In a recent report published on Tuesday, Anthropic, a leading artificial intelligence lab based in the United States, achieved the top score in a semiannual safety evaluation. However, the report highlights a concerning trend in the industry worldwide, indicating that efforts to address potential "existential" risks posed by AI are falling short. The findings underscore the necessity for the tech sector to prioritize safety measures in the rapidly evolving landscape of artificial intelligence. 

Additionally, it's worth noting that the content of this article was generated independently, with no involvement from Barron's news team. For further details, you can visit AFP.com. 

© Agence France-Presse

Think Tank Cautions That Global AI Industry Lacks Adequate Safety Measures

A dimly lit boardroom with heavy shadows, a cracked digital globe at the center glowing with frayed red circuit veins, shadowy figures in suits hunched over a stark report labeled “AI Safety Audit – FAIL,” a single narrow window letting in slivers of muted daylight, red hazard symbols and warning flags scattered across the table, an atmosphere of urgent alarm and looming consequences.

Global AI Industry Falls Short On Safety, Think Tank Warns - Barron's

- Anthropic achieved the highest score in a semiannual AI safety ranking.  
- A report warns the AI industry is not addressing existential threat concerns.  
- Global efforts are insufficient to mitigate potential dangers posed by AI technology.  

US artificial intelligence lab Anthropic scored the highest in a semiannual safety ranking, but globally the industry fails to combat "existential" threats, according to a report released on Tuesday.

The Barron's news department was not involved in the creation of the content above. This article was produced by AFP. For more information go to AFP.com.

© Agence France-Presse

Global AI Industry Falls Short On Safety, Think Tank Warns

Amazon (AMZN) has been making headlines recently for trimming its workforce within portions of its Artificial General Intelligence (AGI) division, even as the company continues to make substantial investments in AI. This development comes amid an internal reorganization, raising questions about Amazon's long-term strategy in the AI space.

As the competition in AI evolves beyond simply launching chatbots, tech giants like Amazon are entering a new phase of expansion driven by the rising demand for AI workloads. Amazon plans to allocate roughly $200 billion toward its AI initiatives by 2026, underscoring its commitment to bolstering its infrastructure. Although these considerable investments have impacted the company's stock, Amazon Web Services (AWS) continues to show strong growth, along with a flourishing advertising sector and an expanding array of AI tools.

Amazon has indicated that the recent job cuts are part of an effort to concentrate on AI projects that deliver maximum value to customers, while still investing in AI models, cloud services, and proprietary chips. 

So, how do these adjustments in Amazon's AGI division reflect its overarching aspirations in the AI domain?

A Closer Look at AI Initiatives

The job reductions within Amazon's AGI unit suggest a shift in focus rather than a retreat from AI altogether. The company has articulated that these changes are designed to enable it to prioritize critical AI projects, thereby allowing teams to streamline their efforts on high-impact work.

Emphasizing its commitment to building large AI models, Amazon appears to be honing its focus on investments that promise the greatest influence across its AWS business and beyond. Recent collaborations reinforce this strategy; for instance, Pinterest (PINS) recently selected AWS to support its next-generation AI-driven visual search platform, demonstrating ongoing reliance on AWS for the computational resources necessary for AI development.

Amazon is also strengthening its AI ecosystem through new collaborations. A notable agreement with Meta Platforms (META) allows running agentic AI workloads on AWS Graviton chips, which points to increasing trust in Amazon’s custom hardware. This partnership underscores that Amazon is not only developing AI models but is also making broad investments across its entire AI ecosystem, from custom chips to cloud infrastructure and AI application development services.

Financial Context of the Layoffs

As one of the globe's largest technology firms, Amazon's diverse revenue streams include online retail, AWS, digital advertising, and various tech services. While recent stock performance has been modest—with an uptick of 2.8% over the past year and a 1.6% increase since January—it still trades at a forward price-to-earnings ratio of 31.93, which is significantly above the sector average of 15.98, suggesting that investors are prepared to pay a premium for its shares.

In its first quarterly report, Amazon saw net sales rise 17% year-over-year to $181.5 billion. Notable growth was observed in North America, which increased 12% to $104.1 billion, while international sales rose 19% to $39.8 billion (or 11% when adjusted for foreign exchange). AWS revenue soared by 28%, reaching $37.6 billion. Improvements in operating income from $18.4 billion to $23.9 billion were recorded, driven by contributions from all business sectors. Notably, AWS emerged as the primary earnings contributor with an operating income of $14.2 billion.

Profitability metrics were notably higher as well. Net income surged to $30.3 billion or $2.78 per diluted share, compared to $17.1 billion or $1.59 per share the previous year, aided by $16.8 billion in pre-tax gains from Anthropic. Operating cash flow experienced a 30% increase, totaling $148.5 billion over the last year. However, free cash flow plummeted from $25.9 billion to $1.2 billion due to a significant $59.3 billion investment in property and equipment aimed at enhancing AI infrastructure.

Analyst Projections

Looking ahead, Amazon’s second-quarter earnings report is slated for July 30, with Wall Street projecting earnings of $1.82 per share, reflecting an 8.33% growth compared to last year. For full-year 2026, consensus estimates indicate projected earnings of $7.75 per share, an 8.09% increase from 2025.

KeyBanc has recently raised its price target for Amazon to $325 from $285, citing robust growth in AWS, improved demand in grocery, and the potential of Amazon Leo. Meanwhile, Goldman Sachs analyst Eric Sheridan has bumped his target to $335 from $325, maintaining a "Buy" rating based on Amazon's strong position in AI and cloud services. Although more conservative adjustments have been made—TD Cowen analyst John Blackledge slightly reduced his target from $350 to $340—he still holds a "Buy" recommendation, reflecting positive sentiment overall.

In summary, of the 57 analysts currently covering Amazon, a consensus rating of "Strong Buy" persists, with an average price target of $315.25 suggesting significant upside potential from current levels.

Final Thoughts

The recent layoffs within Amazon's AGI division signify a tactical realignment rather than a retreat from AI initiatives. The company continues to invest in critical AI infrastructure, enhance AWS's capabilities, forge new enterprise partnerships, and garner backing from financial analysts. This strategy indicates a refined approach to executing its AI vision, rather than a retreat from its ambitions. If Amazon can leverage its substantial AI investments into sustained revenue and earnings growth, the prospects for the stock could remain promising in the long run.

Amazon Resizes Its Artificial General Intelligence Team, Viewing It as a Reset Rather Than a Retreat.

A shadowy corporate conference room bathed in warm amber light. A giant Amazon smile logo dominates the back wall. In front, a sleek podium holds an oversized, glowing red RESET button. Empty swivel chairs and half-packed laptop boxes are strewn about, and one silhouetted figure in a sharp suit reaches hesitantly toward the button, hinting at transformation rather than retreat.

Amazon Trims Its Artificial General Intelligence Team. Think of It as a Reset, Not a Pullback. - Barchart.com

- Amazon is cutting jobs in its AGI group to focus on key AI projects.  
- The company plans to invest $200 billion in AI infrastructure by 2026.  
- Amazon's partnerships and revenue growth indicate strong confidence in its AI strategy.

Despite continuing to invest aggressively in AI, Amazon (AMZN) has begun cutting jobs within parts of its Artificial General Intelligence (AGI) group as part of an internal reorganization. That makes Amazon's latest move particularly interesting.

The AI race is no longer just about releasing the next chatbot. Hyperscalers are entering a new phase of growth as AI workloads continue to increase, prompting them to invest heavily in expanding their infrastructure. Amazon alone plans to spend roughly $200 billion in 2026, and those massive AI investments have weighed on the stock , even as Amazon Web Services (AWS) continues to deliver strong growth, the advertising business expands, and the company rolls out more AI products and services.

The company says the changes are meant to help it focus on the AI projects that matter most to customers while it continues investing in AI models, cloud infrastructure, and custom chips.

So, what does this targeted reduction in Amazon's AGI team really say about the company's long-term AI ambitions?

AI Growth Engines Beyond the Headlines

Amazon's decision to cut jobs in parts of its AGI team looks more like a reorganization than a step back from AI. The company said the changes are meant to help it focus on the AI projects that matter most to customers and allow teams to move faster by putting more resources behind higher-priority work.

Amazon has made it clear that building large AI models remains one of its top priorities. If anything, the layoffs suggest the company is becoming more focused on where it spends its time and money, not less committed to AI. The goal appears to be putting more effort into the projects that can have the biggest impact across AWS and the rest of the business.

Recent developments support that view. Earlier this year, Pinterest (PINS) chose AWS to power the next stage of its AI-powered visual search and discovery platform. The partnership shows that businesses continue to rely on AWS for the computing power needed to build and run AI applications, reinforcing demand for Amazon's cloud infrastructure.

Also, Amazon is strengthening its AI business through new partnerships. In another recent move, Meta Platforms (META) signed an agreement to run agentic AI workloads on AWS Graviton chips. That deal highlights growing confidence in Amazon's custom chips and shows the company is building much more than AI models. It is investing across the entire AI stack, from chips and cloud infrastructure to the services customers use to develop and deploy AI applications.

Financial Strength Behind the Cuts

Amazon is one of the world's largest technology companies, generating revenue from online retail, AWS, digital advertising, subscriptions, and other technology services. Gains have been marginal recently. The stock is up 2.8% over the past 52 weeks and has gained 1.6% since the start of the year.

It also trades at a forward price-to-earnings ratio of 31.93 times , well above the sector average of 15.98 times, showing that investors are willing to pay a premium.

In the first quarter , Amazon grew net sales by 17% year-over-year (YOY) to $181.5 billion. North America revenue rose 12% to $104.1 billion, international sales increased 19% to $39.8 billion (or 11% excluding foreign exchange impacts), and AWS revenue jumped 28% to $37.6 billion. Operating income improved from $18.4 billion to $23.9 billion, with every business segment contributing to the increase. AWS remained the biggest earnings driver, generating $14.2 billion in operating income, while North America and international operations contributed $8.3 billion and $1.4 billion, respectively.

