Welcome back to Daily Zaps, your regularly-scheduled dose of AI news ⚡️ 

Here’s what we got for ya today:

• 🎆 ChatGPT free image generator goes viral

• 💻 Intel, TSMC chipmaking joint venture

• 🤖 Google DeepMind’s 145 page paper on AGI

• 🧑‍🎓 Claude for education

Let’s get right into it!

Sign Up | Advertise | Tools

STARTUPS

ChatGPT free image generator goes viral

OpenAI has made its GPT-4o-powered image generator available to all ChatGPT users, including those on the free tier, though daily usage may be limited to three images. The tool quickly gained popularity, with demand so high that it strained the company's GPUs, and sparked controversy over copyright issues, particularly for images mimicking Studio Ghibli’s animation style.

Some users have also exploited it to create fake receipts, prompting OpenAI to embed metadata in generated images and enforce its usage guidelines. Alongside this rollout, OpenAI announced it raised $40 billion in funding led by SoftBank at a $300 billion valuation, and reported that ChatGPT now boasts 500 million weekly and 700 million monthly active users.

BIG TECH

Intel, TSMC chipmaking joint venture

Intel and TSMC are in preliminary talks to form a joint venture that would give TSMC a 20% stake in Intel’s chipmaking facilities in exchange for sharing its manufacturing expertise and training Intel personnel, rather than contributing capital. The deal, reportedly initiated and supported by the White House and the U.S. Department of Commerce, aims to revitalize Intel’s struggling foundry business, which posted an $18.8 billion loss in 2024.

While Intel and other U.S. firms would retain majority control, internal resistance remains due to concerns over potential layoffs, loss of technological independence, and the possibility of TSMC effectively taking over Intel’s manufacturing operations. TSMC’s dominance in advanced chip production, coupled with geopolitical tensions involving Taiwan, adds further complexity to the deal.

BIG TECH

Google DeepMind’s 145 page paper on AGI

Google DeepMind has published a 145-page paper detailing its approach to AGI (Artificial General Intelligence) safety, predicting the emergence of highly capable AGI systems—defined as matching the top 1% of skilled humans on various tasks—by 2030, potentially leading to “severe harm” or even existential threats.

The paper, co-authored by DeepMind co-founder Shane Legg, critiques the safety strategies of rival labs like Anthropic and OpenAI, arguing that robust training, monitoring, and secure deployment environments are more effective than automation-heavy approaches. While skeptical about the near-term emergence of superintelligent AI, DeepMind warns of the dangers of recursive AI self-improvement and proposes technical safeguards to limit AGI misuse and improve system transparency.

Download 145 page PDF of paper here.

STARTUPS

Claude for education

Anthropic has launched "Claude for Education," a new tier of its AI chatbot Claude tailored for higher education, mirroring OpenAI’s ChatGPT Edu plan. This offering provides students, faculty, and staff with access to Claude, including a “Learning Mode” designed to foster critical thinking through interactive questioning and educational templates. With enterprise-grade privacy and integrations like Canvas via a partnership with Instructure, and support from Internet2, Anthropic aims to expand AI adoption across campuses.

It has already secured agreements with institutions like Northeastern University, LSE, and Champlain College, and plans to grow through student ambassador programs. As AI's role in education remains debated, Claude for Education positions Anthropic to capture a larger share of the academic AI market, supporting its goal to double revenue in 2025.