and what should be done to remain law-abiding in the future. Wartime Therapy (PTSD treatment): An AGI ecosystem system for other AI-YO platform models, specifically used as a solitary or group intervention solution to prevent suicide or de-escalate cognitive effects on the mental health of military veterans exposed to wartime or other military service-connected anomalies.





Frontier Hardware & The Cool World Dating Game

The platform integrates advanced peer-to-peer interactive environments using everyday consumer tech alongside proprietary hardware:

AVK Suit & Beam Bots: High-tech body arrays that track real-time vitals to deploy physical telepresence. It bridges edge-network compatibility for peer-to-peer capabilities with other frontier AI body-ware, such Meta's AI glasses, directly into 3D environments.



High-tech body arrays that track real-time vitals to deploy physical telepresence. It bridges edge-network compatibility for peer-to-peer capabilities with other Meta's AI glasses, directly into 3D environments. QM-Ware Interface: A sub-50 millisecond latency “Quasi-Mind Ware" cognitive link utilizing visual fractal keys.



A sub-50 millisecond latency “Quasi-Mind Ware" cognitive link utilizing visual fractal keys. Cool World Dating: A global peer-to-peer matchmaking arena using QAIAx Omni-AI robotics tools. Ironically, the developers themed this ‘Holo-Day’ platform after the 1990s surreal live-action and animated fantasy film Cool World starring Brad Pitt. In that movie, a human cartoonist is pulled into his own cartoon dimension, involving a fictional intimate interest played by Michelle Pfeiffer. This new quasi-reality game platform similarly pulls two or more end users into a surreal, haptic virtual world to experience professional or social meetings between people, humanoids, and Beam Bots without physically being together, as in-the-flesh human beings would meet at a single location.





QAIA Pro License & Future Mars Pilot

Civic operations, specialized training, and administrative roles within the smart city require a QAIA Pro License . Professional users—including legal case managers, virtual educators, criminal attorneys, healthcare providers, and civic technicians—use this credential to manage automated networks. Similarly, the clinical trial’s resident patients will be given trainable AI-Me robots (“QAIA Minime”) to master technical certifications, mirroring the historic IT licensing booms of the 1990s that were presented by PC/Internet Era pioneers such as IBM and Microsoft.

Looking forward, developers are actively designing the Just Like Mars Mission (QAIAx v3.0 —called “QIII”; an upscaled version of QAIAx’s second-generation of AI City Hall Project—despite no activities with Space Exploration Technologies Corp., and only passive NASA-affiliate program initiatives that paused during the ‘25 federal shutdown. QIII prototype pilot is anticipated to run as its own clinical trial with an anticipated launch later in 2027. USPTO filings for QIII provide details about a continuum of original QAIAx build as an isolated, intranet off-grid testing environment, sending omni-AI humanoids to outer-world destinations to conduct human managed activities.

Clinical Trial Information

Prospective corporate sponsors, institutional partners, and patients seeking to join the active clinical trial waiting list can submit formal regulatory inquiries through the official ClinicalTrials.gov (PRS-NCT07661823) portal, or visit www.vetsrecovery.org .

Related Topics

QAIAx Clinical Trial Program Press Statement from Key Sponsor (Veterans Recovery Network - June 26, 2026)

Implementation of the QAIAx "Cool World" Alternative Sentencing and Cognitive Rehabilitation Program

Homeless Vets Welcome Here: AI City Hall Project Moves on Omni-AI Microcities



Photos accompanying this announcement are available at:

https://www.globenewswire.com/NewsRoom/AttachmentNg/396f6f1b-701f-4e94-9cab-421541b04740

https://www.globenewswire.com/NewsRoom/AttachmentNg/a9b80074-09e3-4b86-911a-e65d2a0beba2

A video accompanying this announcement is available at https://www.globenewswire.com/NewsRoom/AttachmentNg/bcc9f18c-6991-4ad4-b712-83fad4e19e6f

QAIAx Clinical Trial: AGI Alternative Sentencing & Reality Pro-Social Dating (Veterans First for America)

Generate a cinematic news thumbnail for title:QAIAx Clinical Trial: AGI Alternative Sentencing & Reality Pro-Social Dating (Veterans First for America). No text in image.