Further, profitability improved sharply. Net income climbed to $30.3 billion, or $2.78 per diluted share, from $17.1 billion, or $1.59 per share, a year earlier, helped by $16.8 billion in pre-tax gains from Anthropic. Operating cash flow rose 30% to $148.5 billion over the trailing 12 months. Free cash flow, however, fell from $25.9 billion to $1.2 billion because Amazon spent an additional $59.3 billion on property and equipment, with the company saying most of that increase went toward expanding its AI infrastructure.

What Analysts See Ahead

The company’s second-quarter earnings release is set for July 30 after the market closes. Wall Street expects earnings of $1.82 per share, up 8.33% from $1.68 a year ago. Looking further out, consensus estimates call for full-year 2026 earnings of $7.75 per share, an 8.09% increase over 2025.

In April, KeyBanc raised its price target on Amazon to $325 from $285, pointing to continued strength in Amazon Web Services (AWS), improving grocery demand, and the growth potential of Amazon Leo.

Around the same time, Goldman Sachs analyst Eric Sheridan lifted his target to $335 from $325 while maintaining a "Buy" rating, citing Amazon's strong position in AI and cloud computing.

Even the more cautious updates remain positive. TD Cowen analyst John Blackledge lowered his target slightly to $340 from $350, but kept his "Buy" rating, with his target still well above where the stock trades today.

Overall, of 57 analysts covering the stock currently, a consensus rates it a "Strong Buy," and their average price target of $315.25 suggests 34.4% upside from current price levels.

Conclusion

Amazon's AGI layoffs look more like a strategic reset than a sign that it's pulling back from artificial intelligence. The company is still spending heavily on AI infrastructure, expanding AWS's AI capabilities, signing new enterprise partnerships, and earning strong support from Wall Street. That combination suggests Amazon is refining how it executes its AI strategy, not reducing its ambitions. If management continues turning those massive AI investments into stronger revenue and earnings growth, the stock has a solid chance of moving higher over the long term.

Amazon Trims Its Artificial General Intelligence Team. Think of It as a Reset, Not a Pullback.

Amazon has announced job reductions within its artificial intelligence division, specifically targeting its Artificial General Intelligence (AGI) Unit. The decision was confirmed on Thursday to KIRO 7 News.

A company representative stated, "We have been developing extensive AI models for several years, and this remains one of our top priorities. As this field continues to evolve rapidly, we are honing our focus on initiatives that are most beneficial for our customers, allowing us to accelerate our efforts on what truly matters. This refined focus necessitates tough choices, including the elimination of certain positions within our AGI team, while still prioritizing investments in areas crucial for the future of our customers. We are dedicated to aiding those affected during this transition period." Support for U.S. employees who are impacted will include 90 days of salary and benefits, assistance with job searches, transitional health coverage, and potential severance payments. The company expressed gratitude for the contributions of all employees affected.

Artificial general intelligence refers to an advanced AI system with the capability to equal or exceed human intelligence, with the ability to learn, evolve, and operate independently.

As of now, Amazon has not disclosed the specific number of employees affected or their locations. Earlier this year, the company announced the layoff of 16,000 workers, which included staff in Washington state.

Stay tuned for further updates as this story progresses.

Amazon reveals job cuts in its artificial general intelligence division

A dim, tense AGI research lab lit by low tungsten bulbs, empty workstations with neural-network diagrams on screens flashing “Position Eliminated,” silhouettes of professionals walking away in shadow; a large Amazon logo looming on a concrete wall.

Amazon announces layoffs in artificial general intelligence group - KIRO 7 News Seattle

- Amazon is laying off employees in its Artificial General Intelligence Unit.  
- The company is focusing on key initiatives while supporting affected employees during transition.  
- Specific details about the number of layoffs have not been disclosed yet.  

Amazon is laying off some of its employees working in artificial intelligence.

On Thursday, the company confirmed to KIRO 7 News that it’s making cuts to its AGI – Artificial General Intelligence Unit.

It reads:

“We’ve been building large AI models for several years, and it remains one of the most important things we’re working on. This is a fast-moving space, and we’re sharpening our focus on the initiatives that matter most for customers, so we can move faster on what counts. That focus means some difficult decisions, including eliminating some roles within parts of our AGI organization, even as we continue to invest in the areas most important to our customers’ future. We’re committed to supporting impacted employees through their transition. Employees in the U.S. will receive their pay and benefits for 90 days, outplacement job search support, access to transitional health benefits, and be eligible for a severance payment. We’re grateful for the contributions of every impacted employee.”

Artificial general intelligence is a hypothetical type of AI system that can match or surpass human intelligence and can learn, grow, and operate autonomously.

Amazon has not confirmed how many employees are impacted in this round of layoffs or where they are based.

In January, the company confirmed it was laying off 16,000 employees – including some in Washington.

This is a developing story. Check back for updates.

©2026 Cox Media Group

Amazon announces layoffs in artificial general intelligence group

Alphabet, the parent company of Google, is strategically managing its AI accelerator resources to focus on research for artificial general intelligence (AGI), rather than exclusively renting them out to clients. CEO Sundar Pichai shared insights about these priorities during the recent second-quarter earnings call, confirming that the company has begun selling its tensor processing units (TPUs) to select customers.

In response to a question from Goldman Sachs analyst Eric Sheridan regarding how Alphabet balances client demand for TPUs with its own processing needs, Pichai emphasized that the primary focus is on ensuring the necessary resources for advancing AGI. He described AGI, which mimics human intelligence, as foundational to the company’s mission.

Another analyst, Mark Shmulik from Bernstein, inquired about how Google distributes processing power between its search services, cloud operations, and AI model training. Pichai reiterated that the starting point for resource allocation is the pursuit of AGI development. He also pointed out that Alphabet is committed to its key product areas, including Search and YouTube, alongside its Cloud offerings. In the realm of Google Cloud, the company is concentrating compute resources to support initiatives like Vertex and Gemini Enterprise, as well as vital services related to data analytics and cybersecurity.

Google’s core services are experiencing remarkable growth, as reflected in its latest financial results. The company reported an 82 percent year-over-year increase in Cloud revenue, amounting to $24.75 billion for the quarter, which contributed to an impressive profit of $8.8 billion, a 214 percent rise. This surge is attributed to robust demand for AI infrastructure and solutions. Moreover, Google Cloud has a commendable $514 billion in future commitments from clients who have signed up for services yet to be utilized.

The search business saw a 17 percent revenue increase, while YouTube ad revenue grew by 13 percent. Despite initial fears that generative AI might undermine Google’s search advertising, Pichai noted that their AI-enhanced search mode is actually contributing to an overall rise in search queries. 

Implementing AI search requires high-performance hardware that comes with significant costs. Pichai assured stakeholders that the company is effectively managing these expenses, stating that recent engineering advancements have led to the lowest costs for AI mode responses since its inception, all while enhancing AI capabilities.

To support its ambitious AI infrastructure goals, the company is prepared to invest between $195 billion and $205 billion this financial year, a notable increase from the earlier estimation of $180 billion to $190 billion, according to CFO Anat Ashkenazi. However, challenges in supply constraints have prompted Google to expand third-party compute usage in the upcoming quarter as a temporary solution while it builds more in-house capacity.

Pichai explained that acquiring additional capacity is crucial for accommodating large cloud customers. "There are very, very large clients on our Cloud platform that we aim to support during this exceptional period," he highlighted, adding that while short-term costs could be elevated, the long-term partnership potential offers significant returns.

Alphabet also reported total quarterly revenue of $119.8 billion, up 24 percent year over year, with an operating income of $40.8 billion, reflecting a 34 percent increase. Despite these impressive figures, Google experienced a negative free cash flow of $5.9 billion for the first time since 2004, which caused stock prices to dip by four percent in after-hours trading as investors reacted.

Google is stockpiling TPUs to create Artificial General Intelligence.

In a shadowy underground vault, rows of glowing TPU modules glow like high-tech treasure on steel shelves, a massive Google logo emblazoned above the vault door, dim amber spotlights casting deep, ominous shadows, and a metal table in the foreground holds a rolled-out blueprint labeled “AGI.”

Google is hoarding TPUs to develop Artificial General Intelligence - The Register

- Alphabet prioritizes AI accelerators for AGI research over third-party TPUs for customers.  
- Google Cloud revenue surged 82% year-over-year, driven by strong AI demands.  
- Despite record revenue, Google reported a negative cash flow for the first time since 2004.  

Yet the Chocolate Factory is also buying more third-party compute capacity to handle demand for the G-Cloud

Google’s parent company Alphabet is managing its fleet of AI accelerators to prioritize research on artificial general intelligence, rather than renting them all to customers.

CEO Sundar Pichai revealed the company’s priorities during its second quarter earnings call, during which the company confirmed it has delivered on its plan to sell its tensor processing units (TPUs) to some customers. In response to news of those sales, Goldman Sachs analyst Eric Sheridan asked how Alphabet balances demand from customers who want to buy its TPUs and the web giant’s own need for processing power.

“In terms of allocating our TPUs … our first priority is making sure we are allocating what we need to compete at the frontier in terms of AGI development,” Pichai replied, referring to Artificial General Intelligence –AIs that possesses human-like intelligence.

REG AD

“That is the foundation for everything we do,” Pichai added.

REG AD

Another analyst, Mark Shmulik of Bernstein, revisited the matter by asking how Google allocates processing capacity among its search business, cloud operation, and model training efforts.

“On allocation, I think the baseline with which we start is what it takes to continue AGI development at the frontier,” Pichai responded. The CEO said Alphabet is also “prioritizing our core product areas like Search, YouTube, et cetera, as well as Cloud.” And in the G-Cloud, Google is “prioritizing the compute to make sure we can serve our models in the context of Vertex and Gemini Enterprise, and our core solutions, be it data analytics and cybersecurity.”

“Our core services for our core products across consumers and enterprises is where the compute is primarily going, and that’s how we think about it,” the CEO added.

Google’s core services are going gangbusters.

MORE CONTEXT







Google Cloud revenue leapt 82 percent year over year, to $24.75 billion for the quarter, and delivered an $8.8 billion profit which represented 214 percent growth. The Big G said that growth came from “strong demand for AI infrastructure and AI solutions.” There’s probably more to come as the G-Cloud now has $514 billion of cloudy backlog on the books, meaning customers have signed up for services they’re yet to consume.

Search revenue grew 17 percent and YouTube ads grew 13 percent.

