Google blocks Meta's Gemini access while Apple hikes hardware prices

Google has reportedly restricted how Meta can use its Gemini artificial intelligence models.

The decision, first reported by the Financial Times, underscores the growing tension between the two Silicon Valley giants.

Both companies are heavily investing in consumer-facing AI features and developer ecosystems.

This move is expected to force Meta to rely more heavily on its own open-source LLaMA models or alternative partners.

Wayfinder Router has launched to solve the hybrid LLM deployment challenge.

The tool provides a deterministic routing layer to decide whether a query should run locally or go to the cloud.

This approach helps developers control API costs while maintaining low latency for simple tasks.

By running lightweight queries locally and heavy reasoning tasks on hosted endpoints, applications gain significant efficiency.

SoftBank's Masayoshi Son has challenged Elon Musk's proposal to launch AI data centers into Earth's orbit.

Son noted that the math behind space-based computing does not add up, given that power is only a fraction of overall costs.

He emphasized that the physical infrastructure, cooling, and latency challenges of space make terrestrial centers far more practical.

According to Son, the critical phase of the AI race will be decided on Earth within the next few years.

Developers are raising alarms over an unresolved issue in the OpenAI Codex repository.

GitHub issue #2847, which asks for a robust way to exclude sensitive files, remains open.

The lack of a native exclusion mechanism makes it difficult to prevent proprietary code from being ingested.

Developers are calling for clearer controls to ensure local secrets and sensitive files are strictly ignored by AI assistants.

Consumers are footing the bill for something we didn’t ask for, despite record earnings.

The price hikes are “basic economics,” says Tim Derdenger, associate professor of marketing and strategy at Carnegie Mellon University’s Tepper School of Business. As the tech industry has raced to win the AI war, “the price of RAM has skyrocketed because the memory manufacturers have reallocated their production lines to produce new HBM memory for AI data centers and away from consumer DDR5.”

But this isn’t some fluke, or temporary supply chain problem. Companies are choosing data center clients over ordinary buyers because “the same chip earns far more inside an AI server than inside a consumer device,” according to Srikanth Jagabathula, professor of technology, operations, and statistics at the NYU Stern School of Business.

Companies like OpenAI, Google, and Microsoft have thrown around unprecedented amounts of money, outbidding companies like Apple for RAM and storage, creating what even Sam Altman has admitted is a bubble . This imbalance has led to record earnings for companies like Micron , which manufactures memory chips.

But Apple has posted record earnings for at least four quarters in a row , and its margins on hardware sales are much higher than the industry standard. Its markups are estimated to be between 30 and 40 percent , depending on the product. TechInsights and The Wall Street Journal estimate that it’s even higher on the iPhone 17 Pro, perhaps as much as 47 percent .