QAIAx Clinical Trial: AGI Alternative Sentencing & Reality Pro-Social Dating (Veterans First for America) - markets.businessinsider.com

In 2026, the landscape of artificial intelligence (AI) is defined by a dizzying array of announcements, from the introduction of cutting-edge models to corporate discussions about the implementation of intelligent workflows. However, everything gets grouped under the broad term "AI," which risks obscuring critical differences that are crucial for those looking to differentiate between genuine capabilities and mere buzz. AI is not a singular entity; it exists within a three-level hierarchy. Currently, most applications—from ChatGPT to Gemini and various enterprise tools—reside only within the initial tier. Grasping the true state of this technology versus its marketing portrayal is essential for establishing effective AI strategies rather than pursuing an elusive target that remains out of reach.

The framework of machine intelligence is categorized into three levels: Artificial Narrow Intelligence (ANI), Artificial General Intelligence (AGI), and Artificial Superintelligence (ASI). Each category signifies a qualitatively different level of capability, rather than simply an enhancement of the previous stage. ANI excels at specific tasks but lacks the ability to generalize that expertise. In contrast, AGI is designed to match human-like cognitive flexibility across all areas, whereas ASI would surpass human intelligence entirely. As of now, only ANI is a tangible reality; references to AGI or ASI are often aspirational, describing goals or speculative projections rather than currently operational systems.

Currently, all deployed AI products—from large language models and computer vision technologies to virtual assistants—are classified as Artificial Narrow Intelligence, also known as weak AI. These narrow AI systems can exhibit impressive capabilities within their defined functions but cannot extend their skills beyond that. A model proficient in coding, for example, lacks the ability to diagnose medical issues using the same analytical skills. Similarly, a system that can create stunning images cannot draft an email simultaneously. This limitation is a core attribute of ANI, irrespective of how sophisticated or human-like a system’s responses might appear.

Within the category of ANI, systems are categorized based on their information processing capabilities: reactive machines, which respond to stimuli without retaining context, and limited memory systems that utilize recent dialogue history to enhance their responses. Most contemporary consumer AI applications fall into the limited memory category, which explains why they may seem more effective in extended interactions compared to isolated queries.

The widespread implementation of narrow AI is undeniable. According to Stanford University’s 2026 AI Index Report, global adoption of generative AI soared to 53% within just three years following the debut of ChatGPT, outpacing the adoption rate of personal computers and the internet. Similarly, McKinsey’s State of AI 2025 report indicated that 88% of organizations now leverage AI in at least one aspect of their operations. Although termed narrow by definition, its impact is far-reaching.

However, an under-discussed limitation to the growth of ANI lies in its reliance on data. Large models require vast amounts of text, code, and media sourced from the internet's extensive archives. This reservoir is limited, and several analyses highlight a concerning trend: the generation of fresh, high-quality data needed for training is trailing behind the sheer volume of available content. As existing material is increasingly utilized in training processes, further model enhancements are likely to face diminishing returns unless complemented by synthetic data or new methodologies. Consequently, many researchers characterize today’s leading systems as capable ANI with broader scope rather than as early examples of general intelligence.

Artificial General Intelligence denotes a system with the ability to learn, reason, and act across virtually any domain, mimicking human cognitive flexibility when faced with new tasks. Unlike ANI, which falters outside its specialized fields, an AGI system would adapt and apply knowledge in a manner akin to a proficient generalist.

Currently, no AGI exists. Benchmarks provide compelling evidence, as seen in the ARC Prize Foundation’s 2025 assessment of OpenAI's o3 model, which achieved a score of 87.5% on the ARC-AGI Semi-Private Evaluation—an impressive leap, yet still below the AGI benchmark. Likewise, Stanford’s Technical Performance research noted a significant year-over-year gain in frontier models on tests that resist saturation, underscoring ongoing progress. While momentum toward AGI is clear, the milestone itself has not been reached, despite some voices within the industry touting advanced internal capabilities.

Predictions about when AGI might be realized vary significantly, ranging from the near term of 2027 to several decades away, reflecting a genuine divide among researchers concerning the limits of current architectures before radical shifts become necessary.

Artificial Superintelligence remains a theoretical concept situated beyond AGI. While AGI would match human cognitive capabilities, ASI would exceed the collective intelligence of humanity. Philosopher Nick Bostrom emphasizes that ASI would represent intellect far beyond that of the best human minds across various disciplines.