When generative AI came along some pundits suggested it could threaten Google’s search ads biz. That was not a good take because Pichai said the company’s AI Mode for search is “driving an incremental increase in Search queries overall.”

AI search needs specialist hardware that is expensive to buy and run. Pichai said Google is on top of that. “Thanks to our engineering and hardware optimizations, this quarter we reduced the cost of AI Mode responses to its lowest level since launch, even as we’ve brought more advanced AI capabilities.”

REG AD

Google continues to spend megabucks on AI infrastructure – CFO Anat Ashkenazi said the company now plans to spend between $195 billion and $205 billion this financial year, up from a previous estimate of $180 billion to 190 billion.

Ashkenazi said Google can’t get all the kit it needs due to what she described as “the supply-constrained environment.” Google therefore plans to “expand the use of third-party capacity in Q3 as a bridging strategy while we build out more internal capacity.”

Pichai said buying bridging capacity will help Google to land monster cloud clients.

“There are very, very large customers of ours on Cloud who we are trying to support them through this extraordinary moment,” the CEO said. “The incremental opportunities they are bringing to us, while a short-term cost over a few months may be very high, in the lifetime of the deal, as we bring more capacity on, is highly ROI positive.”

“Those are factors we are taking into account. Are you willing to take upfront six-month deal to be able to serve that customer in what is a multi-year opportunity, where the margins and the returns are very, very attractive over that multi-year horizon?”

Alphabet’s quarterly revenue landed at $119.8 billion, up 24 percent year over year. Operating income hit $40.8 billion, up 34 percent.

Yet even those torrents of money couldn’t stop Google’s free cash flow landing at -$5.9 billion – the first time the company hasn’t had spare cash to splash since 2004. Investors seem not to like that and sent the price of the company’s shares down by four percent in after hours trading. ®

Google is hoarding TPUs to develop Artificial General Intelligence

Google's parent company, Alphabet, is strategically managing its array of AI accelerators, focusing primarily on advancing artificial general intelligence (AGI) rather than fully utilizing them for client rentals. 

During the second quarter earnings call, CEO Sundar Pichai outlined the company’s emphasis on AGI development. He confirmed that Alphabet has successfully begun selling its tensor processing units (TPUs) to select customers and addressed a question from Goldman Sachs analyst Eric Sheridan about how the company balances its internal processing needs against external demand.

Pichai explained, “In terms of allocating our TPUs ... our first priority is ensuring we are positioned to compete at the forefront of AGI development,” highlighting the significance of human-like AI capabilities to the company’s overall mission. 

Analyst Mark Shmulik from Bernstein also raised the topic of how Alphabet distributes processing resources among its various sectors, including search, cloud services, and AI model training. Pichai reiterated that AGI remains the baseline for resource allocation, stating that the company is also focusing on essential areas like Search, YouTube, and Cloud. Specifically, he noted that compute resources are being targeted to support models within the Vertex and Gemini Enterprise frameworks, along with critical services in data analytics and cybersecurity.

The performance of Google’s core services has been notably strong. The Google Cloud sector saw an impressive 82% year-over-year revenue increase, reaching $24.75 billion for the quarter and yielding an $8.8 billion profit, a remarkable growth of 214%. This surge is attributed to rising demand for AI infrastructure and solutions. With a substantial $514 billion in future service commitments, the G-Cloud anticipates further growth.

Additionally, search revenue rose by 17%, while YouTube ad revenue increased by 13%. Despite initial concerns that generative AI might disrupt Google’s search ad revenue, Pichai stated that the introduction of AI Mode in search has actually led to an increase in overall search queries.

AI search requires specialized hardware that can be costly to acquire and maintain. Nevertheless, Pichai announced that through engineering and hardware optimizations, Google has successfully reduced the cost of responses from AI Mode to its lowest rate since its introduction, even while enhancing the underlying capabilities.

The company is making significant investments in AI infrastructure, with CFO Anat Ashkenazi indicating plans to allocate between $195 billion and $205 billion in the current financial year, an increase from earlier projections of $180 billion to $190 billion. However, Ashkenazi noted that constraints in supply have necessitated a short-term strategy involving increased reliance on third-party computing capacity as the company builds its own infrastructure.

Pichai mentioned that this strategy for acquiring additional capacity will enable Google to secure major cloud clients. He stated, “We are trying to support very large customers during this extraordinary time,” emphasizing the importance of balancing short-term costs with long-term returns in these multi-year engagements.

For the quarter, Alphabet reported a total revenue of $119.8 billion, up 24% from the previous year, and an operating income of $40.8 billion, a 34% increase. However, despite this financial success, the company recorded a free cash flow of -$5.9 billion, marking the first occurrence of negative cash flow since 2004. This prompted a four percent drop in the company’s stock price during after-hours trading, reflecting investor concerns.

Google is stockpiling TPUs in pursuit of artificial general intelligence.

A dark, industrial warehouse bathed in low, moody shadows, its high shelves and crates overflowing with TPU boards that glint in a single beam of moonlight. A massive Google logo is emblazoned on a reinforced vault door at the far end. In the dim foreground, a lone engineer’s silhouette studies a chalk-drawn AGI brain sketch on a concrete wall—tension crackling in the air as the TPU hoard promises the next leap toward artificial general intelligence.

Google is hoarding TPUs to chase artificial general intelligence - The Register

- Google prioritizes AI research over third-party TPU sales, focusing on AGI development.  
- Google Cloud revenue surged 82%, driven by strong AI infrastructure demand and growth.  
- Cash flow turned negative for the first time since 2004, causing shares to drop.  

Yet the Chocolate Factory is also buying more third-party compute capacity to handle demand for the G-Cloud

Google’s parent company Alphabet is managing its fleet of AI accelerators to prioritize research on artificial general intelligence, rather than renting them all to customers.

CEO Sundar Pichai revealed the company’s priorities during its second quarter earnings call, during which the company confirmed it has delivered on its plan to sell its tensor processing units (TPUs) to some customers. In response to news of those sales, Goldman Sachs analyst Eric Sheridan asked how Alphabet balances demand from customers who want to buy its TPUs and the web giant’s own need for processing power.

“In terms of allocating our TPUs … our first priority is making sure we are allocating what we need to compete at the frontier in terms of AGI development,” Pichai replied, referring to Artificial General Intelligence –AIs that possesses human-like intelligence.

“That is the foundation for everything we do,” Pichai added.

Another analyst, Mark Shmulik of Bernstein, revisited the matter by asking how Google allocates processing capacity among its search business, cloud operation, and model training efforts.

REG AD

“On allocation, I think the baseline with which we start is what it takes to continue AGI development at the frontier,” Pichai responded. The CEO said Alphabet is also “prioritizing our core product areas like Search, YouTube, et cetera, as well as Cloud.” And in the G-Cloud, Google is “prioritizing the compute to make sure we can serve our models in the context of Vertex and Gemini Enterprise, and our core solutions, be it data analytics and cybersecurity.”

“Our core services for our core products across consumers and enterprises is where the compute is primarily going, and that’s how we think about it,” the CEO added.

Google’s core services are going gangbusters.

MORE CONTEXT







Google Cloud revenue leapt 82 percent year over year, to $24.75 billion for the quarter, and delivered an $8.8 billion profit which represented 214 percent growth. The Big G said that growth came from “strong demand for AI infrastructure and AI solutions.” There’s probably more to come as the G-Cloud now has $514 billion of cloudy backlog on the books, meaning customers have signed up for services they’re yet to consume.

REG AD

Search revenue grew 17 percent and YouTube ads grew 13 percent.

When generative AI came along some pundits suggested it could threaten Google’s search ads biz. That was not a good take because Pichai said the company’s AI Mode for search is “driving an incremental increase in Search queries overall.”

AI search needs specialist hardware that is expensive to buy and run. Pichai said Google is on top of that. “Thanks to our engineering and hardware optimizations, this quarter we reduced the cost of AI Mode responses to its lowest level since launch, even as we’ve brought more advanced AI capabilities.”

Google continues to spend megabucks on AI infrastructure – CFO Anat Ashkenazi said the company now plans to spend between $195 billion and $205 billion this financial year, up from a previous estimate of $180 billion to 190 billion.

Ashkenazi said Google can’t get all the kit it needs due to what she described as “the supply-constrained environment.” Google therefore plans to “expand the use of third-party capacity in Q3 as a bridging strategy while we build out more internal capacity.”

Pichai said buying bridging capacity will help Google to land monster cloud clients.

REG AD

“There are very, very large customers of ours on Cloud who we are trying to support them through this extraordinary moment,” the CEO said. “The incremental opportunities they are bringing to us, while a short-term cost over a few months may be very high, in the lifetime of the deal, as we bring more capacity on, is highly ROI positive.”

“Those are factors we are taking into account. Are you willing to take upfront six-month deal to be able to serve that customer in what is a multi-year opportunity, where the margins and the returns are very, very attractive over that multi-year horizon?”

Alphabet’s quarterly revenue landed at $119.8 billion, up 24 percent year over year. Operating income hit $40.8 billion, up 34 percent.

Yet even those torrents of money couldn’t stop Google’s free cash flow landing at -$5.9 billion – the first time the company hasn’t had spare cash to splash since 2004. Investors seem not to like that and sent the price of the company’s shares down by four percent in after hours trading. ®

Google is hoarding TPUs to chase artificial general intelligence

On Wednesday, Amazon announced job cuts affecting its Artificial General Intelligence (AGI) division. The company did not specify the exact number of employees impacted nor the specific areas within the AGI group that were involved in these layoffs. This division is crucial for Amazon's AI development efforts and includes teams dedicated to silicon engineering and quantum computing projects.

An Amazon spokesperson stated, “This is a fast-moving space, and we’re sharpening our focus on the initiatives that matter most for customers, so we can move faster on what counts.” This strategic shift necessitated tough choices, including job eliminations in parts of the AGI team, while the company continues to invest in customer-centric areas.

As Amazon seeks to compete with industry leaders such as OpenAI, Anthropic, and Google, the AGI group plays a vital role in its AI strategy. AGI aims to create systems that can match or surpass human capabilities across various tasks.

In a follow-up comment after inquiries, an Amazon representative emphasized the importance of developing large AI models and reaffirmed that the company is honing its focus on essential initiatives to enhance customer outcomes. This focus has led to the unfortunate decision to reduce some positions within the AGI group.

