☕️ Apple's foldable iPhone faces engineering issues

Hi there, this is your daily ☕️ Techpresso.

In today's Techpresso:

📱 Apple's foldable iPhone faces engineering issues

🤖 Meta to open-source new AI models

📦 Amazon strikes USPS deal keeping 80% of package volume

🏭 Intel joins Elon Musk's $25B Terafab AI chip project

⚡ Anthropic doubles down on Google Cloud TPUs

🏆 Meta employees compete on internal AI usage leaderboard

Plus: 🎁 17 other news you might like, 🧰 6 tools, and 📚 3 papers.

📱 Apple's foldable iPhone faces engineering issues LINK

Apple's foldable iPhone has hit more engineering problems than expected during early test production, and those issues could delay the device's first shipments by several months.

The problems came up during production verification tests, which is the fourth of six steps Apple products must pass before shipping, and component suppliers have already been told about schedule delays.

Apple had been aiming to launch its foldable iPhone in fall 2026 alongside the iPhone 18, but the timeline could now slip into 2027, with April and early May being critical for testing.

🤖 Meta to open-source new AI models LINK

Meta plans to release open-source versions of its next-generation AI models, which are derived from two proprietary frontier models codenamed Avocado and Mango expected to launch this year.

The open-source versions won't include all features found in the closed-source editions, possibly lacking certain neural networks, having smaller parameter counts, or skipping post-training steps.

AI safety is reportedly one reason Meta will hold back features, and the company does not expect its upcoming models to beat competitors like Anthropic and OpenAI across the board.

📦 Amazon strikes USPS deal keeping 80% of package volume LINK

Amazon and the US Postal Service have reached a tentative deal that will reduce USPS package deliveries by 20 percent, far less than the two-thirds cut Amazon originally threatened earlier this year.

Amazon is the USPS's largest customer, bringing in $6 billion in revenue, so even a 20 percent reduction could still mean more than $1 billion in lost revenue for the agency.

The USPS re-engaged with Amazon after bids from rival companies fell short of its volume and revenue expectations, though the new agreement still needs approval from the federal Postal Regulatory Commission.

🏭 Intel joins Elon Musk's $25B Terafab AI chip project LINK

Intel has officially joined Elon Musk's Terafab project, a $20–25 billion semiconductor complex planned for Austin, Texas, partnering alongside Tesla, SpaceX, and xAI to build chips at scale.

The facility aims to produce 1 terawatt per year of compute capacity by manufacturing edge-inference processors for Tesla's FSD systems and radiation-hardened chips for SpaceX satellites and xAI.

Intel CEO Lip-Bu Tan hosted Musk at Intel facilities before the announcement, and the company will contribute its process technology, high-volume fabrication, and packaging expertise to the project.

⚡ Anthropic doubles down on Google Cloud TPUs LINK

Anthropic announced an expanded partnership with Google Cloud, securing access to multiple gigawatts of TPU capacity to train and run its AI models starting in 2027.

The deal delivers Google's Tensor Processing Units through Google Cloud infrastructure with hardware from Broadcom, giving Anthropic enormous compute for its Claude family of AI systems.

Anthropic is also adopting Google Cloud tools like BigQuery, Cloud Run, and AlloyDB, while thousands of companies already access Claude models through Google Cloud today.

🏆 Meta employees compete on internal AI usage leaderboard LINK

Meta has an internal leaderboard called "Claudeonomics" where employees compete to consume the most AI tokens, tracking usage across more than 85,000 workers on the company intranet.

Employees burned through 60 trillion tokens in just 30 days, with the top user averaging 281 billion, though some simply leave AI agents running for hours to pad their numbers.

Despite Silicon Valley treating "tokenmaxxing" as a productivity metric, nobody has put up hard numbers proving that high token consumption actually translates into real business results or revenue gains.

Other news you might like

• Google’s free offline dictation app just made paying $15 a month for Wispr Flow hard to justifyLINK

• Gemini is making it faster for distressed users to reach mental health resourcesLINK

• OpenAI, Anthropic, and Google team up against unauthorized Chinese model copyingLINK

• China actively targeting Taiwan's chip talent and technology, security report saysLINK

• OpenAI asks California, Delaware to investigate Musk's 'anti-competitive behavior' ahead of trialLINK

• Apple plans Supreme Court appeal in Epic Games App Store battle — againLINK

• Three YouTubers accuse Apple of illegal scraping to train its AI modelsLINK

• Bezos' Project Prometheus hires xAI co-founder from OpenAILINK

• Teardown of unreleased LG Rollable shows why rollable phones aren't a thingLINK

• From folding boxes to fixing vacuums, GEN-1 robotics model hits 99% reliabilityLINK

• “The problem is Sam Altman”: OpenAI Insiders don’t trust CEOLINK

• AI startup Rocket offers vibe McKinsey-style reports at a fraction of the costLINK

• US lawmakers aim to ban export of DUV chipmaking and etching tools to leading firms in China — bipartisan proposal would ban lithography equipment for Huawei, SMIC, and othersLINK

• OpenAI launches safety fellowship to advance AI alignment research and assess effects on societyLINK

• Artemis II astronauts capture stunning views of Earth and the moon on historic voyageLINK

🧰 Trending tools

Lessie AI: an AI-powered people search engine that queries 50M+ profiles across LinkedIn, Twitter, GitHub, and other platforms to find leads, talent, and investors.LINK

OpenOwl: a computer-using agent that automates desktop tasks by visually navigating screens, clicking, typing, and extracting data across applications for AI assistants.LINK

SuperShrimp: a webcam-based posture monitoring tool that tracks your sitting position in real-time, sending alerts when you slouch to prevent long-term back pain.LINK

Bibby AI: an AI-powered LaTeX editor that accelerates academic writing with smart citations, paper reviews, and collaborative tools for researchers and students.LINK

Google AI Edge Eloquent: an on-device AI assistant app that runs language models locally on your phone for private, offline conversational AI interactions.LINK

GovAuctions: aggregates government surplus and seizure auctions from multiple sites into a single searchable platform with filters and alerts.LINK

📚 Trending papers & reports

Agentic AI exposure analysis finds that 86% of occupational tasks across multiple regions could be automated by AI systems capable of independent decision making and tool use.LINK

Developer productivity with AI coding tools increases by 26% in speed but produces 41% more bugs, based on controlled experiments measuring both completion time and code quality across multiple tasks.LINK

PostgreSQL vector search performs 2 to 10 times slower when filtering results compared to specialized vector databases, revealing a significant performance gap for AI applications needing filtered semantic search.LINK

Want to get the latest news differently? Find us on:

See you tomorrow for a new dose of ☕️ Techpresso!