The hypothesized transition from AGI to ASI involves recursive self-improvement; a system capable of enhancing its own design would generate a more intelligent successor, continuing in a cycle. Some researchers suggest this transition could occur in a matter of months rather than extending over decades, which raises significant safety and alignment considerations distinct from today’s narrower systems. The challenge of value alignment—ensuring advanced systems pursue objectives aligned with human intentions—remains a central unanswered issue in the journey from AGI to a safe ASI.

Understanding the distinctions between ANI, AGI, and ASI carries real implications for organizations in 2026. Oversimplifying these categories into a single concept of "AI" can result in misaligned governance frameworks and miscalculations regarding the abilities of current technologies. This confusion can lead to either overconfidence in the autonomous capabilities of today's tools or underinvestment in the necessary human oversight that narrow systems still require. McKinsey’s research into AI governance revealed that organizations with robust responsible AI practices report higher maturity levels and are more likely to reap tangible benefits, highlighting how clarity regarding the capabilities being managed is vital.

The rapid evolution of AI technology continues to shape the business landscape, remaining powerful yet distinctly narrow. Recognizing this difference is not just an academic notion but a prerequisite for leveraging AI effectively.

Understanding AI: Distinguishing Between AGI, ASI, and ANI Levels

A bright, sunlit tech lab with tall windows; on a polished table sit three glass domes side by side. Under the first, a robotic arm repetitively assembles a tiny car, labeled “ANI.” The second dome encloses a sleek humanoid robot chatting with a researcher, labeled “AGI.” Above the third dome floats a luminous neural-network sphere swirling like a miniature galaxy, labeled “ASI.” A lone scientist’s silhouette stands in the background, natural light highlighting each dome and casting soft shadows.

AGI vs. ASI vs. ANI: The Levels of AI Explained - AI Insider

- AI is categorized into three tiers: Artificial Narrow Intelligence, General, and Superintelligence.  
- Only Artificial Narrow Intelligence currently exists and powers all commercial AI products today.  
- Confusing AI tiers can lead to governance issues and misjudgments in organizational strategies.  

Every AI headline in 2026, from frontier model launches to boardroom debates about agentic workflows, gets filed under the same catch-all word: AI. That flattening obscures a distinction that matters enormously for anyone trying to separate real capability from hype. Artificial intelligence is not one thing. It is a three-tier hierarchy, and almost everything currently deployed, from ChatGPT to Gemini to enterprise copilots, sits at only the first of those three tiers. Understanding where the technology actually stands, rather than where marketing places it, is the difference between building a sound AI strategy and chasing a moving target that has not yet arrived.

THE THREE TIERS OF MACHINE INTELLIGENCE

Researchers classify machine intelligence into three categories: Artificial Narrow Intelligence (ANI), Artificial General Intelligence (AGI), and Artificial Superintelligence (ASI). Each tier represents a qualitatively different capability threshold, not simply a bigger version of the one before it. ANI performs specific tasks well but cannot generalize beyond them. AGI would match human cognitive flexibility across every domain. ASI would exceed the combined intellectual capacity of humanity itself. As of 2026, only the first category is real. Every other reference to AGI or ASI describes a research target, a forecast, or a marketing claim, not a deployed system.

ARTIFICIAL NARROW INTELLIGENCE: WHERE EVERY DEPLOYED SYSTEM LIVES TODAY

Every commercial AI product in existence, from large language models to computer vision systems to voice assistants, is a form of Artificial Narrow Intelligence, also called weak AI. Narrow systems can be remarkably capable within their lane. They cannot transfer that capability outside it. A model that writes exceptional code cannot autonomously diagnose a medical condition using the same underlying reasoning it applies to software, and a model generating an image cannot simultaneously draft an email in the same operation. That single-task boundary is the defining feature of the category, and it holds true regardless of how fluent or humanlike a given system’s outputs appear.

Within ANI, systems are further divided by how they process information: reactive machines that respond to input without retaining context, and limited memory systems that draw on recent conversational history to produce more relevant answers. Most consumer AI tools today fall into the limited memory category, which is one reason the same model can feel dramatically more useful in a long conversation than in a single isolated prompt.

The scale of narrow AI’s real-world deployment is not in question. Stanford University’s 2026 AI Index Report found that global generative AI adoption reached 53% within three years of ChatGPT’s launch, a faster uptake curve than either the personal computer or the internet achieved. McKinsey’s State of AI 2025 research similarly found that 88% of organizations now use AI in at least one business function. The tier is narrow by definition, but its footprint is anything but small.