The AGI division recently underwent significant leadership changes, with Rohit Prasad, a key figure in the AGI space, leaving at the end of last year, followed by David Luan, head of the AGI Lab, who departed in February. Following these exits, AGI was reorganized under Senior Vice President Peter DeSantis in December 2025 as part of a larger team that also handles silicon and quantum computing efforts.

Reports from online forums indicated that employees from the teams led by Adeeb Shanaa, vice president of AGI data services, and Vishal Sharma, vice president of AGI information, were among those affected by the recent layoffs.

Earlier this year, Amazon had already announced the elimination of around 16,000 corporate positions, bringing the total number of layoffs to about 30,000 since October 2025, with indications that further cuts may still be on the horizon.

Although these roughly 30,000 layoffs represent a mere fraction of Amazon’s 1.58 million workforce—most of whom are engaged in warehouse and fulfillment roles—they constitute nearly 10 percent of the corporate staff. This marks the most significant reduction in Amazon's history, surpassing the approximately 27,000 jobs cut between late 2022 and early 2023.

Amazon plans to terminate staff in its Artificial General Intelligence division.

A moody, dimly lit AGI research wing with red emergency exit lights casting long shadows. A giant Amazon smile logo looms on the far wall. Desolate desks lined with half-booted laptops and fading AI code on monitors. A few professional silhouettes carry boxes of personal items toward a dark exit, heads bowed, as flickering screens go dark.

Amazon to lay off employees in Artificial General Intelligence unit - The American Bazaar

- Amazon is laying off employees within its Artificial General Intelligence (AGI) group.  
- Specific numbers and impacted areas of the AGI unit remain undisclosed by Amazon.  
- This decision comes amid ongoing layoffs totaling about 30,000 corporate jobs since October 2025.

Amazon said Wednesday that it is cutting jobs within its Artificial General Intelligence (AGI) group. The company did not disclose how many employees were affected or which parts of the AGI unit were impacted by the layoffs. The AGI group focuses on building AI models and includes teams working on silicon development and quantum computing initiatives.

“This is a fast-moving space, and we’re sharpening our focus on the initiatives that matter most for customers, so we can move faster on what counts,” an Amazon spokesperson told CNBC in a statement. “That focus means some difficult decisions, including eliminating some roles within parts of our AGI organization, even as we continue to invest in the areas most important to our customers’ future.”

READ: Meta layoffs put four employees on visa at risk of leaving US (July 21, 2026)

The AGI group is a core part of Amazon’s AI strategy as the company tries to keep up with competitors like OpenAI, Anthropic, and Google. AGI generally refers to AI that can perform as well or better than humans on most tasks.

“We’ve been building large AI models for several years, ​and it remains one of the most important things we’re working on,” said an Amazon spokesman following a Reuters inquiry. “We’re sharpening our focus on the ​initiatives that matter most for customers, so we can move ​faster on what counts. That focus means some difficult decisions, including eliminating some ‌roles ⁠within parts of our AGI organization.”

Rohit Prasad, a top Amazon executive overseeing AGI, left the company at the end of last year, and the head of its AGI Lab David Luan left ​in February. AGI work ​was consolidated ⁠under Senior Vice President Peter DeSantis in December 2025 as part of a larger group that also ​includes silicon development and quantum computing.

According to Reuters, employees under Adeeb ​Shanaa, vice ⁠president of artificial general intelligence data services, and Vishal Sharma, vice president of AGI information, reported being impacted by the cuts on ⁠online ​forums on Wednesday.

READ: ‘Reduced to coin flip’: Amazon engineer pursues O-1 visa after H-1B lottery loss (July 17, 2026)

Earlier this year, Amazon announced it had cut about 16,000 corporate jobs, bringing total layoffs to roughly 30,000 since October 2025, and signaled that more reductions could still be ahead.

While roughly 30,000 layoffs account for only a small slice of Amazon’s 1.58 million employees, most of whom work in warehouses and fulfillment centers, the cuts amount to nearly 10 percent of its corporate staff and mark the largest workforce reduction in the company’s three-decade history. The scale surpasses the roughly 27,000 jobs Amazon eliminated between late 2022 and early 2023.

Amazon to lay off employees in Artificial General Intelligence unit

Amazon is undergoing significant changes within its artificial general intelligence (AGI) unit, announcing job reductions on July 22, 2026, as part of a broader strategy to intensify focus on key projects in a fiercely competitive AI environment. 

The AGI division, which aims to create AI systems that could potentially outperform human intelligence in learning and decision-making, is being streamlined as part of the company’s effort to concentrate its resources where they can yield the highest impact. Although Amazon has not specified the number of positions eliminated, the company has previously cut around 16,000 jobs in January 2026, suggesting a concerted effort to refine its workforce.

A spokesperson for Amazon indicated that the decision to reduce AGI personnel was not taken lightly. According to them, these "difficult decisions" are essential for allowing the company to prioritize initiatives that promise substantial benefits for customers. Essentially, the aim is to optimize operations to be more competitive in the AI sector.

Amazon’s approach is driven by a clear objective: to enhance its research and development in areas where it can outpace its rivals. In light of substantial investments from industry giants like Microsoft and OpenAI, Amazon is recalibrating its strategy to ensure it remains a significant contender in the AGI field.

While the specific number of job losses remains undisclosed, the company’s reticence may reflect a desire to manage investor expectations and maintain a positive public image. The ambitious nature of Amazon's AGI project continues to exist, yet the focus is firmly on delivering practical advancements that resonate with current customer needs.

This trend of job cuts within Amazon mirrors a broader theme in the technology sector, which is characterized by an unceasing drive for excellence in AI. AGI, often regarded as the ultimate goal in artificial intelligence, has many tech firms and researchers vying to make significant breakthroughs. The idea of creating machines with intelligence that exceeds human capability may feel like a futuristic fantasy, yet the race among major companies is very much alive.

The implications of advancements in AGI are profound, with the potential to transform various industries, including healthcare and finance. By narrowing its focus, Amazon strives to position itself favorably amidst a landscape where competitors are rapidly advancing their AI capabilities.

For investors, Amazon's workforce reductions may seem counterintuitive given the growth in the tech sector. However, in the high-pressure realm of AI, prioritizing strategic efficiency over sheer numbers could be more advantageous. Leadership in AGI is not solely dependent on the size of a company’s workforce but rather on the ability to deploy resources effectively for faster development of innovative solutions.

Interestingly, while the AGI unit is experiencing cuts, Amazon's commitment to other innovative ventures, including blockchain, appears to be intact. This hints at a broader strategy that spans multiple technology sectors.

In summary, Amazon’s recent actions reflect the escalating competition in the AI sector, demonstrating its commitment to strategic investments and a more streamlined workforce. As the tech industry evolves, Amazon is positioning itself to remain at the forefront of AI developments, eager to play a pivotal role in the next wave of technological breakthroughs.

Amazon reduces workforce in its artificial general intelligence division.

A dimly lit AGI research lab with rows of dark server racks and red warning lights pulsing; a lone silhouette carries a cardboard box of personal items past a large, illuminated Amazon logo etched on a glass wall. Moody, cinematic shadows and a heavy, dark theme convey the sudden shutdown.

Amazon cuts jobs in artificial general intelligence unit - Crypto Briefing

- Amazon's AGI division faces significant job cuts as part of strategic resource realignment.  
- The layoffs aim to streamline efforts towards high-priority initiatives in AI development.  
- Amazon's focus shifts to ensuring competitiveness against rivals like Microsoft and OpenAI.  

Amazon trims its AGI workforce as part of broader strategic refocusing amid intense AI competition

Amazon’s artificial general intelligence (AGI) unit is experiencing significant changes, as the tech giant announced a round of job cuts on July 22, 2026. These layoffs are part of Amazon’s strategic efforts to sharpen focus on high-priority initiatives within the competitive landscape of artificial intelligence development.

The AGI division, an ambitious project aiming to develop AI systems capable of surpassing human-level learning and decision-making, has faced the axe as part of the company’s broader streamlining process. The specifics of how many roles have been affected remain elusive, with Amazon keeping details under wraps. This move follows a significant workforce reduction earlier in the year, where approximately 16,000 employees were let go in January 2026.

Amazon’s strategic shift

The decision to cut jobs in the AGI division wasn’t made lightly. As an Amazon spokesperson noted, these are ‘difficult decisions’ vital for redirecting focus on projects that hold the most promise for customers. In English: trimming the fat to sprint faster in the AI race.

Amazon’s rationale for this reduction is clear: it’s all about bolstering development in areas where they can outpace competitors. With tech behemoths like Microsoft and OpenAI aggressively investing in AI infrastructure and models, Amazon is realigning its resources to ensure it remains a key player in the AGI arena.

Advertisement

Though Amazon hasn’t disclosed the precise number of job cuts in this latest round, the lack of transparency suggests a cautious approach to managing investor expectations and public perception. While the AGI project is ambitiously futuristic, the critical focus for Amazon remains on delivering tangible benefits in areas that customers care about now.

The bigger picture in AI

Amazon’s job cuts reflect a larger trend in the tech industry: the relentless pursuit of AI excellence. AGI, often seen as the holy grail of artificial intelligence, has researchers and companies grappling to be the first to crack its code. The concept of machines with intelligence surpassing that of humans may sound like something from a sci-fi movie, but the race is very real among today’s tech giants.

The stakes in this race are enormous. Developments in AGI could revolutionize industries from healthcare to finance, creating a seismic shift in how businesses operate. By focusing on narrowing its efforts, Amazon aims to ensure it isn’t left behind as competitors drive towards potentially groundbreaking advances in AI capabilities.

What it means for the market

For investors, Amazon’s decision to cut jobs might appear contradictory in an industry that’s booming. But in the high-stakes environment of AI, efficiency and strategic focus could prove more valuable than sheer manpower.

Competition to lead in AGI isn’t just about who has the largest workforce; it’s about who can smartly leverage their resources to bring cutting-edge solutions to market faster. Amazon’s strategy underlines the importance of agility and precision when investing in emerging technologies.

