OpenAI's launch of its next-generation Astra model has drawn widespread attention to a potential leap in AI capability and rekindled the investment narrative around the AI spending cycle.

Rich Privorotsky, head of Goldman Sachs' Delta One trading desk, noted in a latest market briefing that Astra looks "significant" — it has genuinely pulled ahead of its peers, and this might be precisely the kind of capability breakthrough that AI bulls have been waiting for.

That assessment quickly translated into market action: SoftBank jumped more than 10% overnight, with its stock widely viewed as a key proxy for OpenAI's valuation shifts as a major investor; Oracle also rose about 5.5%. But in Privorotsky's view, the share price gains are actually the "least interesting part" of the whole story.

What deserves more attention is the equally aggressive internal assessment of Astra at OpenAI. President Greg Brockman called it a "generational leap" and suggested this moment could come to be seen as a landmark milestone in the arrival of AGI (artificial general intelligence). At the end of the conference call, he went even further, declaring: "Welcome to the AGI era."

From Wall Street to OpenAI's inner circle, Astra is being framed as more than just a model upgrade — it is a signal that a leap in AI capability could reshape a new round of capital expenditure. This expectation has also driven OpenAI's ecosystem to clearly outperform the Anthropic camp this week overall.

Capability Breakthrough, Not Price Cuts: Reshaping the Demand Curve Once Again

In his briefing, Privorotsky drew a sharp distinction between two very different market signals: falling token prices versus a substantive leap in model capability.

His core logic is that when model prices decline, companies simply use the same amount of AI at a lower cost; but when models become "genuinely smarter," businesses discover new use cases that were previously impossible, and the demand space for AI expands accordingly.

"A price cut is one thing; a true breakthrough in intelligence changes the demand curve again," Privorotsky wrote. "It forces every other lab to play catch-up, keeps the spending cycle vibrant, and reignites the belief that there's still something better worth building."

He also clarified that earlier remarks about "token demand growth not keeping pace with the rate of token cost decline" had been taken out of context by some. His full reasoning is that oversupply of compute only becomes a problem if model capabilities fail to achieve a fundamental leap — and the arrival of Astra points firmly in the opposite direction.

Astra Targets AGI: AI Competition Shifts from Price War to "Intelligence War"

Astra's performance on certain AGI-related benchmarks has given OpenAI the confidence to claim it has surpassed Anthropic on some metrics.

While the industry lacks a consensus definition of AGI and no unified testing standards exist, and Brockman himself said the final judgment should rest with users, his remarks during the media call were already quite explicit. When asked whether Astra meets the definition of AGI, he replied: "I think this is the model."

That statement has further reinforced the market's interpretation of this release: competition among leading model makers is shifting from price and efficiency to a head-on contest of intelligence capability. For AI bulls, this is the long-awaited signal — if a leadership edge can be demonstrated in the intelligence of the models themselves, the durability of the AI capital expenditure cycle gains new support.

September AI Calendar: Dense Catalysts, Bullish Sentiment May Continue to Brew

Privorotsky pointed out that September's market calendar looks constructive for the AI sector, with multiple potential catalysts offering a continued channel for bullish sentiment.

Oracle is set to report earnings next week, with the market likely to focus on updates regarding its cloud infrastructure and AI-related businesses. Meanwhile, the tech industry conference season will provide additional catalysts, helping the positive impact from Astra spread further across the broader AI supply chain.

"The return of breakthrough intelligence is exactly what's needed to rekindle both the willingness and the necessity for sustained spending," Privorotsky summarized. However, he also cautioned that energy price movements and next week's CPI data remain the main potential disruptors to current market sentiment and deserve close attention.