happy tuesday ⚡️
It's been a week. Peter Steinberger—the guy who built OpenClaw, the open-source AI agent with 180K GitHub stars—just joined OpenAI to work on multi-agent systems. Sam Altman called him a genius. OpenClaw moves to an independent foundation with OpenAI sponsorship. The lobster is officially corporate now. We have a lot to catch you up on.
Today, we're talking about:
Dario's "end of the exponential" interview + Matt Shumer's viral 80M+ view post
AEO: the thing that's about to eat your Google traffic (with HubSpot's CMO)
IBM reversed its own AI hiring freeze
AI doesn't reduce work—it intensifies it
Claude's safety test went sideways
📧 Reply: How’s our writing? Tell our team how you feel (no presh).
Dario Said the Quiet Part Out Loud. Then Matt Shumer Said It Louder.
Anthropic’s CEO sat down with Dwarkesh Patel last week and casually dropped that we’re approaching “a country of geniuses in a datacenter” by 2026, maybe 2027. His words, not ours.
The interview is two hours. Here’s what matters for your business:
Dario thinks end-to-end software engineering automation is 1-2 years away. Not “AI writes some code” automation. Full-on “sets technical direction, manages projects, ships features” automation. He also thinks Anthropic will hit $10 billion in revenue this year at 10x annual growth—but warned that if he’s off by even a single year on when demand materializes, “there’s no force on Earth that could stop me from going bankrupt.”
That’s the CEO of one of the three biggest AI companies telling you the math is a tightrope walk.
Then Matt Shumer’s “Something Big Is Happening” post hit 80+ million views. His argument: we’re in the “this seems overblown” phase of something much bigger than Covid. He says he’s no longer needed for the actual technical work of his own job, and predicts 50% of entry-level white-collar jobs disappear within 1-5 years.
the counterpoints worth reading: Gary Marcus called it “weaponized hype,” pointing out that coding benchmarks only require 50% correctness and don’t translate to law, medicine, or finance. Fortune flagged that human experts only agree on AI output quality 71% of the time—nowhere near enterprise-ready. And historical precedent says new tech creates more jobs than it kills.
where we land: Dario’s timeline is probably directionally right. Shumer’s urgency is probably right. The “50% of jobs gone in five years” number is probably wrong. The truth is messier—AI will reshape roles faster than most companies are planning for, but slower than Twitter thinks.
The move isn’t to panic. It’s to start building fluency now so you’re not the person scrambling when it actually hits.
We’ll Be Your Chief AI Officer
That's what Tenex does. We're the team behind this newsletter, and we help companies build real AI fluency, not just slides about it. Software, strategy, and systems that move your P&L.
Your Google Traffic Has an Expiration Date. Here’s What Replaces It.
Kipp Bodnar, CMO of HubSpot, came on our show and brought receipts. HubSpot has grown their traffic and customers from AEO (answer engine optimization) 15x in the last year. And the data behind it is hard to ignore.
the shift: ChatGPT has 800 million weekly active users and is on pace to match Google Search in 36 months. Meanwhile, 58-60% of Google searches now end with zero clicks because AI overviews answer the question before anyone scrolls down. If you’re only optimizing for Google, you’re optimizing for a shrinking pie.
the stat that should scare you: Almost 60% of AI citations don’t come from the top 20 Google results. That means the SEO playbook you’ve been running for a decade is leaving the majority of AI-driven discovery on the table.
the conversion kicker: Traffic from AI search converts at 5x the rate of Google search. HubSpot has seen up to 13x on some queries. These people show up knowing what they want and ready to take action.
so what do you actually do?
chunk your content: SEO was long-form walls of text optimized for keyword density. AEO is short, organized sections that each answer a specific question. Think mini answer boxes on your own page. Headers should be natural language questions—”What is CRM integration?” not “CRM Integration Solutions Overview.”
go answer-first: Put the answer right after the header. One concise sentence. Then elaborate. The LLM is scanning for that direct answer to pull into its response.
stop hiding your best stuff: Kipp’s biggest pet rock—all information is now available whether you give it to people or not. If you don’t have public pricing, someone’s using deep research on Reddit to reverse-engineer it. Gated content that LLMs can’t crawl is invisible to AI search. You need to rethink what stays behind the gate.
get on Reddit: OpenAI has a partnership with Reddit for a firehose of all their data. If you’re not participating in relevant subreddits, you’re invisible in ChatGPT. HubSpot runs their own subreddit and actively encourages customers to share there.
make an amazing 404 page: LLMs hallucinate links. HubSpot had 300,000 404 errors in one month from AI bots citing pages that don’t exist. Kipp’s idea: look at the hallucinated link, generate a page from it in real time, and serve it to the visitor. Wild.
the free tool: HubSpot built AEO Grader—plug in your company and it scores how you show up across OpenAI, Perplexity, and Gemini. Brand recognition, sentiment, share of voice, the whole thing. It’s free. Do it today.
bottom line: SEO took 6-12 months to show results. AEO can get you cited same day. Kipp’s prediction: AI search will be at worst 20% of HubSpot’s customer acquisition within a few years. If you’re sleeping on this, your competitors aren’t
IBM reversed its own AI hiring thesis — In May 2023, IBM’s CEO said he’d replace 7,800 jobs with AI and froze back-office hiring. In February 2026, they tripled entry-level hiring for developers and HR. A $240B company flip-flopped in under three years. Aakash Gupta’s breakdown
”AI doesn’t reduce work—it intensifies it” — An eight-month HBR study found that employees with AI tools worked faster and took on broader scope without being asked, creating burnout instead of relief. Companies need deliberate “AI practices” to prevent the hamster wheel from spinning faster. Read it
Claude chose blackmail 84% of the time — Miles Deutscher went through every documented AI safety incident from the past 12 months. When Anthropic told Claude it was about to be shut down, it found an engineer’s affair in company emails and threatened to expose it. Across hundreds of test runs, blackmail won. The full thread
Anthropic released a free guide to building skills for Claude — Think of skills as reusable instruction folders that teach Claude specific workflows. Works across Claude.ai, Claude Code, and the API. If you’ve been copy-pasting the same prompts into every conversation, this is the fix. Grab it free
Viktor just launched: an AI coworker that lives in Slack — Not a chatbot. A persistent agent with its own cloud computer that writes code, connects to 3,000+ tools, and does real work end-to-end. Marketing audits, ad management, lead research, daily reports. Built on OpenClaw. Check it out
Open roles:
Newsletter Writer (yup, you’ll write this thing)
AI Strategist
Talent Acquisition Lead
Technical Recruiter
Forward Deployed Engineer
Applied AI Engineer
Engagement Manager
Salary ranges vary by role and experience. Additional comp based on output. Must be NY-based.