Interestingly, the focus on AGI does not seem to overshadow Amazon’s commitment to other innovative domains like crypto or blockchain. Although the AGI division faces the brunt of recent cuts, projects in blockchain appear unaffected, signaling Amazon’s diversified approach to future tech investments.

As the dust settles, what remains clear is that the battle for AI supremacy is heating up, and Amazon’s latest moves are a testament to the relentless push in the tech landscape. The emphasis on strategic investments and a refined workforce is setting the stage for the next phase of AI breakthroughs—one where Amazon undoubtedly wants a front-row seat.

Amazon cuts jobs in artificial general intelligence unit

Amazon has confirmed that it is reducing its workforce within its artificial general intelligence (AGI) division, as the company continues to streamline operations while heavily investing in AI technologies. The specific number of impacted employees and the particular segments of the AGI organization affected by these layoffs were not disclosed.

The AGI unit is responsible for developing advanced AI models and encompasses teams dedicated to silicon design and quantum computing projects. An Amazon spokesperson explained to CNBC, “In this rapidly evolving field, we are honing in on the initiatives that matter most to our customers, which necessitates difficult choices, including the elimination of certain roles within our AGI division, while we simultaneously invest in areas crucial for our customers’ future.”

Employees primarily engaged in model customization and post-training phases reported their job loss on LinkedIn. Reuters was the first to break the news of these layoffs.

In recent years, Amazon has been retracting its workforce following a significant hiring surge during the pandemic and as it allocates substantial funds toward developing AI infrastructure. The company has laid off over 30,000 employees since October, continuing to make reductions in small increments in subsequent months.

The AGI division is integral to Amazon's artificial intelligence strategy as the company aims to compete with prominent players like OpenAI, Anthropic, and Google. AGI typically refers to AI capable of performing most tasks as well as, or better than, humans.

In 2024, Amazon’s AGI group introduced a series of foundation models known as Nova, and in December, the focus of the unit expanded when Peter DeSantis, a veteran in cloud computing, succeeded Rohit Prasad as the head of the AGI team.

Earlier this year, Amazon also experienced a significant shift when David Luan, the head of its AGI lab, left the organization after joining in 2024 as part of the acquihire of his startup, Adept.

According to the company spokesperson, Amazon has been invested in creating large AI models for several years, emphasizing that it remains a top priority. DeSantis acknowledged in a recent interview that while Amazon’s models have not always been at the leading edge for high-demand workloads, the organization is committed to enhancing their capabilities and aims to produce one of the most powerful intelligent models available.

Looking ahead, Amazon is set to release its second-quarter financial results next week. The company anticipates capital expenditures to hit $200 billion this year, reflecting a more than 50% hike from 2025, and is raising significant amounts of debt to facilitate its AI development initiatives.

Amazon reduces its workforce in the artificial general intelligence division.

A dark, high-tech lab bathed in shadow, a large Amazon logo glowing dimly on a metal wall. Empty workstations line the room, monitors flickering with “Position Eliminated” notices. In the foreground, a lone silhouetted engineer packs a cardboard box beside an inactive humanoid AI chassis, tangled cables pooling at their feet. Overhead lights flicker, casting long shadows across the deserted AGI unit.

Amazon cuts some jobs in its artificial general intelligence unit - CNBC

- Amazon confirmed layoffs in its artificial general intelligence unit as job cuts continue.  
- Affected AGI employees focused on model customization and post-training, per LinkedIn posts.  
- The company has laid off over 30,000 staff since October while investing in AI.

Amazon is laying off some employees in its unit focused on artificial general intelligence, the company confirmed Wednesday, as it continues to cut jobs while pouring money into AI.

The company declined to disclose how many staffers were affected, or which parts of the AGI organization were exposed to the cuts. The AGI unit is focused on building AI models, and it also includes groups working on silicon development and quantum computing initiatives.

"This is a fast-moving space, and we're sharpening our focus on the initiatives that matter most for customers, so we can move faster on what counts," an Amazon spokesperson told CNBC in a statement. "That focus means some difficult decisions, including eliminating some roles within parts of our AGI organization, even as we continue to invest in the areas most important to our customers' future."

AGI employees focused on model customization and post-training were among those who said they were let go, according to LinkedIn posts.

Reuters first reported the layoffs.

Amazon has been downsizing over the past several years following a pandemic hiring binge, and as it commits massive sums to building out AI infrastructure. The company has laid off more than 30,000 staffers since October, and has continued to eliminate roles through smaller rounds in recent months.

The AGI unit is a core part of Amazon's AI strategy as the company tries to keep pace with leaders OpenAI, Anthropic and Google . AGI generally refers to AI that can perform as well or better than humans on most tasks.

In 2024, Amazon's AGI group released a set of foundation models called Nova. The unit took on a more expansive focus last December when Amazon tapped longtime cloud executive Peter DeSantis to replace Rohit Prasad as the head of the group.

In February, Amazon lost the head of its AGI lab, David Luan, who joined the company in 2024 through a so-called acquihire of his startup Adept.

Amazon's spokesperson said the company has been building large AI models for several years and "it remains one of the most important things we're working on."

DeSantis acknowledged in an interview with CNBC last month that Amazon's models "haven't been at the very frontier for the very largest, most demanding workloads."

He said Amazon has been working to shore up its models further, and it hopes to have one of the "most capable intelligent models out there."

Amazon is scheduled to report second-quarter results next week. The company has forecast capital expenditures for the year of $200 billion, an increase of more than 50% from 2025, and is raising tens of billions of dollars in debt to help fund its AI build-out.

WATCH: Amazon raising at least $25 billion in bond sale

Amazon cuts some jobs in its artificial general intelligence unit

The narrative surrounding an artificial intelligence (AI) competition between the U.S. and China has become widespread, fueled by public figures, the media, and tech giants. This "AI race" rhetoric is granting unchecked authority to major tech firms, potentially resulting in detrimental impacts on our communities, invasive surveillance practices, and the monopolization of human knowledge for corporate profit.

However, the premise of a race between the U.S. and China lacks any substantial foundation. The objectives of the respective AI initiatives in both nations are drastically different; hence, they cannot be accurately described as competing in a race.

In America, there is an emphasis on achieving artificial general intelligence (AGI)—the capability of machines to emulate human awareness. This pursuit is perceived as groundbreaking, suggesting that all AI endeavors must align with this singular goal. Conversely, China’s AI strategy prioritizes the integration of AI into practical applications and products rather than striving for AGI. This divergence indicates that the U.S. and China are not engaged in a shared race toward the same objective.

This distinction highlights the unsustainable growth of extensive data centers in the U.S. These sprawling facilities, which can draw upwards of 5 gigawatts of power—sufficient for approximately 3.75 million homes—are fueled by the drive for AGI. China's strategy, focusing on embedded AI applications, does not necessitate such energy-intensive infrastructure.

The mythic AI race serves as a convenient narrative for Big Tech and its allies, effectively insulating them from scrutiny and regulatory measures. By framing discussions around environmental concerns or community opposition as anti-competitive maneuvering against China, these corporations can evade meaningful debate and sidestep regulations that could hinder their expansion.

Additionally, this narrative is weaponized against grassroots movements opposing the proliferation of data centers. Claims from figures like Kevin O'Leary, alleging foreign funding of these movements, are unfounded and have been publicly retracted by media outlets. The opposition to data centers is a homegrown response from communities who see the damaging ramifications of unregulated tech expansion.

The persistent myth of an AI race pushes us toward a future dominated by Big Tech—one characterized by stark inequality and reduced opportunities for the average person. While tech leaders tout the potential for their innovations to solve major societal problems, the underlying agenda seems focused on consolidating wealth and control.

American tech leaders, despite their acknowledgment of potential job losses due to automation, seem to envision a future where a significant portion of the population is left without meaningful employment, relying instead on meager government assistance. Surveillance technologies are already being used to track our movements, and the advent of sophisticated algorithms allows for exploitative pricing strategies.

Even in the context of a hypothetical AI race, the question arises: at what cost do we seek victory if it leads to a dystopian society with an elite tech oligarchy at the helm?

Notably, this ominous vision is not an unavoidable consequence of technological advancement. Each step taken in the development of AI is a political choice, and we still have opportunities to shape our future collectively. Across the nation, resistance to the unchecked expansion of data centers is gaining momentum. Communities are organizing to halt the construction of new facilities, with notable actions such as New York's yearlong suspension of new center projects.

The future does not have to mimic a reality envisioned by figures like Elon Musk. Our capacity to effect change is evident; mobilizing can yield victories against expansive tech interests. The real challenge lies not in competing with China but in confronting Big Tech. This struggle is indeed critical, and it is one we must actively engage in and win.

The Genuine "AI Competition" Isn't Against China

A sunlit racetrack at dawn, a sleek android sprinter with a glowing circuit-patterned brain dashing past a fluttering Chinese flag at the finish line, its gaze fixed on a distant digital horizon shimmering with streaming data, warm natural light illuminating misty air.

The Real “AI Race” Isn’t With China - Food & Water Watch

- The narrative of an AI race between the U.S. and China is misleading and harmful.  
- U.S. and Chinese AI goals differ significantly, lacking a common point of competition.  
- Big Tech exploits the "AI race" myth to avoid regulation and prioritize profit over communities.

If we allow the myth of an AI race with China to give Big Tech free rein on data centers, we face a more polluted, less equal world.

As Big Tech races to build water-guzzling, energy-hungry data centers for its artificial intelligence (AI), talk of an “AI race” between the United States and China has permeated public discourse. Pundits, politicians, and the media have all joined tech corporations in selling this narrative. And it’s giving license to Big Tech and their political handmaidens to ruin our communities, exploit our every action (both online and via AI-powered surveillance), and steal the wealth of human knowledge for private gain.

But the idea of an AI race between China and the U.S. isn’t grounded in reality. The researchers, companies, and governments behind Chinese and U.S. AI development are pursuing completely different goals.

The discourse in the U.S. assumes that achieving artificial general intelligence (AGI) — computers that mimic human consciousness — would be so momentous and earth-shattering that clearly this must be the goal of anyone pursuing AI development. But that’s not the main goal of Chinese AI development. And a competition in which the competitors are running toward different finish lines isn’t a race.

If we allow the myth of an AI race with China to give Big Tech free rein, we face a more polluted, less equal world.

