DOGE AI Tool Targets Massive Regulatory Rollback

Hey folks, Stay ahead with the latest in AI—let’s dive in!

WHAT YOU’LL READ TODAY

• Google’s Opal: AI Mini‑Apps, No Code Needed

• DOGE AI Tool Targets Massive Regulatory Rollback

• ChatGPT Deemed Unsafe for Therapy, Warns OpenAI CEO

•  Zuckerberg Snags Top OpenAI Talent to Power Meta’s AI Push & more…

Google’s Opal: AI Mini‑Apps, No Code Needed

Google Labs has launched Opal, an experimental no‑code tool that lets users build, edit, and share AI-powered mini‑apps using plain language and visual workflows—ideal for rapid prototyping without writing any code. Read More

Huawei’s CloudMatrix 384 Takes On Nvidia with 384 Chips

At WAIC in Shanghai, Huawei unveiled the CloudMatrix 384 AI supernode featuring 384 Ascend 910C chips. With high‑speed optical interconnects, it delivers roughly 300 PFLOPS, outperforming Nvidia’s 72‑chip GB200 NVL72 on several metrics—though at the cost of much higher power use—as part of Beijing’s push for tech self‑reliance under US export curbs. Read More

AI Referrals Surge 357% in June, Hitting 1.13 Billion

In June 2025, AI platforms drove 1.13 billion referrals to the top 1,000 global websites—a 357% year‑over‑year spike. Referrals to news sites soared 770%. Despite this growth, Google Search still dominates traffic with 191 billion referrals, while ChatGPT alone accounts for over 80% of AI‑driven referrals. Read More

DOGE AI Tool Targets Massive Regulatory Rollback

A new AI tool developed by the Department of Government Efficiency (DOGE)—a federal agency formerly led by Elon Musk—is at the center of a controversial push to slash U.S. regulations. According to a leaked internal presentation obtained by The Washington Post, DOGE is using an “AI Deregulation Decision Tool” to help agencies identify outdated rules and generate a “delete list” of unnecessary regulations. The agency reportedly aims to cut federal regulations in half by September 1st, with the AI tool already having flagged 100,000 out of 200,000 analyzed mandates for removal. Notably, the AI has reviewed over 1,000 regulatory sections in the Department of Housing and Urban Development and drafted every proposed deregulation for the Consumer Financial Protection Bureau.

However, the tool’s aggressive approach is drawing serious scrutiny from experts and federal employees alike. While the White House acknowledges the deregulation effort is in its early stages and that no final plans have been approved, insiders at Housing and Urban Development warn the AI is prone to critical errors—frequently misinterpreting legal language and mistakenly classifying valid regulations for deletion. This echoes previous DOGE AI experiments, such as the VA coding tool that famously hallucinated contract sizes, raising broader concerns about whether this tech is ready to shape national policy.

ChatGPT Deemed Unsafe for Therapy, Warns OpenAI CEO

OpenAI CEO Sam Altman has cautioned users against treating ChatGPT like a therapist, emphasizing that conversations with the AI are not private, confidential, or legally protected. In a recent statement, Altman revealed that many young users turn to ChatGPT for emotional support or coaching, but unlike real therapists, lawyers, or doctors, there's no legal safeguard—like doctor-patient privilege—for AI interactions. He noted that even deleted chats can be recovered for legal or security purposes, making users vulnerable if such conversations are ever subpoenaed.

The lack of legal protections could have serious implications for users seeking support through the platform. Altman stressed that while ChatGPT may feel like a safe space, it’s not built for therapy and doesn’t operate under any legal confidentiality frameworks. In legal disputes, OpenAI could be compelled to share chat histories, raising ethical concerns around data privacy and the misuse of sensitive personal disclosures.

Zuckerberg Snags Top OpenAI Talent to Power Meta’s AI Push

Mark Zuckerberg has hired former OpenAI researcher Shengjia Zhao as Chief Scientist of Meta Superintelligence Labs (MSL), the company’s newest AI initiative. Zhao, credited with co-creating ChatGPT, GPT-4, and OpenAI’s first reasoning model o1, is also said to have co-founded MSL alongside Zuckerberg. His move marks a major win for Meta as it races to build the next generation of AI systems.

Zhao will work alongside Alexandr Wang, the ex-Scale CEO who now serves as Meta’s Chief AI Officer, as part of a growing team of elite recruits. Zuckerberg’s hiring spree has already pulled in over 16 engineers and researchers from tech giants like Google, Apple, and OpenAI—signaling Meta’s serious intent to lead the AI arms race.

That’s all for now — thanks for spending a few minutes with us today. We hope you found something valuable to carry into your week. Got thoughts, feedback, or just want to say hi? We’d love to hear from you. Until next time, stay curious and keep moving forward.

-Team AI Paradox