Yazılar

AI Investment Gains Supercharge S&P 500 Second-Quarter Earnings

/in /tarafından

The S&P 500 is closing out an unusually strong second-quarter earnings season, with artificial intelligence investments providing a major boost to corporate profits.

Aggregate earnings for the index are on track to rise about 52% year over year, while technology sector profits are expected to jump roughly 74%. A significant part of that increase comes from large unrealized gains recorded by companies such as Alphabet and Amazon on their investments in fast-growing AI businesses including Anthropic.

Without those mark-to-market gains, S&P 500 earnings growth would be closer to 33%, according to LSEG. That would still represent the strongest quarterly performance since 2021, but the difference highlights how AI-related asset valuations are increasingly influencing reported corporate profits.

Amazon recorded approximately $53.4 billion in pre-tax non-operating income, largely tied to its Anthropic investments, while Alphabet reported a $77.1 billion unrealized gain on equity securities.

The broader AI infrastructure boom is also supporting earnings. Goldman Sachs estimates AI infrastructure companies accounted for roughly one-third of S&P 500 earnings-per-share growth during the quarter.

However, investors are becoming more cautious about elevated AI valuations and the enormous financing commitments required to build new data centers and computing capacity. Analysts warn that mark-to-market gains can reverse quickly if private or public AI valuations decline.

Even beyond technology, corporate earnings remain strong, with seven of the S&P 500’s 11 major sectors expected to post double-digit profit growth.

Google Delays Gemini 3.5 Pro as AI Competition Intensifies

/in /tarafından

Google has reportedly postponed the release of its flagship Gemini 3.5 Pro artificial intelligence model after internal testing failed to meet performance targets, highlighting the growing pressure among leading AI developers to deliver increasingly capable models.

According to reports, Gemini 3.5 Pro had originally been expected to launch in June, but development has fallen behind schedule as Google continues improving the model, particularly in software development and coding tasks—an area that has become a key benchmark in the AI race.

The reported delay reflects the rapidly escalating competition between Google, OpenAI, Anthropic, and other frontier AI companies. Rather than releasing models on fixed timelines, developers are increasingly prioritizing capability, safety, and enterprise performance before public deployment.

Sources cited in the report indicate that Google updated Gemini’s training data in an effort to strengthen its reasoning and coding abilities, but the improvements reportedly did not reach internal expectations. The news also raised concerns among some employees about maintaining competitiveness as rival models continue advancing.

Google responded by stating that it is actively testing Gemini 3.5 Pro and other upgraded models with selected partners while continuing to work with U.S. government agencies on AI-related matters. The company emphasized that it remains focused on delivering powerful models while maintaining cost efficiency for customers.

The delay also comes during a period of heightened regulatory scrutiny across the AI industry. Recent releases from OpenAI and Anthropic have themselves experienced postponements or temporary restrictions due to national security reviews and safety requirements, demonstrating that technical performance is no longer the only factor influencing launch schedules.

For enterprise customers, coding assistants and advanced reasoning models have become some of the most commercially valuable AI applications. As a result, incremental improvements in programming capability can significantly influence adoption across software development, cloud services, and business automation.

The latest developments suggest that competition among frontier AI companies is shifting from rapid release cycles toward balancing performance, reliability, safety, and regulatory compliance. In this environment, delaying a launch to improve model quality may increasingly become a strategic decision rather than a setback.

Google Cloud leads as AI spending tops $700B

/in /tarafından

Google has emerged as the standout performer in Big Tech’s AI infrastructure race after Google Cloud posted a 63% revenue surge, sharply outpacing rivals and reshaping investor expectations.

The strong growth, driven largely by enterprise AI demand, exceeded both Microsoft Azure and Amazon cloud growth rates, reinforcing Google’s strategy of commercializing its AI stack across chips, cloud and business tools.

Across major U.S. tech giants, projected AI-related capital expenditures now exceed $700 billion this year, rising from prior estimates near $600 billion. Google raised its own spending outlook further as demand continues to outstrip available compute capacity.

The market response highlighted a growing divide: investors are increasingly rewarding companies converting AI spending into visible revenue acceleration, while punishing those where returns remain less clear.