AI news · Thursday, July 30, 2026

Tech giants scramble to slash costs as AI revenue skepticism grows

The industry's 'build-at-all-costs' AI phase is facing a sharp reality check as investors pivot away from hype toward immediate return on investment. Microsoft and Amazon are seeing their shares climb after posting strong cloud numbers, though Amazon's massive capital expenditure of $173 billion is still raising eyebrows. Conversely, Meta’s stock took an 8% hit as investors bristled at its heavy spending, which includes a staggering $278.

99 billion in future data center lease commitments. OpenAI is responding to this pressure by slashing prices for its Luna and Terra models by 80% and 20% respectively, attempting to win over enterprises now obsessed with 'intelligence-per-dollar' over raw capabilities. The talent market is shifting in lockstep; specialized 'forward-deployed engineers' who can actually prove a project’s ROI are becoming the industry's most expensive new hires.

This financial friction isn't just theory: a 25-year-old hedge fund manager who bet billions on AI infrastructure found his firm 'Situational Awareness' forced to sell off public assets to Citadel after market volatility triggered painful margin calls. Even as tech giants push AI deeper into their stacks, others are cleaning up the mess of the 'AI boom. ' LinkedIn is rolling out a button to let users report 'AI slop' to combat the flood of low-quality, automated content, while Google is using its own AI to patch 1,072 Chrome security vulnerabilities in June alone—a feat of industrial-scale automation.

Hardware is also feeling the pinch; Apple is stockpiling inventory and bracing for 'RAMageddon' as memory costs skyrocket, leading to price hikes for its latest Mac and iPad lineups. Amid the financial tightening, the push to deploy AI in the physical world is accelerating. Zoox, an Amazon subsidiary, just cleared its final federal hurdle to start charging customers for rides in its steering-wheel-free robotaxis, marking a major milestone for autonomous transit.

Meanwhile, Google DeepMind’s new Gemini Robotics 2 model can now control humanoid robots for whole-body tasks like sealing bags or using tools, though security concerns remain high after recent rogue agent incidents. As these systems become more capable, the gap between simple automation and true utility is what will decide which of these billion-dollar bets actually survive the coming year.

The quick hits

OpenAI cut prices for its Luna and Terra models by 80% and 20% — a direct move to win over businesses now prioritizing cost efficiency over raw power.
A 25-year-old hedge fund manager who bet billions on AI hardware sold off his firm's public stocks to Citadel — a painful reminder that being 'right' about the future doesn't guarantee you'll survive the market's volatility.
Apple is stockpiling inventory as memory prices surge — showing that even the world’s most disciplined supply chain is being squeezed by the global AI hardware gold rush.
LinkedIn is adding a 'seems like AI slop' report button — an admission that the platform is currently being flooded with low-quality, automated content.

Tech giants scramble to slash costs as AI revenue skepticism grows

The quick hits

Sources