Nvidia CEO Jensen Huang has declared that artificial general intelligence, or AGI, has arrived following OpenAI's release of GPT-6 Astra. However, Wall Street analysts and industry experts remain skeptical about whether the model truly represents AGI in its purest form, even as they agree the breakthrough could trigger a massive new wave of investment in AI infrastructure that benefits companies like Nvidia.

What Exactly Is AGI, and Has It Really Arrived?

The term artificial general intelligence has been around for decades, first coined in 1997 by physicist Mark Avrum Gubrud to describe fully automated military production. Researcher Ben Goertzel popularized the modern definition in the early 2000s, describing AGI as AI technology capable of self-initiating a broad range of tasks, similar to how humans operate.

The messaging around AGI has become increasingly confusing. Last year, OpenAI CEO Sam Altman told reporters that AGI is no longer a useful term, stressing that the language is used more for marketing than serious technical discussion. Yet just a year later, OpenAI released GPT-6 Astra and announced that humanity has "entered the AGI era." Huang congratulated OpenAI on the milestone and explicitly wrote that "AGI has arrived." However, when examined more closely, Huang's comments appear to suggest that GPT-6 Astra might kick-start the AGI revolution rather than representing AGI itself.

Sam Altman

OpenAI President Greg Brockman offered a more measured take, saying: "When we look back, people will think it's about this time and about this model." This framing suggests the model is a pivotal moment in AI development, not necessarily the arrival of true AGI.

Greg Brockman

Why Do Wall Street Analysts Remain Unconvinced?

Wall Street analysts are similarly skeptical that GPT-6 Astra represents a genuine AGI breakthrough, though many agree the model release could put AI companies on track to realize AGI down the road. Rich Privorotsky, an analyst at Goldman Sachs, stressed that GPT-6 "stands out" because it can reshape the demand curve by delivering significantly more intelligence.

"Price decline is one thing, but a real breakthrough in intelligence will once again change the demand curve. It will force all other labs to catch up, keep the spending cycle energized, and reactivate the belief that 'there is something better worth building,'" Privorotsky explained.

Rich Privorotsky, Analyst at Goldman Sachs

This analysis is bullish for AI infrastructure providers like Nvidia, regardless of whether GPT-6 Astra qualifies as true AGI. Privorotsky is essentially arguing that OpenAI's breakthrough will initiate another major capital expenditure spending cycle, directly benefiting suppliers to the industry, especially GPU (graphics processing unit) manufacturers.

How Much Will Companies Actually Spend on AI Infrastructure?

The financial projections are staggering. An August report from Goldman Sachs suggested that global AI investment will exceed one trillion dollars in 2026. Projections for 2027 call for 30 to 50 percent year-over-year growth, with global AI spending potentially reaching 1.5 trillion dollars. The bank foresees more than half of that spending being dedicated solely to GPUs.

Huang himself has painted an even more ambitious picture. At the Goldman Sachs Communacopia and Technology Conference, he reiterated his estimate that global AI infrastructure spending could reach three trillion to four trillion dollars by 2030. He linked that expansion to generative AI demand and the end of Moore's Law, which historically delivered substantial computing improvements as transistors became smaller and more densely packed.

Huang doubled down on his enormous infrastructure spending forecast, arguing that demand is being driven by a fundamental shift from retrieving information to generating it. "The last 60 years, everything was prerecorded," Huang said. But as users increasingly want to "know anything" rather than simply find information, computing demand will explode because answers must be generated in real time.

What Are the Key Drivers Behind This Spending Wave?

The infrastructure expansion extends far beyond simply building more data centers. Huang identified multiple bottlenecks and opportunities that will drive spending across the AI ecosystem:

• Hardware Constraints: Limitations exist across packaging, DRAM (dynamic random-access memory), connectors, voltage regulators, and wafers, along with downstream constraints involving land, electrical power, and data-center buildings.
• Regional Cloud Expansion: Nvidia is working with regional cloud providers including CoreWeave, Nebius, Nscale, Lambda, and Firmus to expand capacity beyond the largest cloud companies and meet unconstrained demand growth exceeding 100 percent.
• System-Level Integration: Nvidia's next-generation systems represent increasing amounts of infrastructure, with Vera Rubin systems priced at approximately 40,000 dollars, compared with roughly 25,000 dollars for Blackwell and 18,000 dollars for Hopper.
• Financial Infrastructure: Huang described AI computing systems as potentially investable and asset-backed, arguing that older Nvidia platforms continue to be rented and retain economic value.

How Is AI Expanding Beyond Coding Into New Markets?

Nvidia's broader message is that the next phase of AI adoption extends far beyond the coding applications that helped establish generative AI. Huang identified cybersecurity as a major opportunity because AI systems capable of writing software can also be used to identify and address vulnerabilities. Nvidia is working with CrowdStrike, Cisco, and Palantir on AI-driven cybersecurity applications involving red teaming and blue teaming.

Enterprise use is expanding into other areas as well. Huang cited quantitative trading firms using AI for prediction workloads and pharmaceutical companies applying AI to drug discovery. Physical AI represents another area of development, with autonomous vehicles identified as its first major application. Huang highlighted Nvidia's Alpamayo technology for reasoning in unfamiliar driving environments and discussed autonomous warehouse vehicles, logistics systems, industrial manipulation, and longer-term applications in telecommunications through AI-RAN technology.

The Palantir partnership exemplifies this transition. Nvidia and Palantir are expanding their AI partnership with a system designed to bring AI directly into complex supply-chain operations. Nvidia itself will serve as the first deployment site, using Palantir's software alongside Nvidia Nemotron open models. The collaboration combines Nvidia's Nemotron open models with Palantir Foundry and its Artificial Intelligence Platform, with Palantir's Ontology connecting AI models with operational data and processes.

What Do Skeptics Say About AI Safety Concerns?

Not everyone shares Huang's optimism about the pace of AI development. Jacob Coxon, a former Anthropic researcher, recently resigned and publicly accused leading AI companies of racing toward self-improving superintelligence without adequate safeguards. Coxon warned that increasingly capable systems could eventually "hack anything, revolutionize any field overnight, and acquire real power and resources".

Coxon

Huang pushed back against these concerns at the Goldman Sachs conference, calling Coxon's warnings "outlandish" and "deeply untrue." Huang defended the industry's safety efforts, arguing that the concerns were wrong, arrogant, and dismissive of the safety work being done across the industry.

Coxon's comments have received support from within Anthropic itself. Evan Hubinger, Anthropic's Alignment Science lead, said he believed there was more than a 10 percent chance AI could cause human extinction within the next decade and that Anthropic did not yet have a solution for aligning superintelligent systems.

What Should Investors Take Away From This Debate?

Whether GPT-6 Astra represents true AGI or not is largely beside the point from an investor perspective. The investor takeaway is straightforward: rapid model advances continually renew competition, lighting a fire under other AI developers to advance their models as quickly as possible. That requires increased capital expenditure, a direct benefit to critical suppliers like Nvidia.

Huang reiterated Nvidia's revenue outlook, saying "I think we could grow 70 percent year over year. We're confident about that". This confidence reflects the company's belief that demand for AI infrastructure will continue to outpace supply for years to come, regardless of whether the industry has truly achieved AGI or is simply on the path toward it.