A Race Doesn’t Have Two Different Finish Lines

While the U.S. is focused on artificial general intelligence (AGI) powered by Large Language Models (LLMs), Chinese developers are focused on AI embedded in products. It’s ChatGPT versus robots.

Yes, China is developing LLMs, although largely in an open-source way as opposed to the for-profit competition in the U.S. Recent news stories report that China is “catching” the U.S. in LLM development. Indeed, the latest Chinese model outperforms leading U.S. models. But this isn’t evidence of an LLM/AGI race. Instead, it shows that without making AGI its main focus, China is capable of developing its own models almost as quickly as U.S. companies.

More to the point, LLM development in China is incidental to the country’s real goal for AI. Its focus remains on products embedded with AI and robots. Or, as AI policy researcher Liang Zheng says, in China, “the first priority is to use it to benefit ordinary people” (debatable, but indicates the kind of AI they are pursuing).

In the U.S., the first priority is to exploit people so that the tech oligarchs can profit. It’s chatbots all the way down.

This isn’t to argue that China is doing it “right” and the U.S. is doing it “wrong.” Either approach will lead to a future in which citizens become increasingly disempowered. In which work becomes more scarce and less lucrative for most people. And in which a handful of billionaires grow wealthier and more powerful.

But the arguments being hauled out to support the destructive growth of hyperscale data centers are based on a fallacy. There is no need to “beat China.” China and the U.S. are racing on separate tracks, in different races, with different finish lines.

These two separate approaches also explain the mind-boggling scale of the data center invasion we currently face. The massive hyperscale data centers — recent proposals would demand up to 5 GW, enough electricity to power roughly 3.75 million U.S. households — are only “required” because the U.S. is racing toward AGI. Meanwhile, the embodied AI dominating in China does not require the same amount of computational power.

The truth is, we don’t need hyperscale data centers to “beat China.” We aren’t racing China. We’re killing ourselves so Silicon Valley can race itself.

How Big Tech Is Weaponizing the “AI Race” to Build Data Centers

Yet, the “AI race” story is a convenient lie for Big Tech and its political protectors. It is a neat political argument designed to insulate the industry from criticism and regulation.

If every environmental review, every question raised by a community, every issue around water usage and electricity prices can be dismissed or lessened as “helping China,” then meaningful political debate can be silenced. Real regulation — if even possible — can be avoided. Fear becomes a substitute for policy.

At the same time, data center developers and their minions in Washington have tried to weaponize false claims about foreign ties to the anti-data center movement. Kevin O’Leary of Shark Tank fame has explicitly said that our movement is being funded by China. He has no evidence because evidence of a falsehood can’t exist. In fact, Fox News has been forced to retract its coverage of his claims.

We understand why O’Leary and the other Tech Broligarchs don’t understand the grassroots opposition to data centers building across the country. It’s hard to spot the grassroots from the window of a private jet. But the opposition is real and organic, and no amount of disinformation and pushing the “AI race against China” scare tactic will derail the movement.

Big Tech Is Selling Us a Dystopian Future

The myth of an AI race with China threatens to propel us into Big Tech’s vision of the future — one that’s more unequal than ever. Yes, tech leaders suggest their algorithms will cure cancer, but their real goal is and has been to increase their power and wealth at the expense of the rest of us.

Even as American Tech Broligarchs have distanced themselves from earlier prophecies of widespread job loss, their vision of the future will see the vast majority of us out of meaningful work. We’ll be subject to living off whatever meager handouts are created in an attempt to mollify us.

At the same time, our movement will be traced whether or not we use the electronic gadgets they sell us. Already, surveillance devices linked to AI can recognize our faces, record our license plates, and report our movements. Companies and governments can buy this data in order to track us.

Meanwhile, “surveillance pricing” allows companies to change prices in an instant so that they can exploit our needs for their profit. Deepfake videos have already added to the rapidly decaying trust in a commonly shared world and a basic set of facts necessary for a functioning democracy. And this is but a scratch of the surface.

Even if there were an AI race, is it worth running, let alone winning, if the prize is a dystopian future of mass misery with a thin layer of super wealthy tech oligarchs at the top?

Big Tech’s Dystopian Vision Isn’t Inevitable

But we should be clear. This dystopian vision is not the inevitable outcome of unstoppable technological “progress” as the Tech Broligarchs would have us believe. Each and every decision being made to advance AI is a political decision. And, for now, we still have the ability to determine our political future.

Across the country, there is a growing resistance to the nightmare being shoved down our throats. Communities are coming together to fight the spread of destructive hyperscale data centers. Already in 2026, more than $130 billion-worth of proposed data center projects have been defeated and canceled.

Communities are taking control of their futures by placing moratoriums on new data centers. New York enacted a one-year pause on new centers, and there is growing support for a nationwide pause in Congress.

We aren’t destined to live in Elon Musk’s fever dream. We have the power to stop him and his fellow Broligarchs. When we organize, we win. That’s the real “AI Race.” Not China versus the U.S., but us versus Big Tech. That’s not only a race worth running — it’s one we have to win.

The Real "AI Race" Isn't With China

Monitor your investments without any cost using Simply Wall St, a portfolio management platform that has gained the trust of over 7 million investors worldwide.

Microsoft (MSFT) is garnering attention again after it announced participation in the AGI Summit SF 2026, scheduled for July 18. This prestigious event focused on artificial general intelligence could shape investor perceptions regarding the company's AI initiatives.

Recent analysis of Microsoft highlights that the stock's value has been on the rise, showcasing a one-day increase of 2.15% and a 30-day gain of 6.03%. Yet, the year-to-date performance reflects a decline of 14.94%, with the total shareholder return over the past year dropping by 20.50%. On a more positive note, long-term investors have fared better, with total shareholder returns of 19.33% and 45.02% over the three- and five-year periods, respectively.

If Microsoft's focus on AGI has you contemplating investments beyond individual stocks, now might be an opportune time to explore the broader landscape of AI infrastructure and its associated stocks, which number 54.

Following a significant drop and subsequent rebound, Microsoft’s stock is now trading at a notable difference from both analysts’ targets and intrinsic value estimations. This raises the question: where should fair value realistically be positioned within this disparity, and what is the current cost of investment?

Current Analysis: 13.7% Undervalued

With a recent closing price of $402.29 compared to a perceived fair value of $466, Microsoft appears to be considerably undervalued. The reasoning presented highlights factors such as cash generation, capital allocation, and the impact of regulatory challenges.

Previously, I estimated Microsoft's value to be between $393 and $400 per share through a cautious discounted cash flow analysis. However, developments from Q3, alongside clarified capital expenditures and adjustments regarding its competitive advantages, have led to an updated fair value assessment of $466 per share in the base scenario. The optimistic scenario suggests a value of $804 per share, while the pessimistic outlook drops to $229 per share.

For a comprehensive understanding of the distinctions among these bear, base, and bull cases for Microsoft, the narrative hinges on specific expectations regarding revenue, margins, and cash flow projections that may not be immediately clear from the most visible figures.

Conclusion: Fair Value Set at $466 (UNDERVALUED)

To fully comprehend the forecasts and valuations mentioned, it's worthwhile to delve into the complete analysis.

Nevertheless, this assessment of Microsoft’s potential could face challenges if regulatory frameworks around cloud and software licensing become stricter, or if investments in AI infrastructure do not yield sufficient returns.

To understand the risks associated with this Microsoft outlook, make sure to read about the crucial factors that could impact its situation.

Next Steps

Given the potential opportunities and concerns surrounding Microsoft, it would be prudent to examine the relevant data yourself. Take the time to consider the detailed breakdown of five key advantages and one significant cautionary signal.

Is Microsoft (MSFT) Still Priced Too Low as AGI Summit Highlights Valuation Again?

A sunlit modern conference hall with skylights, a large “AGI Summit” banner hanging behind a podium topped by a bold Microsoft logo. In the foreground, a polished golden balance scale holds a small glowing Microsoft icon on one side and a hefty plaque labeled “Valuation” on the other, the scale tipping as attendees in business attire look on with intrigued expressions.

Is Microsoft (MSFT) Still Undervalued As AGI Summit Focus Puts Valuation Back In View? - Yahoo Finance Australia

- Microsoft will present at AGI Summit SF 2026, highlighting its AI ambitions.  
- Recent Microsoft share price rebound: 2.15% in one day, down 14.94% year-to-date.  
- Analysts suggest Microsoft is undervalued at $402.29, with a fair value of $466.

Track your investments for FREE with Simply Wall St, the portfolio command center trusted by over 7 million individual investors worldwide.

Microsoft (MSFT) is back in focus after confirming it will present at AGI Summit SF 2026 on July 18, a high profile forum on artificial general intelligence that could influence how investors frame its AI ambitions.

See our latest analysis for Microsoft.

That AGI Summit appearance comes as Microsoft's share price has recently rebounded, with a 1-day share price return of 2.15% and 30-day share price return of 6.03%, although the year-to-date share price return is down 14.94% and the 1-year total shareholder return has declined 20.50%. However, the 3-year and 5-year total shareholder returns of 19.33% and 45.02% indicate that longer term holders have still seen gains.

If this AGI focus has you thinking beyond a single stock, it could be a good moment to scan the broader opportunity set in AI infrastructure and related enablers through the 54 AI infrastructure stocks

After a sharp pullback and a recent bounce, Microsoft now trades at a wide gap to both analyst targets and intrinsic value estimates. Where might fair value actually sit along that spread, and what are you really paying for today?

Most Popular Narrative: 13.7% Undervalued

At a last close of $402.29 versus a narrative fair value of $466, Microsoft is presented as materially undervalued, and the author's reasoning focuses on cash generation, capital allocation and regulatory pressure.

In March I valued Microsoft at roughly $393 to $400 per share with a conservative discounted cash flow analysis. The Q3 print, the CapEx clarity, and the moat adjustments produce a different number now. Base case fair value: $466 per share. Bull case: $804 per share. Bear case: $229 per share.

Read the complete narrative.

Want to see what sits behind that spread between bear, base and bull cases for Microsoft? The narrative relies on specific revenue, margin and cash flow paths that are not obvious from the headline numbers.

Result: Fair Value of $466 (UNDERVALUED)

Have a read of the narrative in full and understand what's behind the forecasts.

