Bots Online, Agents On-Device
· The Fluency Briefing
The Fluency Briefing
Your Guide to What's Happening in AI and Why It Matters to You
Thursday, June 4, 2026
Everyone assumes the internet is made by humans, for humans. As of this week, that's officially wrong. Cloudflare's CEO confirmed bots now generate more web traffic than people do - a milestone nobody expected until 2027. Meanwhile, courts are scrambling to figure out what happens when ChatGPT plays lawyer, Apple is betting its entire AI reputation on a privacy-first Siri reboot, and Stanford just dropped a framework that could make your phone smarter than most cloud services.
Today in AI:
- Humans Are Now the Minority Online - Cloudflare CEO Matthew Prince confirmed that bot traffic has surpassed human traffic on the internet for the first time, hitting a 57.5% to 42.5% split. He'd predicted this wouldn't happen until 2027. Agentic AI - bots shopping, comparing flights, scraping prices - is the main driver. Tom's Hardware
- ChatGPT Goes to Court, Gets Mixed Reviews - Federal courts are split on whether AI-generated legal documents count as protected work product. One Michigan court said yes; a New York court said no. Meanwhile, an insurance company is suing OpenAI, alleging ChatGPT practiced law without a license. MIT Technology Review
- Apple's WWDC Could Finally Deliver the Siri We Were Promised - Apple's keynote on June 8 is expected to unveil the long-delayed AI-powered Siri, running on Google's Gemini models with a privacy-first architecture. This will also be Tim Cook's last keynote before handing the CEO role to John Ternus in September. Engadget
- Apple's Privacy Gambit: Encrypted AI on Nvidia Chips - Apple's deal with Google reportedly requires Gemini models to run on Nvidia chips that support encrypted processing, meaning Google can't peek at your queries. Three tiers of privacy - on-device, Apple's servers, then Google Cloud - aim to make AI feel safe. 9to5Mac
- Stanford's OpenJarvis Puts a Personal AI Agent on Your Phone - Stanford researchers released OpenJarvis, an open-source framework that runs AI agents, memory, and learning entirely on-device. It performs within 3.2 points of the best cloud model at roughly 800 times lower cost. Your phone just got a lot more capable. MarkTechPost
- Bezos Bets $500M on Brain-Inspired AI - A startup called Flourish, backed by Jeff Bezos, wants to reverse-engineer the brain's efficiency to fix AI's energy problem. The human brain runs on 20 watts; a single AI chip uses 30 times that. At a reported $2.5 billion valuation, it's neuroscience meets Silicon Valley ambition. Wired
- Computex 2026: AI Drives Hardware Prices Up, Innovation Sideways - This year's Computex theme was "AI Together," but the real story is AI's appetite for hardware driving up RAM and storage prices. AMD re-released a four-year-old CPU, and new SATA SSDs are making a comeback. Your next PC build might be weirdly retro. Tom's Hardware
- A $65,000 AI School That Isn't Actually a School - Alpha School opened a Manhattan campus charging $65,000 a year, but New York's education department declined to approve it as a school. The state found its AI-powered "2 Hour Learning" model had "little to no supervision or competent teacher delivering instruction." It now operates as a homeschooling center. Wired
Today's Takeaway:
Here's the thing about bots surpassing humans online: it doesn't just mean more automated traffic. It means the internet itself is becoming a place built primarily for machines to talk to other machines. According to Tom's Hardware, Cloudflare's data shows the split at 57.5% bot versus 42.5% human. These aren't the spam bots of old - they're AI agents comparing flight prices, reading product pages, handling customer service, and shopping on your behalf. Translation: the web is becoming an API that humans occasionally visit.
Now layer in a few other stories from Thursday. Stanford's OpenJarvis wants to run those same kinds of agents locally on your phone, not in the cloud. Apple is restructuring its entire AI stack around privacy tiers specifically because it knows agents will be handling your most personal requests. And courts, as MIT Technology Review reports, are already dealing with the fallout of people treating chatbots as their personal lawyers. The pattern is clear: AI agents are becoming the primary way we interact with digital infrastructure, and our institutions - legal, educational, commercial - are scrambling to catch up. The bots didn't just cross a traffic threshold. They crossed a cultural one.
🔍 Myth Buster
The myth: "The internet is still mostly human - bots are a future problem"
The reality: Bots have already surpassed human internet traffic for the first time in history, according to Cloudflare's CEO, and this tipping point arrived ahead of schedule - analysts didn't expect agentic AI traffic to eclipse real human traffic until 2027. This isn't a gradual trend; the explosion of autonomous AI agents fetching data, filing requests, and interacting with services has fundamentally changed the composition of the web faster than even industry insiders predicted.
The nuance: The concern is legitimate and urgent: when bots dominate traffic, it distorts analytics, strains infrastructure, inflates costs for web services, and makes it harder to design experiences for actual humans - problems that are already materializing, not hypothetical.
The Bottom Line
The Pattern: AI isn't arriving someday - it's already the dominant force on the internet, in courtrooms, and increasingly on your personal devices. Thursday's stories all point to the same uncomfortable truth: the world is being reorganized around AI agents, and humans are adapting to them, not the other way around.
Why It Matters: If bots are already the majority of web traffic, every business with an online presence needs to think about who - or what - is actually visiting their site. If courts can't agree on whether ChatGPT output is privileged legal work, anyone using AI for anything consequential is operating in a gray zone. The rules are being written in real time, and ignorance isn't a defense.
Your Move: Pick one thing you do online every week - price comparison, research, scheduling - and ask yourself whether an AI agent could do it better. Then ask: what does your business look like when your customers start doing the same thing?
What We're Working On
✨ Founding Cohort Special - 60% Off! - Use code MAF20 to join for just $20/month (regularly $50). Get weekly group sessions & workshops, self-paced courses for all levels, access to tools & templates, challenges with peer feedback, and 24/7 support community. → Join Now
✨ Free 30-Minute AI Consultation - Discover how My AI Fluency can help your business unlock the potential of AI. We'll discuss your goals, explore practical AI opportunities for your industry, and outline clear next steps. → Schedule Free Call
✨ How AI-Fluent Are You? - Test your AI fluency with our interactive quiz. See how you stack up and discover what to learn next. → Take the Quiz
💬 Community | 📞 Book a Consultation | 🌐 Website
Fluently yours, The My AI Fluency Team