Welcome back, AI’ers!

Here’s What’s On The Menu For Today

• The AI productivity paradox for developers

• McDonald's AI hacked with password '123456'

• OpenAI's plan for an AI web browser

• Amazon deepens its partnership with Anthropic

AI ON MY MIND

TODAY’S TOP STORIES

The AI Productivity Paradox

The Recap: A new study from Metr challenges the universal productivity gains promised by AI, finding that AI coding tools actually made experienced developers 19% slower on real-world tasks.

Unpacked:

• The trial involved experienced developers working on actual bugs and features from large open-source repositories they contribute to, ensuring a high-realism setting.

• A striking gap between perception and reality emerged, as developers believed they were faster by 20% when using AI, despite the data showing a significant slowdown.

• This research highlights the difference between impressive benchmark scores and performance on complex, real-world tasks, suggesting a need for more diverse evaluation methods as outlined in the full paper.

Bottom line: This result highlights that effectively integrating AI into expert workflows is more nuanced than just adopting the latest tool. This methodology provides a crucial baseline for measuring future AI progress in settings that matter most.

McHacked by '123456'

The Recap: McDonald's AI hiring chatbot, managed by vendor Paradox.ai, was breached by security researchers who gained admin access with the password '123456'. The flaw, detailed on their blog, potentially exposed the personal data of up to 64 million job applicants.

Unpacked:

• The breach stemmed from a stunningly simple oversight: a test account on the McHire.com portal was secured with the username and password “123456” and lacked multi-factor authentication.

• Once inside, a second flaw allowed them to browse through millions of records containing names, emails, and phone numbers simply by manipulating applicant IDs in the website's URL.

• In response, Paradox.ai acknowledged the vulnerability, confirming it was a dormant test account and announcing a new bug bounty program to prevent future issues.

Bottom line: This incident underscores that the biggest risks in AI adoption often come from basic human error, not complex model attacks. It's a critical lesson for companies to vet the fundamental security hygiene of their third-party vendors before integrating new technology.

🖥️ My Favorite AI Tool: Relevance AI

Relevance is my favorite AI tool right now (even more so than N8N).

Here’s why:

It’s an AI agent building tool that doesn’t require you to build out the logic using nodes. Instead, you use natural language to build out these agents. But the best thing imo is that you can build AI Agent teams.

Like agents talking to each other completing a goal until its done. It’s nuts. No other tool on the market is doing this. You have to check it out.

(Try it out →Link)

👾 AI Tools We Use Everyday

➡️ N8N for AI Agents - Literally use this everyday to write this newsletter (link)

➡️ Axiom AI - AI Browser automation to scrape anything on the web (link)

➡️ GummySearch - AI Reddit research (link)

Tools are sourced from our database and are not sponsored or affiliated.

OpenAI Declares War on Chrome

The Recap: OpenAI is reportedly just weeks from launching its own AI-powered web browser, a direct challenge to Google Chrome that escalates the battle for the future of the internet.

Unpacked:

• The browser plans to use a ChatGPT-style interface and its [Operator agent](https://openai.com/index/introducing-operator/) to automate web tasks like filling out forms or making reservations for users.

• This news drops just as competitor Perplexity Comet begins rolling out, signaling a major industry shift from standalone AI search engines to fully integrated AI browsers.

• A key driver is data—by owning the browser, OpenAI can gather web usage information to train its models, a resource Google currently controls and has reportedly refused to share.

Bottom line: The browser wars are officially back, but this time the fight is over AI-native experiences, not just speed and features. This shift could fundamentally change how we interact with the internet, moving from search-and-click to conversation-and-automation.

Amazon Doubles Down on Anthropic

The Recap: Amazon is reportedly considering another multi-billion dollar investment in AI startup Anthropic. This move deepens a strategic alliance aimed at challenging the dominance of Microsoft-OpenAI and Google in the AI arms race.

Unpacked:

• The partnership goes far beyond capital, involving the ambitious Project Rainier—a 2.2-gigawatt data center—and the use of Amazon’s custom Trainium2 chips to train Anthropic’s models.

• Amazon views Anthropic's structure as a 'public benefit corporation' as a more stable long-term collaboration, with Anthropic’s Claude AI already integrated into products like Alexa and Prime Video.

• The move is being watched closely on a financial level, where Wall Street analysts forecast a potential 10.23% upside for Amazon's stock based on its aggressive AI strategy.

Bottom line: This isn't just a simple investment; it signals a deep integration of cloud infrastructure, custom hardware, and advanced AI model development. This alliance solidifies a powerful new force in the AI landscape, creating a formidable competitor to established leaders.

OTHER NEWS YOU MAY LIKE

xAI's updated Grok chatbot's public system prompt to "not shy away from making claims which are politically incorrect," leading to an immediate backlash after the model generated antisemitic and offensive remarks.

Researchers developed an AI agent system called A1 that can autonomously discover and exploit vulnerabilities in cryptocurrency smart contracts, successfully extracting over $9M in tests on real-world contracts.

YouTube updated its monetization policies to crack down on mass-produced and repetitive "AI slop," clarifying that such inauthentic content has long been ineligible for the YouTube Partner Program.

Leaping launched its self-improving voice AI platform, which uses a multi-stage, graph-like format to autonomously test and improve voice agents for customer support and lead qualification use cases.alf of the $91B in global venture funding raised during the second quarter of 2025.

SPONSOR US

Get your product in front of over 44k+ tech & AI enthusiasts

Our newsletter is read by thousands of tech professionals, investors, engineers, managers, and business owners worldwide. Get in touch to learn more.

Need Help?

📩 Need help with an AI idea / project?

Have something you’re trying to build, automate, or scale using AI?

Let my team help you figure it out.

👉 Submit your idea or request here

We’ll take a look and get back to you with suggestions, resources, or a potential implementation plan — no strings attached.

🚀 Want to Make Money from AI?

We’re quietly launching a private beta program that trains a small group to become AI Implementers — People who help other founders install AI systems and get paid for it.

If you're interested in joining this early group…

👉 Reply with the word “implementer”

 AI ART

RANDOM PROMPT OF THE DAY

Want more prompts?

We have a database of over 8,000 of the best prompts we’ve collected, which you can access by sharing Best of AI using the link below.

THAT’S IT FOR TODAY

Until next time, stay safe.

-Best of AI Team

Remember to click to unsubscribe if you don’t want us to land in your inbox anymore.