However, this Microsoft narrative could be knocked off course if regulators tighten remedies around cloud and software licensing, or if AI infrastructure spending fails to earn adequate returns.

Find out about the key risks to this Microsoft narrative.

Next Steps

Seeing both the opportunities and the concerns around Microsoft in this article, it makes sense to review the underlying data yourself and decide where you stand with the full breakdown of 5 key rewards and 1 important warning sign.

Is Microsoft (MSFT) Still Undervalued As AGI Summit Focus Puts Valuation Back In View?

After a seven-year hiatus, ZHU Songchun made a notable return to the World Artificial Intelligence Conference (WAIC) on July 19, 2026. Previously, he had not participated since 2019, marking his absence as a significant statement against the prevailing trends in the AI field. 

In recent years, the spotlight at WAIC has been on large language models, scaling laws, and an optimism that equates massiveness with breakthrough advancements. ZHU, the President of the Beijing Institute for General Artificial Intelligence (BIGAI), has stood out as a prominent critic of this prevailing mindset. Following the surge in interest around large models ignited by the launch of ChatGPT in late 2022, ZHU has consistently asserted that these models do not inherently pave the way to Artificial General Intelligence (AGI). He argues that merely increasing the scale of parameters, data, and computation is insufficient, emphasizing that the financial motivations and international narratives driving this expansion often overshadow authentic technological advancements.

In his recent address during the "Thinkers Forum," ZHU characterized the AI boom as a narrative fueled by capital and geopolitical interests, linked to what he refers to as the "Silicon Valley-Wall Street-Washington triad." He pointed out that the Chinese tech sector has become ensnared in a cycle of reverse informational flows from the U.S., where information is fed back to China and restated by local experts, leading to a repetitive cycle.

ZHU highlighted what he has termed the "Musk Faith," referring to investors' unwavering belief in the ambitious claims of tech leaders, despite a track record of unmet promises. He noted instances where projects championed by Musk, such as the Hyperloop and full self-driving vehicles, have faced persistent delays or outright cancellations, with market reactions often fervently exaggerated in response to these grand pledges.

The tech market worldwide has recently experienced heightened volatility, particularly since June 2026, as major U.S. tech firms continue to escalate their investment in AI while facing diminishing free cash flow. In his address, ZHU reflected on how these market trends may lead to greater receptivity to his critiques as signs of a potential bubble begin to appear.

Notably, the consensus in the tech industry still leans towards larger models, with key players like OpenAI, Google DeepMind, and Anthropic continuing to advocate for their development. They assert that there remains significant potential in refining these architectures and enhancing reasoning capabilities through multimodal integration. Many leaders within China's AI sector believe that while large models may not represent the culmination of AGI, they currently serve as the most effective tools for commercialization.

Throughout the last three years, ZHU's core perspectives have remained steadfast. He has likened the journey towards AGI to a trek up Mount Everest versus a lunar landing, emphasizing the complex differences between achieving large model capabilities and true general intelligence. In his 2025 publication, *Giving Minds to Machines*, he described large models as "brains in vats"—able to produce language but lacking a true understanding of the world, as there exists no genuine connection between their generated words and the real environment.

By 2026, ZHU assessed the field's many booms and busts—covering various eras from computer vision advancements to the metaverse hype—and concluded that the industry's faith in computational power and data-driven models is being systematically challenged by real-world results. 

In contrast to mainstream AI trends heavily reliant on transformer technologies, ZHU has initiated research centered around causal reasoning and intrinsic value-driven methodologies, aiming to develop autonomous general agents capable of task generation, learning, and action. His approach has led to the creation of a framework known as CUV, where C represents cognitive architecture, U encompasses capabilities related to perception, cognition, and action, while V symbolizes intrinsic motivations and value systems.

His institution has also rolled out the TongOS general artificial intelligence operating system, the TongPL programming language, and a unique agent named "Tongtong." Unlike typical large model offerings, Tongtong operates on intrinsic values, facilitating an understanding of the environment, generating tasks, and planning actions autonomously. Preliminary evaluations suggest some of its cognitive abilities are akin to those found in five- to six-year-old children.

In March 2026, the latest version of Tongtong was unveiled, showcasing advancements in spatial and cognitive intelligence, as well as social intelligence, alongside the introduction of "Tong Brain," an embodied intelligence engine intended to integrate general agents with physical robotics. 

ZHU's team has gained notable attention from platforms like *Science*, detailing their vision for transforming isolated AGI agents into cohesive AGI societies that focus on multi-agent interactions and the development of artificial civilizations. At WAIC, ZHU emphasized the importance of "social intelligence" as the final hurdle toward reaching AGI, underlining the aspect of comprehending complex social dynamics—responsibilities, intentions, and collaborative interactions—in which current large models still fall short.

As the conversation around ZHU’s framework unfolds, the pressing question remains: before Tongtong's commercial viability is established, does his approach reflect a more enlightened technical direction or an ambitious academic hypothesis awaiting validation?

Engaged in thoughtful dialogue at WAIC, ZHU expanded upon his critical views regarding the potential of large models in achieving AGI, addressing the existence of a bubble within the AI sector, assessing China’s innovative initiatives, and discussing how cognitive architectures can transition to practical applications in the industry.

ZHU's stance is clear: while the vibrant atmosphere of innovation at WAIC is commendable, a more profound inquiry into the current technical trajectory is necessary. Despite advancements in robotic technologies, the question persists—will the current methodologies yield the comprehensive general intelligence that stakeholders envision?

Zhu Songchun: China's AI Sector Should Steer Clear of "Musk Belief" Misconceptions

A dimly lit high-tech lab tinted in deep reds and blacks: a stern Chinese AI researcher in business attire stands before a wall-mounted Chinese flag, turning his back on a looming shadow silhouette of Elon Musk’s recognizable hairline labeled “Musk belief,” while he clutches a glowing tablet titled “AI Roadmap,” lit by a single warm overhead lamp.

Zhu Songchun: China's AI Industry Should Not Be Misled by the "Musk Belief" - 36 Kr

- ZHU Songchun returns to WAIC after seven years, criticizing the focus on large models.  
- He argues that scaling doesn't guarantee AGI, emphasizing intrinsic value and cognitive architecture.  
- ZHU introduces "Tongtong," an autonomous agent, highlighting social intelligence as vital for AGI.  

The last time ZHU Songchun appeared at the World Artificial Intelligence Conference (WAIC) was back in 2019. For the following seven years, he was almost entirely absent from China's largest AI industry showcase.

That absence itself was a statement of stance.

Over the past few years, WAIC's center stage has belonged to large language models, the Scaling Law, and the belief in "miracles born of sheer scale." ZHU Songchun, President of the Beijing Institute for General Artificial Intelligence (BIGAI), has been one of the most consistent and persistent critics of this dominant narrative.

Ever since ChatGPT ignited the large model wave at the end of 2022, he has repeatedly emphasized that large language models cannot naturally lead to Artificial General Intelligence (AGI), and that simply scaling up parameters, data, and computing power will not automatically bring about AGI emergence. In his view, the capital-driven marketing and international geopolitical narrative behind this boom are far more powerful than genuine technological breakthroughs.

On July 19, 2026, ZHU Songchun made his return to WAIC.

During his one-hour speech at the "Thinkers Forum," he summarized the past several years of the AI boom as a capital and geopolitical narrative jointly promoted by the "Silicon Valley-Wall Street-Washington trio." China's technology community has long been trapped in problems defined by the United States: "Information flows from the U.S. to us, then we invite American experts to restate it, creating a cycle of reverse information import and amplification."

He even singled out a phenomenon he called the "Musk Faith": investors' blind adherence to grand tech leader narratives, noting that "80% of what Musk promises never gets delivered." Projects like the Hyperloop have been shut down, full self-driving has been delayed more than a dozen times, and brain-computer interfaces have fallen far short of expectations — yet the grander the promises and the slower the delivery, the more fanatical the market reaction becomes.

From June 2026 onward, global tech stocks have experienced dramatic volatility. U.S. tech giants keep expanding their AI capital expenditures, putting free cash flow under pressure; SpaceX's stock fell below its IPO price after going public; Meta has also begun exploring commercializing its surplus computing power externally. ZHU Songchun's judgments are partially coming to pass. "When the bubble is stretched to its limit and shows signs of starting to burst, people might be more willing to listen to what I have to say." That is part of the reason ZHU Songchun returned to WAIC this year.

Yet the mainstream consensus in the industry remains on a different track. OpenAI, Google DeepMind, and Anthropic are still doubling down on large models, arguing that there is far more untapped potential in architectural improvements, reasoning enhancements, and multimodal fusion. Leading domestic AI companies generally hold that large models are not the end point of AGI, but they are currently the most efficient technical tool and commercialization path — and the ecological advantages already established should not be easily abandoned. Both sides have their own evidence, and their own vested interests.

ZHU Songchun's core judgments have never changed over the past three years. In early 2024, he used the metaphor of "climbing Mount Everest vs. landing on the moon" to describe the distance between large models and general artificial intelligence. In his 2025 book *Giving Minds to Machines*, he characterized large models as "brains in vats": they can speak, but they do not understand the real world, lacking a genuine connection between "words" and "reality." By 2026, he reviewed four successive boom-and-bust cycles: the AI "Four Little Dragons" era in computer vision, the metaverse craze, the "hundred-model battle," and embodied intelligence. His conclusion is that the faith in computing power, the faith in big data, and the faith in large models are all being dismantled one by one by real-world outcomes.

Technically, ZHU Songchun has taken a path distinct from the Transformer paradigm: centered on causal reasoning and intrinsic value-driven mechanisms, to build general agents that can autonomously generate tasks, learn, and act.

Along this path, his team has proposed the CUV general intelligence framework. Here, C represents the agent's cognitive architecture, U covers perception, cognition, and action capabilities, and V stands for the intrinsic motivation and value system. ZHU Songchun has also divided AGI innovation into five layers: philosophy, mathematical and physical frameworks, models, algorithms, and implementation. In his view, most current industrial innovations are still concentrated on the algorithm and implementation layers, and truly original breakthroughs require tracing back further to the philosophical and fundamental theory layers.