THE DATA CEILING FACING TODAY’S NARROW SYSTEMS

A less discussed constraint on ANI’s trajectory is the supply of the material it depends on: data. Large models are trained on enormous volumes of text, code, and media scraped from archives that have accumulated online for decades. That stockpile is finite, and multiple analyses have flagged the same pattern, that the growth of freshly generated content is outpacing the growth of genuinely new, high-quality training data, much of which has already been absorbed into prior training runs. Once a model has consumed the readily available corpus of human-generated material, further scaling runs into diminishing returns unless synthetic data, new data-generation techniques, or new architectures compensate for the shortfall. This is one reason serious researchers describe today’s frontier systems as very capable narrow AI with unusually wide narrow ranges, rather than as early-stage general intelligence.

ARTIFICIAL GENERAL INTELLIGENCE: THE THRESHOLD NOBODY HAS CROSSED

Artificial General Intelligence describes a system that can learn, reason, and act across virtually any domain with the same flexibility a human brings to unfamiliar problems. Given a task it was never explicitly trained for, an AGI system would reason from what it already knows and transfer relevant knowledge the way a competent generalist would, rather than failing or producing nonsense the way a narrow system does outside its lane.

No AGI exists today. The clearest evidence comes from benchmark data rather than marketing claims. The ARC Prize Foundation’s 2025 evaluation of OpenAI’s o3 model recorded a score of 87.5% on the ARC-AGI Semi-Private Eval, a substantial jump from earlier systems but still short of the benchmark’s own AGI threshold. Stanford’s Technical Performance research reached a related conclusion from a different angle, finding that frontier models gained roughly 30 percentage points in a single year on Humanity’s Last Exam, a test built specifically to resist saturation, and that evaluations designed to hold up for years are being exhausted in months. Progress toward AGI is real and accelerating. Crossing the threshold is a separate event that has not yet happened, whatever individual executives have suggested about internal capabilities already having “whooshed by” undetected.

Expert timelines for when AGI might arrive vary widely, spanning from 2027 to several decades out, reflecting genuine disagreement among researchers about how far current architectures can be pushed before a fundamentally different approach becomes necessary.

ARTIFICIAL SUPERINTELLIGENCE: THE HYPOTHETICAL BEYOND

Artificial Superintelligence sits beyond AGI on the same continuum and remains entirely hypothetical. Where AGI would match human cognitive ability, ASI would surpass the collective intelligence of all humans simultaneously, across creativity, scientific reasoning, and strategic planning at once. Philosopher Nick Bostrom’s framing, that ASI would be an intellect vastly beyond the best human minds in practically every field, remains the reference point most researchers still use.

The theorized path from AGI to ASI runs through recursive self-improvement, a system smart enough to improve its own architecture, which then produces a still-smarter successor capable of the same feat. Researchers studying this transition describe it as potentially compressing into months rather than decades once it begins, which is precisely why the safety and alignment questions attached to ASI differ structurally from anything associated with today’s narrow tools. The value alignment problem, ensuring a highly capable system pursues objectives the way its designers actually intended, is treated by safety researchers as the central unresolved challenge standing between AGI and any safe transition to ASI.

WHY THE DISTINCTION MATTERS FOR ORGANIZATIONS IN 2026

Collapsing ANI, AGI, and ASI into a single undifferentiated idea of “AI” has practical costs. A governance framework built for narrow classifiers looks nothing like what a system approaching general intelligence would require, and confusing the two leads organizations to either overestimate what today’s tools can be trusted to do autonomously, or underinvest in the human-in-the-loop safeguards that narrow systems still need. McKinsey’s State of AI Trust research found that organizations investing meaningfully in responsible AI governance report materially higher maturity scores and are more likely to realize measurable benefit, a pattern that holds specifically because those organizations are reasoning clearly about which tier of capability they are actually managing.

The technology reshaping business today is powerful, fast-moving, and, by any rigorous definition, still narrow. Recognizing that distinction is not a caveat on AI’s importance. It is the precondition for using it well.

References and Further Reading

As excitement for A.I. grows in Silicon Valley, Pope Leo has some comments.

More from Artificial General Intelligence

DeepMind's CEO reiterates the need for a frontier AI standards organization.

DeepMind's CEO Claims AGI Will Outshine Electricity and Fire

Demis Hassabis warns that humanity has a 'valuable opportunity' to guarantee AGI safety.

Alan Turing's key assumption about AI might have been incorrect.