Built around this system, the Beijing Institute for General Artificial Intelligence has developed the TongOS general artificial intelligence operating system and the TongPL programming language, and launched the general intelligent agent "Tongtong." Unlike large model products that are primarily driven by external instructions, "Tongtong" is designed as a prototype of a general agent driven by an intrinsic value system, capable of autonomously understanding the environment, generating tasks, and planning actions. According to the team's evaluation standards, some of its cognitive and behavioral capabilities have reached the level of a five- to six-year-old child.

In March 2026, "Tongtong" 3.0 made its debut at the Zhongguancun Forum, with upgrades in spatial intelligence, cognitive intelligence, and social intelligence. The concurrently released "Tong Brain" was positioned as the core embodied intelligence engine connecting general agents to physical robots.

In 2025, *Science* consecutively reported on ZHU Songchun's team's research around the general intelligent agent "Tongtong" and their roadmap from individual AGI agents to AGI societies, focusing on their efforts to extend general intelligence from individual cognition to multi-agent interaction, social simulation, and artificial civilization research.

At this WAIC appearance, ZHU Songchun formally brought "social intelligence" to the forefront: "understanding responsibilities, rights, and interests, inferring intentions, reading subtext, and participating in complex social collaboration." He argues this is the final barrier to AGI, and the dimension where all current large models are still handing in blank papers.

For the industry, however, the more practical question remains: before "Tongtong" completes a viable commercialization loop, does the path ZHU Songchun has chosen represent a more forward-looking technical judgment, or an academic bet yet to be proven?

During WAIC, Q had an in-depth conversation with ZHU Songchun. He further elaborated on his views around topics such as whether large models can lead to AGI, whether bubbles exist in the current AI boom, whether China truly lacks original technical paths, and how cognitive architectures can move toward industrial implementation.

01 Scorching Weather, Hottest Large Models, and Industrial Bubbles

Q: This year's WAIC is still buzzing with activity, with notable new progress in areas like robots and intelligent agents. What's your overall impression of the current AI industry?

ZHU Songchun: There are huge crowds and a lively atmosphere, and robotic technology has indeed made noticeable progress compared to previous years. Yet a fundamental question still hangs over the entire industry: can the technical path we are currently taking ultimately deliver the general artificial intelligence everyone talks about?

Back when the large model boom first started from late 2022 to early 2023, I pointed out that large language models cannot naturally lead to AGI. Over three years later, large models have improved in language generation, dialogue, and tool use, but there remains an essential gap between them and true general intelligence. The most prominent problem right now is equating incremental improvements in partial capabilities with the imminent arrival of AGI, which in turn creates overinflated technical expectations and capital valuations.

Q: Is the question of whether large models can lead to AGI a separate matter from whether large models can generate industrial value?

ZHU Songchun: Absolutely two different things. As a tool, large models can play a role in scenarios like chat, content generation, and code assistance. But having practical application value does not make a system AGI, nor does it prove that continuing to scale parameters, data, and computing power along the same path will necessarily achieve AGI.

Scenarios that have truly formed stable commercial closed loops are still limited. The "Agent" concept is very hot, and there are many products, but moving from demo to real production environments requires solving a whole series of issues around accuracy, long-term planning, result verification, cost, and accountability boundaries. A model that can chat and call a few tools is still very far from a general agent that can autonomously understand the world, set goals, and complete open-ended tasks.

Q: Why do you frame this round of AI boom within the context of international politics and capital narratives?

ZHU Songchun: Any major technological wave is never just a technical line — behind it lies a narrative co-constructed by politics, capital, and communication. Around 2014 to 2016, when DeepMind was acquired by Google, OpenAI was founded, and AlphaGo entered the public consciousness, the globalist narrative was also shifting toward the "America First" paradigm. Artificial intelligence, and AGI in particular, gradually became a key strategic anchor for the U.S. to reshape its technological competitive advantage.

This narrative combines policy needs, Silicon Valley companies' fundraising needs, and Wall Street's capital needs, drawing global capital, talent, and infrastructure investment to concentrate in the U.S. Even if the bubble bursts in the future, the U.S. will already have secured tangible benefits such as data centers, electricity, chips, talent, and tax revenues. So to understand AI, we cannot only look at model leaderboards — we also need to see who is defining the problems, setting the standards, and organizing the resources.

Q: Are chips and computing power the most critical bottleneck for China's AI development?

ZHU Songchun: Chips are certainly important, and China must build self-reliant and controllable software and hardware foundations. But we cannot reduce every problem to "not having enough GPUs." Computing power only generates value when combined with correct technical architectures, clear tasks, and real-world demands. Massive chip purchases and intelligent computing center construction can still lead to resource waste if there is a lack of effective utilization and industrial closed loops.

The Scaling Law portrays the continuous expansion of data, parameters, and computing power as a sufficient condition for reaching AGI, but I believe this logic has fundamental flaws. It's like a person who never builds a knowledge structure or thinking ability, but only hoards piles of exam papers — they will never develop true general competence. AI's real bottlenecks also include cognitive architectures and value systems.

Q: Do you think the AI industry has already formed a bubble? What is the essence of this bubble?

ZHU Songchun: I believe there is an obvious bubble right now. The essence of the bubble is prematurely and excessively discounting the future value of an immature technology into the present. Capital markets rely on grand narratives to sustain expectations, companies keep switching buzzwords to secure financing and valuations, while real technological maturity, user demand, and business revenues fail to keep pace.

From large models and Agents to embodied intelligence and world models, hot topics keep rotating — there is genuine progress, but also a huge amount of labeling and homogenization. Judging bubbles should not only look at stock price movements, but also check whether input and output match, whether the technology solves real problems, and whether the business model is sustainable. When training and operating costs keep rising, models need rapid iteration, but revenues cannot cover investments, this structure is extremely fragile.

Q: How do you expect this bubble to adjust? Can bubbles also have positive effects?

ZHU Songchun: Bubbles are not entirely without positive effects. High-density capital investment accelerates infrastructure construction, attracts talent into the field, and may drive breakthroughs in some key technologies. After the dot-com bubble burst, companies with real core technologies and viable business models survived — the AI industry may go through a similar process.

The problem is that the costs of bubbles are equally massive, including capital losses, talent misallocation, distorted public expectations, and damage to the long-term scientific research ecosystem. A huge number of teams chase short-term fads, investment institutions only back projects that can be quickly packaged and exited, while truly original research requiring 5 or 10 years of accumulation struggles to get support. Only when the bubble recedes can the industry have the opportunity to refocus on the essence of technology and real-world scenarios.

Figure: ZHU Songchun speaking at the Thinkers Forum

02 Stop Misusing AGI as a Marketing Buzzword

Q: There is no unified definition of AGI in the industry right now. How do you define true general artificial intelligence?

ZHU Songchun: AGI is not a marketing term that can be arbitrarily interpreted. Strictly speaking, general artificial intelligence should possess physical common sense and social cognition, be capable of autonomously generating tasks and completing infinite open-world tasks, and its behaviors should be driven by intrinsic values aligned with human values and social norms.

I believe there are at least three fundamental characteristics: first, infinite task generalization, breaking through predefined tasks and closed scenarios; second, autonomous task generation, being able to form goals, plan actions, and adjust based on feedback; third, value-driven, knowing why to act, what should be done, and what should not be done. A system that only answers questions based on external instructions, or performs pattern matching within existing data distributions, cannot be called true AGI.

Q: Why do you think large language models cannot cover all the capabilities required by AGI?

ZHU Songchun: Large language models primarily process statistical correlations in language sequences. Language ability is important, but AGI also encompasses many domains including vision, robotics, cognitive reasoning, multi-agent collaboration, social intelligence, and social minds. A system that can fluently generate text does not truly understand the physical world, causal relationships, other people's intentions, or social norms.

The current mainstream path tends to mistake external performance for internal intelligence: when a model scores high on certain exams, or completes a sequence of tool calls, people interpret it as "emergent" general intelligence. But without stable world representations, causal mechanisms, goal generation, and value judgment capabilities, these performances can hardly generalize consistently and reliably in open environments.

Q: Why do you consider "cognitive architecture" and "value" more important than model parameters?

ZHU Songchun: Architecture determines what kinds of capabilities an agent can develop. Different biological species have different cognitive architectures — no matter how much data and training you give them, they will never naturally gain capabilities beyond the boundaries of their architecture. Today's many large models focus heavily on parameters, data, and computing power, yet lack stable world models, causal mechanisms, goal systems, and value structures.

We proposed the CUV framework to uniformly describe agents through the cognitive architecture C, the capability or potential function U, and the value function V. The cognitive architecture organizes information and makes decisions; the capability system supports perception, language, movement, and learning; the value system drives goal selection and behavioral direction. True general intelligence requires the three to co-evolve. The idea of "giving minds to machines" means endowing agents with internal goals, value judgments, and structures for understanding others.

Q: You divided AGI innovation into five layers: philosophy, mathematical and physical frameworks, models, algorithms, and implementation. Why start from the philosophical layer?

ZHU Songchun: Most current industrial competition is concentrated on the model, algorithm, and implementation layers — for example, parameter scaling, training tricks, inference speed, and engineering deployment. These efforts are certainly important, but if the underlying assumptions about "what intelligence is" never change, all innovations are just optimizations within the existing paradigm.

The philosophical layer determines how we understand the subject, goals, and values of intelligence; the mathematical and physical framework layer translates these fundamental understandings into computable, verifiable forms; only then come models, algorithms, and engineering implementation. Truly original breakthroughs must be able to redefine problems, not just run faster on someone else's track.

Q: How can we prove that an original architecture is closer to AGI than the Transformer path?

ZHU Songchun: Originality cannot rely solely on conceptual claims — it must ultimately be tested through a set of open, systematic standards. AGI testing should not only focus on language Q&A or single benchmarks, but cover capabilities such as perception, reasoning, causal understanding,

Think Tank Cautions That Global AI Industry Lacks Adequate Safety Measures

More from Artificial General Intelligence

Amazon Resizes Its Artificial General Intelligence Team, Viewing It as a Reset Rather Than a Retreat.

Amazon reveals job cuts in its artificial general intelligence division

Google is stockpiling TPUs to create Artificial General Intelligence.

Google is stockpiling TPUs in pursuit of artificial general intelligence.